Unique space time adaptive system (USS)

Description

CROSS REFERENCE TO RELATED APPLICATIONS

This application claims the benefit, Under 35 U.S.C. § 119(e), of U.S. Provisional Application No. 60/677,576, filed May 4, 2005, which is hereby incorporated by reference.

BACKGROUND OF THE INVENTION

1. Field of Invention

The field of the invention relates generally to radars and more specifically to a radar mounted on a moving platform employing an electronic scanned array with a transmission array and receive array (channel) or arrays (channels).

2. Description of the Related Art

In the field of this invention, detecting radar returns as moving targets and rejecting others such as clutter and others has been a challenge for many years. Obtaining a moving object's velocity, azimuth, and measurement of its parameters have been an objective of many radar systems. In relatively recent years with the increased processing and storage of improved integrated circuits, space time adaptive processing (“STAP”) has become more practical.

U.S. Pat. No. 5,563,601 (the “'601 patent”) issued Oct. 8, 1996 and entitled TWO PORT CLUTTER SUPPRESSION INTERFEROMETRY SYSTEM FOR RADAR DETECTION OF MOVING TARGETS is incorporated, in it's entirety, by reference herein. The '601 discloses, in part, a two port radar system for detecting and measuring range, azimuth and velocity of radar returns. This patent utilizes the detection of shadows to locate the targets azimuth and not employing any other technique in combination with it and it is meant for land clutter and not sea clutter.

U.S. Pat. No. 6,633,253 (the '253 patent”) issued and entitled DUAL SYNTHETIC APERTURE RADAR SYSTEM is incorporated, in its entirety, by reference herein. The '253 patent discloses, in part, a system which utilizes an electronic scanned array with two receive channels (arrays) on an airborne platform. The system employs displaced phase center antenna (DPCA) techniques to cancel clutter and detect moving targets and measures its range, velocity and azimuth accurately.

SUMMARY

The UNIQUE SPACE TIME ADAPTIVE SYSTEM (USS) system is unique in a number of ways. For example, it does not require clutter cancellation of any kind and it may employ as few as one channel or as little as one pulse. It is as accurate or more accurate as any known STAP system, hardware and dwell time and processing are a minimum. It utilizes a combination of techniques, but basically employs single or two or more equal receive arrays or pulses with or without DPCA methodology, detects returns and measures their velocity, azimuth and range. It also takes different looks in time to measure differences with time, for purpose of determining radial velocity unambiguously and tangential and vertical tangential velocity.

In one embodiment, a system utilizes one, two, or more equal receive channels processed with “M” pulses of time data. In another embodiment, the system implements as few as one or two pulses of time data and has “N” channels of data employed and processed.

Both systems will detect a return or number of returns in the same range doppler or range azimuth bin such as clutter, target, noise, jamming, etc and identify them out measure there three dimensional position and velocity. This is performed without employing any clutter cancellation of any kind.

The returns are processed to detect returns and detect moving targets of interest and reject unwanted returns such as clutter, sidelobes, movers, multipath returns, others.

USS does not require clutter cancellation therefore there is no clutter covariance matrix or training data involved or knowledge aided information. Channel matching is not required. USS has the ability to handle more returns than clutter and target per range doppler bin (RDB) or range azimuth bin (RAZ) and determine there range, velocity and azimuth accurately. Thresholding and post processing of the returns will determine if returns are clutter, moving target, antenna sidelobes, filter sidelobes, multipath returns, etc. It has the ability to handle ground moving targets, high speed targets, ship detection and identification in a unique and simplified manner.

BRIEF DESCRIPTION OF DRAWINGS

So that the manner in which the above recited features of the present invention can be understood in detail, a more particular description of the invention, briefly summarized above, may be had by reference to embodiments, some of which are illustrated in the appended drawings. It is to be noted, however, that the appended drawings illustrate only typical embodiments of this invention and are therefore not to be considered limiting of its scope, for the invention may admit to other equally effective embodiments.

FIG. 1—depicts top views of some of the positions that ships may be as relative to the radar and the position the ship is detected and the actual ship position;

FIG. 1A—depicts the Ship moving towards the radar;

FIG. 1B—depicts the Ship heading away from the radar;

FIG. 1C—depicts the Ship to the left of and perpendicular to the radar;

FIG. 1D—depicts the Ship to the right of and perpendicular to the radar;

FIG. 2—depicts a high level block diagram in accordance with the invention;

FIG. 3—depicts an embodiment of a chart for determination of the height of ships when detected and the solution of the height dependent on various factors;

FIG. 4—depicts an embodiment of a Ship Parameter Measurement Chart—in accordance with the invention;

FIG. 5—depicts an embodiment of a Chart of Proposed Systems and Characteristics in accordance with the invention;

FIG. 6—depicts an illustrative of techniques for all systems such as over ocean, over land, and in the air;

FIG. 7—depicts an embodiment of a Delta T technique with additional DPCA delays in accordance with the invention;

FIG. 8—depicts an embodiment of a Delta D technique with additional time delays in accordance with the invention;

FIG. 9—depicts an embodiment of a block diagram of a system in accordance with the invention;

FIG. 10—depicts an exemplary illustration of a relationship of detection phase to radial velocity phase to azimuth phase in accordance with the invention;

FIG. 11 depicts an exemplary illustration of delta T combined with delta I (interleaved pulses—two in example) and aperture change synchronized with interleaved data in accordance with the invention;

FIG. 12 depicts examples of processing returns with the association of returns processed in adjacent range and/or doppler bins in accordance with the invention;

FIG. 13 depicts an illustrative single aperture system with interleaved pulses with change in aperture in accordance with the invention; and

FIG. 14 depicts an illustrative single aperture system with no interleaved pulses with change in aperture in accordance with the invention.

DETAILED DESCRIPTION

I. A—Basic Concept of a Unique Space Time Adaptive Radar System (USS)

It is to be noted that the description of this disclosure are illustrative and cannot show all possible implementations that may be employed from the information in the disclosure by a person of ordinary skill in the state of the art. Therefore this disclosure is illustrative and not limited of the scope of the proposed invention and not limited in the scope of the role in obtaining the objective of this disclosure.

USS processing system is applicable to all frequencies. The system is able to utilize all radio frequencies limited on the low end by the size of the antenna and on the high end by the practical limitations of short waves.

Illustrated in the patent application is a one dimensional array. The implementation may also be with two dimensional arrays as well.

II. B Basic Operations, Equations and Methodology Fundamental to all of the Systems Employed in this Disclosure

A—Fundamental Techniques Employed:

The mathematical fundamental equations, the radar analysis and the computer simulation for all systems to obtain the objective of the disclosure are presented. The techniques are as follows:

• • 1) Basic System is with one aperture, one PRF and one transmission frequency. If the ambiguous range and or velocity are to be increased another PRF and/or transmission frequency is implemented in the second aperture. The basic techniques that are employed are the following:
    • a) Change in time (ΔT)
    • b) Change in delay (ΔD)
    • c) Change in frequency (ΔF)
    • d) Change in range (ΔR)
    • e) Interleaved pulses (ΔI)
    • f) Groups of data (ΔG)
    • g) Change in channel (ΔC)
    • h) Change in aperture (ΔA)
    • i) Simultaneous beams(antennas) (ΔS)
    • j) And/or any combinations of the techniques above and correlated

IIC—USS Section

1—Processing multiple returns (greater than three) employ a number of techniques to reduce processing such as known clutter reduces number of unknown returns by one, association of known returns close to processed returns may reduce number of unknown returns by one or more and analytic solution which performs well with three or less unknown returns, and the candidate phi technique which will correlate with the other techniques.

As the number of expected returns increases, the solution is attainable but more difficult and complicated. To aid in this problem any prior knowledge such as clutter is present, we know one of the returns has “0” or near “0” velocity and/or an association technique where adjacent RDBs processed indicates the return under test velocity of some or more of its returns. An analytic technique has been developed but greatly facilitated by previous techniques.

2—ΔF processing is performed as a check on the azimuth, it takes another sample(s) in frequency close to initial sample and process this data the same as the initial sample and should have the same solutions for velocity of returns but the resulting returns amplitude and phase change gives the indication of where the peaks of the returns resides and therefore accurate azimuth determination. Also if addition data is processed it should remain the same

3—ΔR processing is performed to determine the range more accurately, it takes another sample in range close to initial sample and process this data the same as initial sample and should have the same solutions for velocity of returns but the resulting returns amplitude and phase change gives the indication of where the peaks of the returns resides and therefore the accurate range determination Also if addition data is processed it should remain the same

4—ΔH processing. If implementing a planar array, processing as performed on the first linear array is done on the other linear arrays, the same or close to same solutions for Φ_O and Φ₁ (radial velocity of first and second return) and Φ_AO and Φ_A1 (azimuth of first and second returns). Also if addition data is processed it should remain the same.

Taking next the solutions for M0 and M1 for processing each linear array solves for the change in phase of M0 and M1, since the amplitude changes and is proportional to the height of the return.

If the returns are considered far field and the vertical spacing of the arrays are λ/2 and λis the wavelength and the angle the radar waves are making with the return is θ, the sine θ is proportional to the height of return divided by the slant range. The phase differences between the linear arrays is λ/2* sine θ*π/180 in radians will give the height of the return where sine θ=H/R and H=R*sine θ (H—height and R—range). The phase difference will give the height of the return (illustrated in FIGS. 3 and 4). The phase difference from linear array to linear array should be same or close to same.

In the multi pulse technique processing a significant time later the data is processed identically the difference in height(phase) will give the vertical tangential velocity together with the horizontal tangential velocity give the total tangential velocity vector and add that to unambiguous radial velocity gives the total three dimensional velocity vector.

5—Applying to all systems, if we obtain multiple looks at the return we will determine the unambiguous radial velocity and tangential velocity and greater accuracy in determining the return range, velocity and azimuth.

Multilooks is defined as a look with data point 1 to N and delay the data a portion of the N point such as N/4 and adding N/4 points at the end and performing the same operations. This will result in the increased capability as stated.

When the data is delayed and reprocessed as the first set of data the returns radial motion will be measured by the number of range bins or part of a bin traveled in this time (ΔR/ΔT) gives the true velocity of the return and will resolve the unambiguous velocity, if any, of the return without resorting to another PRF and saves time. It will also be a check on the radial velocity.

The horizontal tangential velocity will also be determined which could not be determined before as a measure of (ΔD/ΔT) doppler bins moved in the time difference and the phase difference between linear arrays will give the height difference, ΔH/ΔT, hence vertical tangential velocity. Hence, the total velocity of the return is determined and not only the radial velocity, the ratio of the radial velocity to the horizontal and vertical tangential velocity will give the angle the return is pointing in space.

Multi-looks of the return will result in better parameter estimation of the return and estimation of the range, velocity and azimuth.

6—IMPLEMENTING ANOTHER PRF OR PRFs in the multipulse system adds additional capability in determining unambiguous radial velocity, ability to separate out the number of returns that are detected in the same RDB and determine there velocities, reduced clutter area competing in same RDB with other returns in both PRFs. Also other returns that occur in same PRF in one PRF do not occur in the same manner in the other PRF.

a. The clutter is present in one PRF is different in the other PRF and consequently the clutter common to the two PRFs may be reduced by as much as 50% and help significantly in determining the nature of returns by reducing the number of returns detected in one range bin by at least one. The returns will occur in the same range bin or close to same range bin when one PRF is followed by the other PRF. This occurs when the ambiguous velocity is greater than that due to the lower PRF.

b. When the last condition exists as in previous paragraph the clutter and returns in the first PRF occur in different doppler bins and have the ability in one PRF a much easier ability to process many returns than occur in the other RDB.

c. When one solution for velocity is determined in one PRF then knowing the PRFs the other radial velocity is determinable in the other PRF consequentially making it easier to determine the solution for velocities when four or more returns are detected in the same RDB.

d. When a return is detected in both PRFs the ability to determine the unambiguous velocity is increased significantly therefore the ΔR evaluation of obtaining the unambiguous velocity is much more accurate and effective.

e. All of the above makes the ability to determine the identification of returns when the number per RDB is high significantly more easily attainable. Of course more than two any number of PRFs may be implemented.

7—Post Detection Processing

Since as a result of processing we have to sort out the type of returns detected. This may be movers, clutter, multipath returns, antenna sidelobes, filter side lobes, land sea interface and others. This presents a challenge to mathematically analyze the returns and/or gave a data base that assist in categorizing the returns.

Isodop correction for the velocity of the return, focusing the array may be performed to enhance the accuracy of the system.

Motion compensation relative to the boresight of the antenna is assumed.

Add ISAR processing for further classification of the ship when the parameters are processed.

8—DPCA principles are applied in a number of techniques. To illustrate the principles and the equations involved an example will be presented.

Antenna length—eleven feet

TRANSMISSION ARRAY—ELEVEN FEET

Number of receive arrays—two of 5.5 feet each

PRF-1000 HZ

Velocity of platform—500 feet/second

Number of pulses—64

For DPCA compensation due to array two not traveling the ideal distance of 2.75 feet in time equivalent of the distance is a phase in the frequency domain. Expressing this in equation form we have the following:

If there is no delay between array 1 and array 2 this is a distance equal to 2.75 feet. This distance in time equivalent equals 2.75 feet/500 feet/sec (velocity of the platform) which is 5.5 milliseconds (D/2V). Φ CO = 2 ⁢ Π ⁢   ⁢ FT where T = D / 2 ⁢ V and F = F R ⁢ K / N where F R ⁢   ⁢ is ⁢   ⁢ the ⁢   ⁢ PRF and N ⁢   ⁢ is ⁢   ⁢ the ⁢   ⁢ number ⁢   ⁢ of ⁢   ⁢ PRIs = 2 ⁢ Π ⁢   ⁢ F R ⁢ K / ND / 2 ⁢ V = Π ⁢   ⁢ F R ⁢ KD / NV where K ⁢   ⁢ is ⁢   ⁢ the ⁢   ⁢ filter ⁢   ⁢ number

• • is the phase compensation
    Φ_CO is the phase compensation for DPCA OPERATION
    Φ_CE is the phase compensation error for DPCA OPERATION
    Φ_CE=πF_RKD/2NV

The phase compensation error may now be calculated for each delay starting with no delay where time equal 5.5 milliseconds and substituting all parameters in the equation we get 9.8 degrees and for next time for succeeding pulse is 7.0 degrees and the next is 4.2 degrees and the next is 1.4 degrees. It is observed that the change in phase compensation (ΔK_D0) is constant.

Since this is true regardless of change in increment since it is constant (Δx). The first phase compensation depends on time and position in filter of return. The phase compensation term is known (K_D0) but not the position in the filter (Yo).

9—Other Factors

• • a. Ocean combined with overland. capability and employing the phase corrections and phase coefficient if necessary as explained in DSARS patent.
  • b. The mode of operation depends on many factors such as range, surveillance, tracking or spot light operation, sea state, etc.
  • c. Surveillance mode may be combined with spot image mode.
  • Wake and bow wave signatures of ships signatures as a function of their velocity and direction of the ship in help in classification of ships.
  • d. Surveillance mode could be a one antenna transmit and receive system.
  • e. High sea state conditions create shadow conditions that could be employed for better processing.
  • f. In the ocean the place where the ship actually exists there is no return have there azimuth at this location
  • g. Surveillance plus spot image mode may be combined to reduce dwell time and obtain maximum information per unit time.
  • h. This technique may be extended to space borne operations

Problem Areas

• • a) High clutter sea states are a big problem area and challenge to perform meaningful operations.
  • b) Long dwell time to perform accurate determinations. This is the reason for decreased spacing of doppler bins and increased sampling per range bin might ameliorate that condition.
  • c) Long range makes things very difficult.
  • d) Performing surveillance and tracking at the same time as classification of ships.
  • e). Operate without change in time operation for the surveillance mode.

10—Block Diagram of System

For the over ocean capability this is a simplified system as illustrated in FIGS. 1 to 5. FIG. 1 shows different positions of ship and effects due to these positions. FIG. 2 illustrates some techniques employed. FIGS. 3 and 4 shows the height and shadow determinations. FIG. 5 depicts the techniques employed as well as different affects detected and evaluated indicating the basic data is received from the radar in digital form and stored and processed in any or selected techniques described in the disclosure. The data is spectrally processed and the detection of the ship is performed together with the detection of the shadow and black hole to determine the ships radial velocity, azimuth and range, as well as the measurement of the ship parameters to classify the ship with as much accuracy as possible.

The delay data is processed to determine radial velocity unambiguously and to determine the tangential velocities and the ship parameters more accurately.

11—The following analysis may be applied to all VSS systems especially over the ocean detection of ships as follows:

1—ONE CHANNEL-MANY PULSES-ONE RETURN SIMPLE SOLUTION

The equations are as follows:
V₀=M₀ channel 1 time 1
V′₀=M₀X₀ channel 2 time 1
V₁=M₀e^jΦD0 channel 1 time 2
V′₁=M₀X₀e^jΦ′D0e^jθ0 channel 2 time 2
e^jΦD0=V₁/V₀
Φ_D0=Φ′_D0
e^jΦ′D0e^jθ0=V′₁/V′₀
e^jθ0=(V′₁/V′⁰)/(V₁/V₀):θ₀=Φ₀+ΔK_D0X₀

M′₀/M₀ Where the ratio of the amplitude at aperture 1 to aperture 2 as a function of azimuth is determined a prior by taking antenna measurements or real measure relatively high amplitude clutter only data. Like equations above will give curve of ratio of outputs-vs-azimuth.

Where the definition of terms are as follows:

V₀—is the return of first aperture output

V′₀—is the return of second aperture output

V₁—is the return of first aperture output delayed

V′₁—is the return of second aperture output delayed

Φ_D0—is the phase of the first aperture proportional to its velocity plus azimuth

Φ′_D0—is the phase of the second aperture proportional to its velocity plus azimuth

The phase of M′₀/M₀ is Φ_D0/2 proportional to its phase of its velocity plus azimuth. The phase Φ₀=Φ_D0−Φ_A0 will give the phase proportional to radial velocity.

Φ₀—phase of return proportional to radial velocity

Φ_D0—phase of return proportional to radial velocity plus azimuth

Φ_A0—phase of return proportional to azimuth

12—ONE PULSE-MANY channels-ONE RETURN SIMPLE SOLUTION

The equations are as follows:
V₀=M₀
V′₀=M′₀
V₁=M₀e^jΦD0
V′₁=M′₀e^jΦ′D0
e^jΦD0=V₁/V₀
e^jΦ′D0=V′₁/V′₀
Φ_D0=Φ′_D0

Where the definition of terms are as follows:

M′₀/M₀ Where the ratio of the amplitude at aperture 1 to aperture 2 as a function of azimuth is determined a prior by taking antenna measurements or real measure relatively high amplitude clutter only data. Like equations above will give curve of ratio of outputs-vs-azimuth.

Where the definition of terms are as follows:

V₀—is the return of first aperture output

V′₀—is the return of second aperture output

V₁—is the return of first aperture output delayed

V′₁—is the return of second aperture output delayed

Φ_D0—is the phase of the first aperture proportional to its velocity plus azimuth

Φ′_D0—is the phase of the second aperture proportional to its velocity plus azimuth

The phase of M′₀/M₀ is Φ_D0 proportional to its phase of its velocity plus azimuth. The phase Φ₀=Φ_D0−Φ_A0 will give the phase proportional to radial velocity.

Φ₀—phase of return proportional to radial velocity

Φ₀—phase of return proportional to radial velocity plus azimuth

Φ_A0—phase of return proportional to azimuth

V₀—is the return of first pulse output

V₁—is the return of second pulse output

Φ_A0—is the phase of the return proportional to its velocity

13a—TWO CHANNEL-MANY PULSES-ONE RETURN-delta D technique where M1 return much larger than other returns for example a ship return in the ocean or in a shadow area employing DPCA methodology.
V₀=M₀ channel 1 time 1-M delay 0  (1)
V₁=M₀X_O channel 2 time 1-M delay 0  (2)
V₂=M₀X_Oe^jθ0 channel 2 time 2-M+1 delay 1  (3)
X_O=1/W_M0=A_M0e^{j(ΨM0+KD0Y0)}=A₀  (4)
A₀e^jθ0=V₂/V₁=A₀e^{j(Φ0−ΔKD0X0)}  (5)
A_M0=|V₂/V₁|:(6′)e^jθ0=phase of V₂/V₁  (6)
θ₀=Φ₀−ΔK_D0X₀  (7)

Solve equation (7) for θ₀ is known and X₀ are unknown and therefore from the peak of where the return is detected which is Φ_D0 and therefore X₀ and Φ₀ has been determined and phase Φ_A0=Φ_D0−Φ₀ will give the phase proportional to azimuth.

M1—RETURN

With clutter only returns, it could be utilized for producing the curve ratio between apertures-vs-azimuth.

V0—first channel output at time 1

V1—second channel output at time 1

AM0—Channel balancing amplitude factor

Ψ_M0—Channel balancing phase factor

KD1—Phase coefficient to correct for detection of return not at center of doppler bin

X0—Distance return detected from center of doppler bin

WM0—Factor that makes channel two equal to channel 1

Φ₀ phase of return proportional to radial velocity

Φ_D0—phase of return proportional to radial velocity plus azimuth

Φ_A0—phase of return proportional to azimuth

13b TWO CHANNEL-MANY PULSES-ONE RETURN-Φ_D0 technique

The equations are as follows:
V₀=M₀ channel 1 time 1
V′₀=M₀X₀ channel 2 time 1
V₁=M₀e^jΦD0 channel 1 time 2
V′₁=M₀X₀e^jΦ′D0e^jθ0 channel 2 time 2
e^jΦD0=V₁/V₀
Φ_D0=Φ′_D0
e^jΦ′D0e^jθ0=V′₁/V′₀
e^jθ0=(V′₁/V′⁰)/(V₁/V₀):θ₀=Φ₀+ΔK_D0X₀

Solve equation (7) for θ₀ is known and X₀ are unknown and therefore from the peak of where the return is detected which is Φ_D0 and therefore X₀ and Φ₀ has been determined and phase Φ_A0=Φ_D0−Φ₀ will give the phase proportional to azimuth.

The same definition of terms as in 18a.

14a—TWO PULSE-MANY channels-ONE RETURN-delta C technique employing DPCA methodology.
V₀=M₀ pulse 1 channel 1-N delay 0  (1)
V₁=M₀X_O pulse 2 channel 1-N delay 0  (2)
V₂=M₀X_Oe^jθ0 pulse 2 time 2-N+1 delay 1  (3)
X_O=1/W_M0=A_M0e^{j(ΨM0+KD0X0)}=A₀  (4)
A₀e^jγ0=V₂/V₁=A₀e^{j(ΦA0−ΔKD0X0)}  (5)
A_M0=V₂/V₁:(6)e^jγ0=phase of V₂/V₁  (6)
γ₀=Φ_A0−ΔK_D0X₀  (7)
Solve equation (7) for γ₀ is known and X₀ are unknown and therefore from the peak of where the return is detected which is Φ_D0 and therefore Φ_A0 and Φ₀ has been determined and phase Φ₀=Φ_D0−Φ_A0 will give the phase proportional to radial velocity
M1—return
With clutter only returns, it could be utilized for producing the curve ratio between apertures-vs-azimuth.
V0—first pulse output at channel 1
V1—second pulse output at channel 1
AM0—Channel balancing amplitude factor
Ψ_M0—Channel balancing PHASE factor
KD0—Phase coefficient to correct for detection of return not at center of doppler bin
X0—Distance return detected from center of doppler bin
WM0—Factor that makes phase channel two equal to channel 1
Φ₀ —phase of return proportional to radial velocity
Φ_D0—phase of return proportional to radial velocity plus azimuth
Φ_A0—phase of return proportional to azimuth

14 b TWO CHANNEL-MANY PULSES-ONE RETURN-Φ_D0 technique

The equations are as follows:
V₀=M₀ pulse 1 channel 1N
V′₀=M₀X₀ PULSE 2 channel 1-N
V₁=M₀e^jΦD0 pulse 1 channel 2-N+1
V′₁=M₀X₀e^jΦ′D0e^jγ0 PULSE 2 channel 2-N+1
e^jΦD0=V₁/V₀
Φ_D0=Φ′_D0
e^jΦ′D0e^jγ0=V′₁/V′₀
e^jγ0=(V′₁/V′⁰)/(V₁/V₀):γ₀=Φ₀+ΔK_D0X₀

Solve for γ₀ is known and X₀ are unknown and therefore from the peak of where the return is detected which is Φ_D0 and therefore Φ_A0 and Φ₀ has been determined and phase Φ₀=Φ_D0−Φ_A0 will give the phase proportional to radial velocity

The same definition of terms as in 19a.

15—Implementing Another Transmission Frequency

Against broadband jamming change the transmission frequency to avoid jamming but not enough to change the implementation.

III—Basic Equations and Methodology

Detection of wanted returns and reject unwanted returns such as clutter. In previous STAP approaches where a number of pulses (M) of a particular PRF and a number channels (N-receive arrays) are implemented.

Considering that channels returns differs in both elevation and azimuth pattern(magnitude as well as phase), employing a minimum number of channels reduces the channel matching problem and reduces the storage and processing requirements significantly since the number of channels proposed to be a few as one instead of the usual ten or more. It proposes to perform as well or better than any of the known STAP systems.

Also considering as in most proposed STAP implementations some form of clutter cancellation are employed such as clutter covariance matrix or training data or knowledge aided clutter detection for non homogeneous data or high discretes of clutter for better detection of returns of interest. Returns of interest have to be thresholded above the clutter residue which may be significant and lead to many false alarms and/or missed detections. In the proposed USS no cancellation of clutter is required and all the above clutter reductions are eliminated. Consequentially when thresholding for possible meaningful returns they are not competing with clutter and simpler to detect and measure their radial velocity and azimuth very accurately.

This STAP methodology employs two or more channels, process the pulse data (slow data) first into its frequency spectrum and consequently localize clutter into its own range doppler bin (RDB) together with any other returns that may be detected in that same RDB such as target, thermal noise, and others.

The subsequent processing of each RDB will separate out returns doppler wise and since clutter has zero (0) radial velocity and other returns in the same (RDB) will have different velocities. From determining the velocities of the returns the azimuths will be calculated therefore no additional channels are required.

The knowledge aided STAP will be involved to determine from the detected returns in the RDB which are clutter, targets of interest, sidelobe returns, land sea interface, thermal noise, etc. The knowledge aided STAP will be not be involved in canceling clutter but in determining the returns of interest so they may be detected and there parameters measured and determine the nature of the return.

The following sections will be an analysis of various techniques with their mathematical development to accomplish these ends. It is assumed the data in time has been processed by FFT into their individual RDBs where there exist in the cases of interest clutter (0-velocity) and other returns (non “0” velocity). Initially two (2) returns will be developed; it may be clutter and moving target or two moving targets or more returns.

Two or three more returns in one RDB will be considered but more than three returns can be processed and determine the nature of the returns.

III-A Two channels at a time-“M”-PULSES in time data-two returns-Φ_D technique

The analysis may be performed with a one antenna transmit and two (channel) receive system. This system is called ΔT methodology where the data will be delayed one or more time increments in channel 1 and 2 as required for a solution. We will consider two returns clutter and target and the “M” pulse data (time data) has been spectrum processed into its individual RDBs and each will be treated as follows:

Each set of data, each time the data point is delayed it is multiplied by a suitable weighting function and its spectrum is obtained with such as FFT. In processing a particular RDB where we have two returns we have the following equations:
V₀₀=M₀+M₁ Channel 1 Pulse data 1-M Delay 0  (1)
V₀₁=M₀e^jΦD0+M₁e^jΦD1 Channel 1 Pulse data 2-M+1 Delay 1  (2)
V₀₂=M₀e^j2ΦD0+M₁e^j2ΦD1 Channel 1 Pulse data 3-M+2 Delay 2  (3)
V₀₃=M₀e^j3ΦD0+M₁e^j3ΦD1 Channel 1 Pulse data 4-M+3 Delay 3  (4)
Above equations are for two returns where
V₀₀ —is the return in the RDB channel 1 being processed at time 1
V₀₁ —is the return in the RDB channel 1 being processed at time 2
V₀₂—is the return in the RDB channel 1 being processed at time 3
V₀₃—is the return in the RDB channel 1 being processed at time 4
M₀—is the first return vector
M₁—is the second return vector
Φ_D0—is the phase of the first return where the phase is proportional to the phase due to radial velocity plus that due to the azimuth of the return.
Φ_D1—is the phase of the second return where the phase is proportional to the radial velocity plus that due to the azimuth of the return

It is noted with each delay in time of the data the vectors of the returns phase is increased proportional to the delay which represents the phase of the return proportional to velocity and that due to its azimuth position in the antenna beam. Zero velocity returns such as clutter will have phase shift equal to zero due to its velocity but one due to its azimuth position in the antenna beam and other returns will have phase shifts directly proportional to their radial velocity and one due to its azimuth position in the main beam. When returns are detected in the same RDB the sum of their phases (Φ_D0 or Φ_D1) (frequency) are detected in the same in RDB but are different in phase value and it is on this basis the returns are analyzed, processed and separated out.

Taking equations (1) and (2) and treating M₀ and M₁ as the variables and solving for M₀ and M₁ we have:
M₀=(V₀₀e^jΦD1−V₀₁)/(e^jΦD1−e^jΦD0)  (1′)
M₁=(V₀₀−V₀₁e^jΦD0)/(e^JΦD1−e^jΦD0)  (2′)
Taking equations (2) and (3) and treating M₀e^jΦD0 and M₁e^jΦD1 as the variables and solving for M₀e^jΦD0 and M₁e^jΦD1 we have:
M₀e^jΦD0=(V₀₁e^jΦD1−V₀₂)/(e^jΦD1−e^jΦD0)  (1″)
M₁e^jΦD1=(V₀₁−V₀₂e^jΦD0)/(e^jΦD1−e^jΦD0)  (2″)
Equation (2″)/Equation (2′) or Equation (1″)/Equation (1′) are the following:
e^jΦD0=(V₀₁e^jΦD1−V₀₂)/(V₀₀e^jΦD1−V₀₁)  (3′)
e^jΦD1=(V₀₂−V₀₁e^jΦD0)/(V₀₁−V₀₀e^JΦD0)  (4′)
Equation (3′) or Equation (4′) is easily solved for Φ_D0 and Φ_D1 which are proportional to the total phase of return 0 and return 1 respectively. If return “Mo” is clutter then Φ₀=0 corresponding to clutter having zero (0) velocity.

Now employing equations (1) and (2) and solving for M₀ and M₁ knowing Φ_D0 and Φ_D1 we are now are to find Φ₀ and Φ₁ which are proportional to velocity of return 0 and return 1 respectively. M₀ and M₁ are returns that are detected in the same RDB.

Having the second channel data and performing the same operations as in channel 1 and the equations are as follows:
V′₀₀=M₀X₀₁+M₁X₁₁ Channel 2 Pulse data 1-M Delay 0  (1′)
V′₀₁=M₀X₀₁e^jΦ′D0+M₁X₁₁e^jΦ′D1 Channel 2 Pulse data 2-M+1 Delay 1  (2′)
V′₀₂=M₀X₀₁e^j2Φ′D0+M₁X₁₁e^j2Φ′D1 Channel 2 Pulse data 3-M+2 Delay 2  (3′)
V′₀₃=M₀X₀₁e^j3Φ′D0+M₁X₁₁e^j3Φ′D1 Channel 2 Pulse data 4-M+3 Delay 3  (4′)
X₀₁=e^jDΦ0e^jKD0X0/W_M0=|1/A_M0|e^{j(KD0X0−ΨM0)} where D=0X₀₁=e^jKD0X0/W_M0=|1/A_M0|e^{j(KD0X0−ΨM0)}
X₁₁=e^jDΦ1e^jKD0X1/W_M1=|1/A_M1|e^{j(KD0X1−ΨM1)} where D=0X₁₁=e^jKD0X1/W_M1=|1/A_M1|e^{j(KD0X1−ΨM1)}

Solving equations (1′) to (4′) in the same manner as equations (1) to (4) we solve for Φ′_D0 and Φ′_D1 and M₀X₀₁ and M₁X₁₁. Φ′_D0 and Φ′_D1 solution should be the same as for Φ_D0 and Φ_D1 since in channel 2 we have the same returns detected in the same RDB with the same velocity components. solution for M₀X₀₁ and M₁X₁₁ in channel 2/solving for M₀ and M₁ in channel 1 yields X₀₁ and X₁₁ Having the second channel data delayed and performing the same operations as in channel 1 and 2 and the equations are as follows:
V″₀₀=M₀X₀₂+M₁X₁₂ CHANNEL 2 PULSE data 2-M+1 Delay 1  (1″)
V″₀₁=M₀X₀₂e^jΦ″D0+M₁X₁₂e^j2Φ″D1 CHANNEL 2 PULSE data 3-M+2 Delay 2  (2″)
V″₀₂=M₀X₀₂e^j2Φ″D0+M₁X₁₂e^j2Φ″D1 CHANNEL 2 PULSE data 4-M+3 Delay 3  (3″)
V″₀₃=M₀X₀₂e^j3Φ″D0+M₁X₁₂e^j3Φ″D1 CHANNEL 2 PULSE data 5-M+4 Delay 4  (4″)
X₀₂=X₀₁e^j(Φ0+ΔKD0ΔX⁰⁾×X₀₁e^j(θo): X₁₂=X₁₁e^j(Φ1+ΔKD0ΔX⁰)=X₁₁e^j(θ1)
V″₀₀=M₀X₀₁e^jθ0+M₁X₁₂e^jθ1 Channel 2 Pulse data 2-M+1 Delay 1  (1″)
V″₀₁=M₀X₀₁e^j2θ0e^jΦ″D0+M₁X₁₂e^j2θ1e^jΦ″D1 Channel 2 Pulse data 3-M+2 Delay 2  (2″)
V″₀₂=M₀X₀₁e^j3θ0e^j2Φ″D0+M₁X₁₂e^j3θ1e^j2Φ″D1 Channel 2 Pulse data 4-M+3 Delay 3  (3″)
V″₀₃=M₀X₀₁e^j4θ0e^j3Φ″D0+M₁X₁₂e^j4θ1e^j3Φ″D1 Channel 2 Pulse data 5-M+4 Delay 4  (4″)
Solving equations (1″) to (4″) the same manner as equations (1) to (4) we solve for (1D0 and Φ_D1 and M₀X₀₁ and M₁X₁₁. Φ″_D0 and Φ″_D1 solution should be the same.

solution for M₀X₀₁e_jθ⁰ and M₁X₁₁e^jθ1 in channel 2 delayed/solving for M₀X₀₁ and M₁X₁₁ in channel 2 yields e^jθ0 and e^jθ1 and

θ₀=Φ_0+ΔKD0X₀ where θ₀ is known and X₀ is unknown and therefore Φ₀ has to be determined

θ₁=Φ_1+ΔKD0X₀ where θ₁ is known and X₁ is unknown and therefore Φ₁ has to be determined

If the location is taken at the center of the filter the error in determination of azimuth is plus or minus a half a RDB. If a more accurate determination is desired a point of frequency close to first filter is created and processed like that of first filter.

This gives the same Φ₀ and Φ₁ and different M₀ and M₁ and the ratio of the M₀ and M₁ should give a good estimate of where the position Φ_D0 and Φ_D1 is detected at in the RDB. From this an estimate of azimuth of both returns are determined. To get a more accurate determination another frequency may be processed or a slight change in the range processed and results correlated for best results.

From a second set of data a small known change in frequency from the first set of data. We assume there will be no change in the channel balancing terms A_M0, Ψ_M0 and X₀ and Φ₀ which are the amplitude and phase term but an unknown DPCA term ΔK_D0 X₀ where X₀ is the unknown change in position of new filter and ΔK_D0 DPCA known constant therefore the term is unknown. Performing the operations on the second set of data, the solutions are the same for Φ₁ and Φ₀

The returns change due to frequency change from

M₀ to M′₀ where X_F0=M′₀/M₀ and M₁ to M′₁ where X_F1=M′₁/M₁. Therefore we have determined the ratio of the returns from which we estimate the position of the peak where is the first return and from that calculate the azimuth of the return. We can analogously perform that for the second return.

Reference: section I J on DPCA calculations will give the derivations of Φ_AO, Φ_D0, Φ₀ and Φ_A1,Φ_D1,Φ₁,and ΔK_D0, ΔK_D1, Δ_X0, Δ_X1, X₀, X₁, θ₀, θ₁

Solving for phase proportional to azimuth in both returns we have the following:
Φ_AO=Φ_D0−Φ₀ and Φ_A1=Φ_D1−Φ₁
Definition of terms not defined previously:
ALL “V” TERMS ARE MEASURED TERMS.—
X01—CHANNEL 2 TERM THAT makes relates channel 2 to channel 1 for return 1
X02—CHANNEL 2 TERM THAT makes relates channel 2 to channel 1 for return 2
ΔK_D0—the difference factor for different delays for return 1 and 2
X₀—the position in filter for return 1
θ₀—the difference in angle between different delayed data of return 1
θ₁—the difference in angle between different delayed data of return 2
Φ_AO—phase proportional to azimuth of return 1
Φ_A1—phase proportional to azimuth of return 2
A_M0—Amplitude balancing term between channel 1 and 2 for return 1
A_M1—Amplitude balancing term between channel 1 and 2 for return 2
Ψ_M0—Phase balancing term between channel 1 and 2 for return 1
Ψ_M1—Phase balancing term between channel 1 and 2 for return 2

Comments and observations on technique:

1—All solutions Φ_D0, Φ′_D0, Φ″_D0 should be equal and Φ_D1, Φ′_D1, Φ″_D1 should be equal

2—Solving for M0 and M1 by this approach solves for the location of their peaks therefore they have a phase shift equal to zero at this point.

3—To solve for the channel balancing terms three sets of equations are required but for solving for velocity and azimuth only last two sets are required.

4—Correlate with other ΔT-DPCA-technique in the following manner:

a) same solution

b) all variables are the same value such as M0, M₁, ETC

c) ΔF, ΔR and ΔH results should have the same values and correlate

Analogously a small change in range bin may be taken and we determine X_R0 and X_R1 which determines where the peak of the returns in range, this does not help in the evaluation in azimuth. If and evaluation in resolving velocity ambiguity with the taking of a meaningful delay in time and processing again. Thus we can determine the peak of each return in range and azimuth to obtain the maximum amplitude for each return for further use. The change in amplitude and phase of the range bin in conjunction with a delay in time gives an accurate determination of velocity which will resolve velocity ambiguity.

The change in amplitude and phase of the doppler bin in conjunction with a delay in time gives an accurate determination of horizontal tangential velocity.

The change in amplitude and phase in the different linear arrays of the doppler bin in conjunction with a delay in time gives an accurate determination of vertical tangential velocity.

Thus we have obtained the three dimensional positions and velocities of all returns.

B. Two channel “M” pulse data in time-three returns-Φ_D technique

The previous analysis was for two returns possible per RDB processed; this will be for three (3) returns per RDB.
V₀₀=M₀+M1+M2 Channel 1 Pulse data 1-M Delay 0  (1)
V₀₁=M₀e^jΦD0+M₁e^jΦD1+M₁e^jΦD2 Channel 1 Pulse data 2-M+1 Delay 1  (2)
V₀₂=M₀e^j2ΦD0+M₁e^j2ΦD1+M₂e^j2ΦD2 Channel 1 Pulse data 3-M+2 Delay 2  (3)
V₀₃=M₀e^j3ΦD0+M₁e^j3ΦD1+M₂e^j3Φ2 Channel 1 Pulse data 4-M+3 Delay 3  (4)
All terms previously defined except the following:
M2—Third return
Φ_D2—phase proportional to radial velocity plus azimuth of third return
Φ_A2—phase proportional to azimuth of third return
Φ₂—phase proportional to radial velocity of third return
Taking equations (1) and (2) and (3) and treating M₀ and M₁ and M₂ as the variables and solving the determinant equation for A₀ we have: Δ 0 = ⅇ jΦ D ⁢   ⁢ 0 1 ⁢ ⅇ jΦ D ⁢   ⁢ 1 1 ⁢ ⅇ jΦ D ⁢   ⁢ 2 1 = 1 ⁢ ⁢ ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⅇ jΦ D ⁢   ⁢ 2 - 1 ⁢ ⁢ ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ jΦ D ⁢   ⁢ 2 + 1 ⁢ ⁢ ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⁢ Δ 0 = ⁢ ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⅇ jΦ D ⁢   ⁢ 2 - ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 + ⁢ ⅇ jΦ D ⁢   ⁢ 2 ⁢ ⅇ j2Φ D ⁢   ⁢ 0 + ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 - ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 0 = ⁢ function ⁢   ⁢ of ⁢   ⁢ Φ D ⁢   ⁢ 0 , Φ D ⁢   ⁢ 1 ⁢   ⁢ and ⁢   ⁢ Φ D ⁢   ⁢ 2 ⁢ ⁢ and ⁢   ⁢ solving ⁢   ⁢ for ⁢ ⁢ M 0 _ = M 0 * Δ 0   M 0 _ = V 01 V 00 ⁢ ⅇ jΦ D ⁢   ⁢ 1 1 ⁢ ⅇ jΦ D ⁢   ⁢ 2 1 = V 00 ⁢ ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⅇ jΦ D ⁢   ⁢ 2 - V 01 ⁢ 1 ⁢ ⁢ 1 + V 02 ⁢ 1 ⁢ ⁢ 1 ⁢ ⁢ V 02 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ V ⁢   ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⁢ ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⅇ jΦ D ⁢   ⁢ 2 ⁢ ⁢ M 0 _ = ⁢ V 00 ⁡ ( ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ jΦ D ⁢   ⁢ 2 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ) - ⁢ V 01 ( ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ j2Φ D ⁢   ⁢ 1 ) + V 02 ⁡ ( ⅇ jΦ D ⁢   ⁢ 2 - ⅇ jΦ D ⁢   ⁢ 1 ) = ⁢ function ⁢   ⁢ of ⁢   ⁢   ⁢ Φ D ⁢   ⁢ 1 ⁢   ⁢ and ⁢   ⁢ Φ D ⁢   ⁢ 2 ⁢ ⁢ and ⁢   ⁢ solving ⁢   ⁢ for ⁢ ⁢ M 1 _ = M 1 * Δ 0 ( 6 ) M 1 _ = ⅇ jΦ D ⁢   ⁢ 0 1 ⁢ V 01 V 00 ⁢ ⅇ jΦ D ⁢   ⁢ 2 1 = V 00 ⁢ ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ jΦ D ⁢   ⁢ 2 - V 01 ⁢ 1 ⁢ ⁢ 1 + V 02 ⁢ 1 ⁢ ⁢ 1 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ V 02 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ⁢ ⁢ ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ jΦ D ⁢   ⁢ 2 ⁢ ⁢ M 1 _ = ⁢ V 01 ⁡ ( ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ jΦ D ⁢   ⁢ 2 ) - ⁢ V 01 ( ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ j2Φ D ⁢   ⁢ 0 ) + V 02 ⁡ ( ⅇ jΦ D ⁢   ⁢ 2 - ⅇ jΦ D ⁢   ⁢ 0 ) = ⁢ function ⁢   ⁢ of ⁢   ⁢ Φ D ⁢   ⁢ 0 ⁢   ⁢ and ⁢   ⁢   ⁢ Φ D ⁢   ⁢ 2 ⁢ ⁢ and ⁢   ⁢ solving ⁢   ⁢ for ⁢ ⁢ M 2 _ = M 2 * Δ 0 ( 7 ) M 2 _ = ⅇ jΦ D ⁢   ⁢ 0 1 ⁢ ⅇ jΦ D ⁢   ⁢ 1 1 ⁢ V 01 V 00 = V 00 ⁢ ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ jΦ D ⁢   ⁢ 1 - V 01 ⁢ 1 ⁢ ⁢ 1 + V 02 ⁢ 1 ⁢ ⁢ 1 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ V 02 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⁢ ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⁢ M 2 _ = ⁢ V 00 ⁡ ( ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 - ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⅇ jΦ D ⁢   ⁢ 0 ) - ⁢ V 01 ( ⅇ j2Φ D ⁢   ⁢ 1 - ⅇ j2Φ D ⁢   ⁢ 0 ) + V 02 ⁡ ( ⅇ jΦ D ⁢   ⁢ 1 - ⅇ jΦ D ⁢   ⁢ 0 ) = ⁢ function ⁢   ⁢ of ⁢   ⁢ Φ D ⁢   ⁢ 0 ⁢   ⁢ and ⁢   ⁢   ⁢ Φ D ⁢   ⁢ 1 ( 8 )
Solving for Φ_D0, Φ_D1 and Φ_D2 and substituting these values in equations (1), (2) and (3) we determine M₀, M₁ and M₂

Taking equations (2), (3) and (4) and treating M₀e^jDΦ0, M₁e^jDΦ1 and M₂e^jDΦ2 as the variables and solving the determinant equation for Δ₀ is the same and performing the same operations as with the first set of equations we have the following:
M₀e^jΦD0=function of (Φ_D1,Φ_D2)  ( 6)
M₁e^jΦD1=function of (Φ_D0, Φ_D2)  ( 7)
M₂e^jΦD2=function of (Φ_D0,Φ_D1)  ( 8)
Equation ( 6)/(6), ( 7)/(7) and ( 8)/(8)
( 6)/(6)=e^jΦD0=function (Φ_D1, Φ_D2)
( 7)/(7)=e^jΦD1=function Φ_D0, Φ_D2)
( 8)/(8)=e^jΦD2=function (Φ_D0,Φ_D1)
Solving for Φ_D0, Φ_D1 and Φ_D2 and substituting these values in equations (2), (3) and (4) we determine M₀e^jDΦ0, M₁e^jDΦ1 and M₂e^jDΦ2.
V′₀₀=M₀X₀₁+M₁X₁₁+M₂X₂₁ Channel 2 Pulse data 1-M Delay 0  (1′)
V′₀₁=M₀X₀₁e^jΦ′D0+M₁X₁₁e^jΦ′D1+M₂X₂₁e^j2′D2 Channel 2 Pulse data 2-M+1 Delay 1  (2′)
V′₀₂=M₀X₀₁e^j2Φ′D0+M₁X₁₁e^j2Φ′D1+M₂X₂₁e^j2Φ′D2 Channel 2 Pulse data 3-M+2 Delay 2  (3′)
V′₀₃=M₀X₀₁e^j3Φ′D0+M₁X₁₁e^j3Φ′D1+M₂X₂₁e^j3Φ′D2 Channel 2 Pulse data 4-M+3 Delay 3  (4′)
X₀₁=e^jDΦ0e^jKD0X0/W_M0=|1/A_M0|e^{j(KD0X0−ΨM0)} where D=0X₀₁=e^jKD0X0/W_M0|1/A_M0|e^{j(KD0X0−ΨM0)}
X₁₁=e^jDΦ1e^jKD0X1/W_M1=|1/A_M1|e^{j(KD0X1−ΨM1)} where D=0X₀₁=e^jKD0X1/W_M1|1/A_M1|e^{j(KD0X1−ΨM1)}
X₂₁=e^jDΦ2e^jKD0X2/W_M2=|1/A_M2|e^{j(KD0X2−ΨM2)} where D=0X₀₁=e^jKD0X2/W_M2|1/A_M2|e^{j(KD0X2−ΨM2)}

Performing the same operations on equations (1′), (2′) and (3′) with the variables M₀X₀₁, M₁X₁₁ and M₂X₂₁ and Δ₀ remains the same. The analysis is analogous and the result is the following: and ⁢   ⁢ solving ⁢   ⁢ for ⁢   ⁢ M 0 ⁢ X 01 _ = M 0 ⁢ X 01 * Δ 0   M 0 ⁢ X 01 _ = ⁢ V 01 ⁡ ( ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 1 ′ ⁢ ⅇ j ⁢   ⁢ Φ ⁢   D ⁢   ⁢ 2 ′ - ⅇ j ⁢   ⁢ 2 ⁢   ⁢ Φ D ⁢   ⁢ 1 ′ ⁢ ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 2 ′ ) - ⁢ V 02 ⁡ ( ⅇ j ⁢   ⁢ 2 ′ ⁢ Φ D ⁢   ⁢ 1 - ⅇ j ⁢   ⁢ 2 ⁢ Φ D ⁢   ⁢ 2 ′ ) + V 03 ⁡ ( ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 1 ′ - ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 2 ′ ) = ⁢ function ⁢   ⁢ of ⁢   ⁢ Φ D ⁢   ⁢ 1 ′ ⁢   ⁢ and ⁢   ⁢ Φ D ⁢   ⁢ 2 ′ ⁢ ⁢ and ⁢   ⁢ solving ⁢   ⁢ for ⁢   ⁢ M 0 ⁢ X 11 _ = M 1 ⁢ X 11 * Δ 0 ( 6 ′ ) = function ⁢   ⁢ of ⁢   ⁢ Φ D ⁢   ⁢ 0 ′ ⁢   ⁢ and ⁢   ⁢ Φ D ⁢   ⁢ 2 ′ ⁢ ⁢ and ⁢   ⁢ solving ⁢   ⁢ for ⁢   ⁢ M 2 ⁢ X 21 _ = M 2 ⁢ X 21 * Δ 0 ( 7 ′ ) = function ⁢   ⁢ of ⁢   ⁢ Φ D ⁢   ⁢ 0 ′ ⁢   ⁢ and ⁢   ⁢ Φ D ⁢   ⁢ 1 ′ ( 8 ′ )
Taking equation (6′)/(6) we have the following:
X₀₁=M₀X₀₁/M₀

Taking equation (7′)/(7) we have the following:
X₁₁=M₁X₁₁/M₁  

Taking equation (8′)/(8) we have the following:
X₂₁=M₂X₂₁/M₂ V₀₂(e^jΦ″D0e^j2Φ″D1−e^j2Φ″D0e^jΦ″D1)−V₀₃(e^j2′Φ″D1−e^j2Φ′D0)+V₀₄(e^jΦ″D1−e^jΦ″D0)/V₀₁(e^jΦ′D0e^j2Φ′D1−e^j2Φ′D0e^jΦ′D1)−V₀₂(e^j2′ΦD1−e^j2Φ′D0)+V₀₃(e^{jΦ′D1−ejΦ′D0})
Performing the same operations on equations (2′), (3′) and (4′) with the variables M₀X₀₁, M₁X₁₁ and M₂X₂₁ and Δ₀ remains the same. The analysis is analogous and the result is the following:
M₀e^jΦD0=function of (Φ_D1, Φ_D2)  (6′)
M₁e^jΦD1=function of (Φ_D0, Φ_D2)  (7)
M2e^jΦD2=function of Φ_D0, Φ_D1)  (8′)
Equation ( 6′)/(6′), ( 7′)/(7′) and ( 8′)/(8′)
( 6′)/(6′)=e^jΦD0=function (Φ_D1,Φ_D2)
( 7′)/(7′)=e^jΦD1=function (Φ_D0, Φ_D2)
( 8)/(8′)=e^jΦD2=function (Φ_D0, Φ_D1)
Solving for Φ_D0, Φ_D1 and Φ_D2 and substituting these values in equations (2), (3) and (4) we determine M₀X₀₁, M1 X₀₂ and M2X₀₃

Taking the next set of equations as follows:
V″₀₀=M₀X₀₂+M₁X₁₂+M₁X₂₂ Channel 2 Pulse data 2-M+1 Delay 1  (1″)
V″₀₁=M₀X₀₂e^jΦ′D0+M₁X₁₂e^jΦ″D1+M₂X₂₂e^jΦ″D2 Channel 2 Pulse data 3-M+2 Delay 2  (2″)
V″₀₂=M₀X₀₂e^j2Φ″D0+M₁X₁₂e^j2Φ″D1+M₂X₂₂e^j2Φ″D2 Channel 2 Pulse data 4-M+3 Delay 3  (3″)
V″₀₃=M₀X₀₂e^j3Φ″D0+M₁X₁₂e^j3ΦD1+M₂X₂₁e^j3Φ″2 Channel 2 Pulse data 5-M+4 Delay 4  (4″)
X₀₂=X₀₁e^{j(Φ0+ΔKD0X0)}=X₀₁e^j(θ0): X₁₂=X₁₁e^{j(Φ1+ΔKD0X0)}=X₁₁e^j(θ1):X₂₁=X₂₁e^j(Φ2+K^D0X0)=X₂₁e^j(θ2)
Rewriting equations (1″ to 4″) we have the following:
V″₀₀=M₀X₀₁e^jθ0+M₁X₁₁e^jθ1+M₂X₁₂e^jθ2 Channel 2 Pulse data 2-M+1 Delay 1  (1″)
V″₀₁=M₀X₀₁e^jθ0e^jΦ″D0+M₁X₁₁e^jθ1e^jθ′D1+M₂X₁₂e^jθ2e^jΦ″D2 Channel 2 Pulse data 3-M+2 Delay 2  (2″)
V″₀₂=M₀X₀₁e^jθ0e^j2Φ″D0+M₁X₁₁e^jθ1e^j2Φ″D1+M₂X₁₂e^jθ2e^j2ΦD2 Channel 2 Pulse data 4-M+3 Delay 3  (3″)
V″₀₃=M₀X₀₁e^jθ0e^j3Φ′D0+M₁X₁₁e^jθ1e^j3Φ′D1+M₂X₁₁e^1θ2e^j3Φ′D2 Channel 2 Pulse data 5-M+4 Delay 4  (4″)
Performing the same operations on equations (1″), (2″) and (3″) with the variables M₀X₀₁e^jθ0, M₁X₁₁e^jθ1 and M₂X₂₁e^jθ2 and Δ₀ remains the same. The analysis is analogous and the result is the following: and ⁢   ⁢ solving ⁢   ⁢ for ⁢   ⁢ M 0 ⁢ X 01 ⁢ ⅇ j ⁢   ⁢ θ 0 _ = M 0 ⁢ X 01 ⁢ ⅇ j ⁢   ⁢ θ 0 * Δ 0   M 0 ⁢ X 01 ⁢ ⅇ j ⁢   ⁢ θ 0 _ = ⁢ V 02 ″ ⁡ ( ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 1 ′ ⁢ ⅇ j ⁢   ⁢ Φ ⁢   D ⁢   ⁢ 2 ′ - ⅇ j ⁢   ⁢ 2 ⁢   ⁢ Φ D ⁢   ⁢ 1 ′ ⁢ ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 2 ′ ) - ⁢ V 03 ″ ⁡ ( ⅇ j ⁢   ⁢ 2 ′ ⁢ Φ D ⁢   ⁢ 1 - ⅇ j ⁢   ⁢ 2 ⁢   ⁢ Φ D ⁢   ⁢ 2 ′ ) + V 04 ″ ⁡ ( ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 1 ′ - ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 2 ′ ) = ⁢ function ⁢   ⁢ of ⁢   ⁢ Φ D ⁢   ⁢ 1 ″ ⁢   ⁢ and ⁢   ⁢ Φ D ⁢   ⁢ 2 ″ ⁢ ⁢ and ⁢   ⁢ solving ⁢   ⁢ for ⁢   ⁢ M 1 ⁢ X 11 ⁢ ⅇ j ⁢   ⁢ θ 1 _ = M 1 ⁢ X 11 ⁢ ⅇ j ⁢   ⁢ θ 1 * Δ 0 ( 6 ″ ) M 1 ⁢ X 11 ⁢ ⅇ j ⁢   ⁢ θ 1 _ = ⁢ V 02 ⁡ ( ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 0 ″ ⁢ ⅇ j ⁢   ⁢ 2 ⁢   ⁢ Φ ⁢   D ⁢   ⁢ 2 ″ - ⅇ j ⁢   ⁢ 2 ⁢   ⁢ Φ D0 ″ ⁢ ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 2 ″ ) - ⁢ V 03 ⁡ ( ⅇ j ⁢   ⁢ 2 ″ ⁢ Φ D ⁢   ⁢ 0 - ⅇ j ⁢   ⁢ 2 ⁢   ⁢ Φ D ⁢   ⁢ 2 ″ ) + V 04 ⁡ ( ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 0 ″ - ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 2 ″ ) = ⁢ function ⁢   ⁢ of ⁢   ⁢ Φ D ⁢   ⁢ 0 ′ ⁢   ⁢ and ⁢   ⁢ Φ D ⁢   ⁢ 2 ′ ⁢ ⁢ and ⁢   ⁢ solving ⁢   ⁢ for ⁢   ⁢ M 2 ⁢ X 21 ⁢ ⅇ j ⁢   ⁢ θ 2 _ = M 2 ⁢ X 21 ⁢ ⅇ j ⁢   ⁢ θ 2 * Δ 0 ( 7 ″ ) M 2 ⁢ X 21 ⁢ ⅇ j ⁢   ⁢ θ 2 _ = ⁢ V 02 ⁡ ( ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 0 ″ ⁢ ⅇ j ⁢   ⁢ 2 ⁢   ⁢ Φ ⁢   D ⁢   ⁢ 1 ″ - ⅇ j ⁢   ⁢ 2 ⁢   ⁢ Φ D0 ″ ⁢ ⅇ j ⁢   ⁢ Φ D1 ″ ) - ⁢ V 03 ⁡ ( ⅇ j ⁢   ⁢ 2 ′ ⁢ Φ D ⁢   ⁢ 1 ″ - ⅇ j ⁢   ⁢ 2 ⁢   ⁢ Φ D ⁢   ⁢ 0 ″ ) + V 04 ⁡ ( ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 1 ″ - ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 0 ″ ) = ⁢ function ⁢   ⁢ of ⁢   ⁢ Φ D ⁢   ⁢ 0 ″ ⁢   ⁢ and ⁢   ⁢ Φ D ⁢   ⁢ 1 ″ ( 8 ′ )
Taking equation (6″)/(6′) we have the following:
e^jθ0=M₀X₀₁e^jθ0/M₀X₀₁=V″₀₂(e^jΦ′D1e^j2Φ′D2−e^j2Φ′D1e^jΦ′D2)−V″₀₃(e^j2′ΦD1−e^j2Φ′D2+V″₀₄(e^jΦ′D1−e^1ΦD′2)/V₀₁(e^jΦ′D1e^j2Φ′D2−e^j2Φ′D1e^jΦ′D2)−V₀₂(e^j2′ΦD1−e^j2Φ′D2)+V₀₃(e^jΦ′D1−e^jΦ′D2)  (12)
Taking equation (7″)/(7′) we have the following:
e^jθ1=M₁X₁₁e^jθ1/M₁X₁₁=V₀₂(e^jΦ″D0e^j2Φ″D2−e^j2Φ″D1e^jΦ″D2)−V₀₃(e^j2″ΦD1−e^j2Φ″D2+V₀₄(e^jΦ″D1−e^jΦD″2)/V₀₁(e^jΦ′D0e^j2Φ′D2−e^j2Φ′D0e^jΦ′D2)−V₀₂(e^j2′ΦD0−e^j2Φ′D2)+V₀₃(e^jΦ′D0−e^jΦ′D2)  (13)
Taking equation (8″)/(8′) we have the following:
e^jθ2=M₂X₂₁e^jθ2/M₂X₂₁=V₀₂(e^jΦ″D0e^j2Φ″D1−e^j2Φ″D0e^jΦ″D1)−V₀₃(e^j2′Φ″D1−e^j2Φ″D0)+V₀₄(e^j2′Φ″D1−e^jΦ″D0)/V₀₁ (e^jΦD0e^j2Φ_D1−e^jΦD1e^j2ΦD0)−V₀₂(e^j2ΦD1−e^j2ΦD0)+V₀₃(e^jΦD1−e^jΦD0)  (14)
Performing the same operations on equations (2″), (3″) and (4″) with the variables M₀X₀₁e^j0, M₁X₁₁e^jθ1 and M₂X₂₁e^jθ2 and Δ₀ remains the same Substituting equation (13) into equation (12) we have the following:
The analysis is analogous and the result is the following:
M₀e^jΦD0=function of (Φ_D1,Φ_D2)  ( 6)
M₁e^jΦD1=function of (Φ_D0, Φ_D2)  ( 7)
M₂e^jΦD2=function of (Φ_D0,Φ_D1)  ( 8)
Equation ( 6)/(6), ( 7)/(7) and ( 8)/(8)
( 6)/(6)=e^jΦD0=function (Φ_D1, Φ_D2)
( 7)/(7)=e^jΦD1=function Φ_D0, Φ_D2)
( 8)/(8)=e^jΦD2=function (Φ_D0,Φ_D1)
Solving for Φ_D0, Φ_D1 and Φ_D2 and substituting these values in equations (2), (3) and (4) we determine M₀X₀₀e^jθ0 M₁X₀₁e^jθ1 and M₂X₀₂e^jθ2
e^jθ0=M₀X₀₀e^jθ0/M₀X₀₀=function of Φ_D1 and Φ_D2
e^jθ1=M₁X₀₁e^jθ1/M₁X₀₁=function of Φ_D0 and Φ_D2
e^jθ2=M₂X₀₂e^jθ2/M₂X₀₂=function of Φ_D0 and Φ_D1

From the determination of M₀X₀₁,M₁X₁₁ and M₂X₂₁ and M₀X₀₀ e^jθ0, M₁X₀₁e^jθ1 and M₂X₀₂e^jθ2 we have determined θ₁, θ₂ and θ₃ And θ₀=Φ_0+ΔKD0X₀: θ₁=Φ_1+ΔKD0X₁:θ₂=Φ_2+ΔKD0X₂ where everything is known except Φ₀, Φ₁ and Φ₂ and X_O, X₁ and X₂ are determined as in two return case and having determined Φ_D0 and Φ_D1 and Φ_D2 since Φ_A0=Φ_D0−Φ₀: Φ_A1=Φ_D1−Φ₁:Φ_A2=Φ_D2−Φ₂ the azimuth of the returns have been attained.

Φ_D0 and Φ_D1 and Φ^D2 in the three sets of matrix data of three returns as follows Substituting in equation Φ_D1 as a function of Φ_D0 and Φ_D2 we have the following: Φ_D0 as a function of Φ_D2

Substituting in equation Φ_D1 as a function of Φ_D0 and Φ_D2 we have the following: Φ_D0 as a function of Φ_D2

Substituting equation Φ_D2 as a function of Φ_D0 and Φ_D1 we have the following: Φ_D1 as a function of Φ_D0

Similarly substituting Φ_D1 as a function of Φ_D2 we have the following: Φ₁ as a function of Φ_D0

Similarly substituting Φ_D1 as a function of Φ_D0 we have the following:

Similarly substituting Φ_D1 as a function of Φ_D0 we have the following:

Φ₁ as a function of Φ_D2

we have the following:

Φ_D1 as a function of Φ_D2

we have two equations two unknowns and solvable in Φ_D0 and Φ_D2 and similarly for Φ_D1 and Φ_D2.and Φ_D0 and Φ_D1 and Φ_D1, Φ_D2 and Φ_D3

Thus we have solved for θ₁, θ₂ and θ₃ twice.

B1 Assist in Processing Three or More Returns Per Detected Bin

We can perform these operations for the three sets of data for the three returns and the solutions should be the same or close to the same.

The methodology in simplifying and more robust solutions when there are three or returns is the following:

• • 1. Determine from processing adjacent bins for the detection of other returns that would also be detected in the processed bin such as clutter then we have one or more known solutions.
  • 2. Employing the candidate solution technique, that is substituting all possible solutions which are very limited in number (detected only in the beamwidth of the antenna).
  • 3. If radar returns are received from more than one range Doppler bin then we know that there is an object associated with the radar returns located in the overlap from the adjacent range Doppler bins (i.e., narrowing the possible location of the object). Thus, multiple objects in a single bin are easier to locate when overlap from other range Doppler bins is also considered.
  • 4. The solution entails Φ_D1, Φ_D2, Φ_D3, etc where the amplitudes are equal.
    From the second set of data a small frequency change from the first set of data this gives the same Φ_D1, Φ_D2 and Φ_D3 for the solution but different M₀, M₁ and M₂. The ratios of the M′₀/M₀,M′₁/M₁ and M′₂/M₂ should give a good estimate of where the position Φ_D0 and Φ_D1 and Φ_D2 are detected at their peak in that RDB and a check on the solutions determined. From this the azimuth of the returns are determined. From the ratio of the second set of data to the first set of data we obtain XF0 and XF1 and XF2 from which is the ratio of M′₀/M₀=X_F0 and M′₁/M₁=X_F1 and M′₂/M₂=X_F2 where all other terms are known. From the previous determinations of the estimate Φ_D0 and Φ_D1 and Φ_D2 which is the position in the filter where the returns are detected at there peak in the RDB. This gives a good estimate of the azimuth of the returns. To get more accurate determinations other close frequency points to initially processed data are processed.

From XF0 an estimate of where the returns are detected at there peak in the RDB. From the following equation Φ₀A=Φ_D0−Φ₀ where Φ₀A is the phase of the return proportional to the azimuth of the return, and Φ_D1, Φ_D2 and Φ_D3 is the phase of the return proportional to the peak of the return where, Φ₀ is the phase of the return proportional to the velocity of the return.

Similarly this is performed for XF1 and XF2 hence finding the azimuth of the second and third return.

Analogously a small change in range bin may be taken and we determine XR0, XR1 and XR2 which determines where the peak of the returns in range, this does not help in the evaluation in azimuth. If and evaluation in resolving velocity ambiguity with the taking of meaning delay in time and processing again. Thus we can determine the peak of each return in range and azimuth to obtain the maximum amplitude for each return for further use.

A more accurate determination of Φ_D0, Φ_D1 and Φ_D2 is determined by taking the four sets of equations and employing the candidate Φ technique substituting all possible solutions which are restricted to the values that can be in only one range doppler bin (RDB) but for when greater accuracy is required the number of candidate solutions increase. The candidate solutions should be very close in value for all four sets of data giving very robust and accurate solutions which determines all the parameters of the returns.

Also another correlating and checking operation is to repeat the processing with a close frequency and the results should be very close plus obtaining and obtaining change in return vector (XF). This would be the same for all sets of four sets of data and also would be a check on the Φ_D0, Φ_D1 and Φ_D2 solutions. This is illustrated in FIG. 12.

Analogously this would process a close sample in the range direction and also for processing other linear arrays. Attaining the precise range, azimuth, height, unambiguous radial velocity and vertical and tangential velocity is performed the same as in other single channel many pulse systems. RECAPPING—We have determined all Φs, in the two return and three return case and as consequently it may be performed for more than three Φs. The significance of this development is as each RDB that is processed clutter, target, noise, and other returns may be detected and thresholded for importance and later post processed to determine if clutter, movers, sidelobes, multipath targets, and others.

Correlation Factors

1—Time delay as many pulses at a time as the number of returns and process data determine all the Φs

2—Additional time delay processed again and all (Ds should agree

3—Other RDB processed that have the same return data related to each such as mover should agree

4—IF more than two channels other dual channels is processed and results should agree.

5—Other techniques that are related as to be shown later in document and results should agree

6—If a planar array is implemented all other linear arrays should obtain the same results and height and vertical tangential velocity obtained.

The aforementioned system has many advantages such as the following:

1—No clutter cancellation of any kind is required therefore as follows:

• • a) no clutter covariance matrix;
  • b) no training data; and
  • c) no special clutter knowledge required.
    2—No channel matching required
    3—Returns and clutter do not compete with each other in there detection and therefore clutter and returns are thresholded separately and returns are ideally are competing with white noise only and make for an excellent return ratio to noise.
    4—Very simple system-less storage-less processing-less hardware and less dwell time and very accurate.
    5—Full transmit and receive antenna employed with full antenna gains
  • a) smaller antenna sidelobes;
  • b) full antenna gain; and
  • c) narrow clutter band width
    6—May be applied to two arrays and two arrays or three arrays. Each dual array processed as two channel system and should have same results and correlated.
    7—Correlation factors as stated in previous paragraph.
    8—The significance of the ability to process many returns in same RDB and determining there amplitude and phases and radial velocity gives the ability to separate clutter and either returns such as bona fide targets, moving clutter, multi path returns, etc. Knowledge aided information would aid in categorizing these returns.
    IIII. Two channel “M” Pulses-Two return analyses-DPCA TYPE OPERATION
    The following analysis may be applied for a one transmit with the two or more receive antennae (channels) utilizing DPCA techniques to find the range, radial velocity and azimuth. This set of equations is for a two channel ΔT system.
    A—This development is for two(2) returns with DPCA operation.
    V₀₀=M₀+M₁ Channel 1 Pulse data 1-M Delay 0  (1)
    V₀₁=M₀X₀₁+M₁X₁₁ Channel 2 Pulse data 1-M Delay 0  (2)
    V₀₂=M₀X₀₂e^jθ0+M₁X₁₁e^jθ0 Channel 2 Pulse data 2-M+1 Delay 1  (3)
    V₀₃=M₀X₀₃e^j2θ0+M₁X₁₁e^j2θ1 Channel 2 Pulse data 3-M+2 Delay 2  (4)
    X₀₁=e^1DΦ0/W_M0 where D=0:X₀₁=e^j(KD0X0)/W_M0  (5)
    X₁₁=e^jDΦ1/W_M1 where D=0:X₁₁=e^j(KD0X1)/W_M1  (6)
    X₀₂=e^{jDΦ0+ΔKD0X0/W}_M0 where D=1:X₀₁e^{j(Φ0+ΔKD0X0)}  (7)
    X₁₂=e^jΦ0+ΔKD0X1/W_M1 where D=1:X₁₁e^1ΦKD0X1)  (8)
    X₀₃=e^{jD(Φ0+ΔKD0X0)}/W_M0 where D=2:X₀₁e^{j2(Φ0+KD0X0)}  (9)
    X₁₃=e^{jD(Φ1+KD0X1)}/W_M1 where D=2:X₁₁e^j2(Φ^0+KD0X1)  (10)

Rewriting equations (1) to (4) incorporating development above we have the following:
V₀₀=M₀+M₁ Channel 1 Pulse data 1-M Delay 0  (1)
V₀₁=M₀X₀₁+M₁X₁₁ Channel 2 Pulse data 1-M Delay 0  (2)
V₀₂=M₀X₀₂e^jθ0+M₁X₁₁e^jθ0 Channel 2 Pulse data 2-M+1 Delay 1  (3)
V₀₃=M₀X₀₃e^j2θ0+M₁X₁₁e^j2θ1 (Channel 2 Pulse data 3-M+2 Delay 2  (4)
where θ₀=ΔK_D0X₀ and θ₁=ΔK_D0X₁
All terms previously defined except X₀₁ and X₁₁ where X₀₁ and X₀₂ is the DPCA factor plus that which makes the channel 1 and channel 2 equal and where A_M0 is the amplitude matching antenna factor and Ψ_M0 is are the phase matching factor and K_D0X₀ is the factor of the detection is from the center of filter. AMO and Ψ_M0 represents the imperfect matching between channel 1 and 2. K_D0X₀ is the factor that corrects if the return is not detected in the center of the filter. K_D0 is the DPCA constant that corrects off for mismatch from center of filter. X₀ is the distance the return is detected from the center of filter. X₁ and analogously the same as for W_M1. A_M1 is the amplitude factor and Ψ_M1 and K_D0 X₁ are the phase factor. A_M1 and W_M1 represent the imperfect matching between channel and 2. K_D0X₁ is the factor that corrects if the return is not detected in the center of the filter. K_D0 is the DPCA constant that corrects off center of filter. X1 is the distance the return is detected from the center of filter.
Equation (1) is the first channel and equation (2) is the second channel delayed no time delay.
Equation (3) is second channel delayed one times.

Taking equations (2) and (3) and treating M₀X₀₁ and M₁X₁₁ as the variables and vying for M₀X₀₁ and M₁X₁₁ we have:

Taking equations (3) and (4) and treating M₀X₀e^jθ0 and M₁X₁e^jθ1 as the variables and solving for M₀X₀e^jθ0 and M₁X₁e^jθ1 we have:
M₀X₀e^jθ0=(V₀₂e^jθ1−V₀₃)/(e^jθ1−e^jθ0)  (2′″)
M₁X₁e^jθ1=(V₀₂=V₀₃e^jθ0)/(e^jθ1−e^jθ0)  (3′″)
Equation (2′″)/Equation (2″) and equation (3′″)/equation (3″) is the following:
e^jθ0=(V₀₂e^jθ1−V₀₃)/(V₀₁e^jθ1−V₀₂)  (4)
e^jθ1=(V₀₂−V₀₃e^jθ0)/(V₀₁−V₀₂e^jθ0)  (5)
Equation (4) or Equation (5) is easily solved for θ₀ and θ₁ which are solved for Φ₀ and Φ₁ proportional to the radial velocity of return 0 and return 1 respectively. If return “Mo” is clutter then Φ₀=0 corresponding to clutter having zero (0) velocity. Now employing equations (2) and (3) and solving for MoXo and M1X1 knowing Φ₀ and Φ₁ we are now to find Φ_D0 and Φ_D1 (defined previously) which are proportional to return 0 and return 1 respectively where they are detected in there filter.
We have determined Φ₀ and Φ₁ and MoXo and M1X1 and both returns are detected in the same RDB but the location in that RDB is not known. If the location is taken at the center of the filter the error in determination of azimuth is plus or minus a half a RDB. If a more accurate determination is desired a point of frequency close to first filter is created and processed like that of first filter.
This gives the same Φ₀ and Φ₁ and different M₀ and M₁ and the ratio of the M₀ and M₁ should give a good estimate of where the position Φ_D0 and Φ_D1 is detected at in the RDB. From this an estimate of azimuth of both returns determined. To get a more accurate determination another frequency may be processed or a slight change in the range processed and results correlated for best results.
From a second set of data a small known change in frequency from the first set of data. We assume there will be no change in the channel balancing terms A_M0,Ψ_M0 which are the amplitude and phase term but a known DPCA term ΔK_D0 X₀ where X₀ is unknown position of new filter and ΔK_D0 DPCA known constant therefore the term is unknown and performing a small change in frequency as done previously second set of data, the solutions is then determined for Φ₁ and Φ₀ and the changes are in X₀₁ to X₀₂ e^jΔX0*ΔKD0 where the phase term is known and represents the known change in frequency and known position change in peak of filter. The same is for the second return X₁₁ to X₁₂e^jΔKD0X1 and analogous definitions of terms. The returns change due to frequency change from M₀ to M′₀ where X_F0=M′₀/M₀ and M₁ to M′₁ where X_F1=M′₁/M₁ and solving for M₀X₀ and M₀X₀e^j(KD1X0) in the first and second set of data and dividing the terms we get X_F0 e^j(KD1X0)=M′₀/M₀ e^j(KD1X0). Therefore we have determined the ratio of the returns from which we estimate the position of the peak where is the first return and from that calculate the azimuth of the return. We can analogously perform that for the second return. The aforementioned analysis may be applied to many returns per RDB but as the number of returns increases the difficulty of determining the phases of the returns becomes much more difficult up to four and more returns is cumbersome are complicated. More said about that later in the document.
B—This development is for three (3) returns with DPCA operation.
V₀₀=M₀+M₁+M₂ CHANNEL 1-TIME 1  (1′)
V₀₁=M₀X₀₁+M₁X₁₁+M₂X₂₁CHANNEL 2-TIME 1  (2′)
V₀₂=M₀X₀₂e^jθ0+M₁₂X₁₂e^jΦ1+M₂X₁₂e^jΦ2 CHANNEL 2-TIME 2  (3′)
V₀₃=M₀X₀₃e^j2Φ0+M₁₃X₁₃e^j2Φ1+M₂X₂₃e^j2Φ2 CHANNEL 2-TIME 3  (4′)
V₀₄=M₀X₀₄e^j3Φ0+M₁₄X₁₄e^j3Φ1+M₂X₂₄e^j3Φ1 CHANNEL 2-TIME 4  (5′)
X₀₁=e^jΦ0/W_M0=e^jΦ0/A_M0e^{j(ΨM0+KD0X0)} where Φ₀=0  (6′)
X₁₁=e^jθ0/W_M1=e^jΦ1/W_M1=e^jΦ1/A_M1e^{j(ΨM1+KD0X1)} where Φ₁=0  (7′)
X₁₂=e^jΦ2/W_M2=e^jΦ2/A_M2e^{j(ΨM2+KD0X2)} where Φ₂=0  (8′)
X₀₂=X₀₁e^jΦ0/W_M0=1/A_M0e^{j(Φ0−ΔKD0X0−ΨM0)}  (9′)
X₁₂=e^jΦ1/W_M1=e^jΦ1/A_M1e^{j(Φ1−ΔKD0X1−ΨM1)}  (10′)
X₂₂=e^jΦ2/W_M2=e^jΦ2/A_M2e^{j(Φ2−ΔKD0X2−ΨM2)}  (11′)
X₀₃=X₀₁e^jΦ1/W_M0=1/A_M0e^{j2(Φ0−ΔKD0X0) -ΨM0)}  (12′)
X₁₃=e^jΦ1/W_M1=e^jΦ1/A_M1e^{j2(Φ1−ΔKD0X1)−ΨM1)}  (13′)
X₂₃=e^jΦ2/W_M2=e_jΦ²/AM2e^{j2(Φ2−ΔKD0X2)−ΨM2)}  (14′)

Rewriting equations (1) to (5) incorporating development above we have the following:
V₀₀=M₀+M₁+M₂ CHANNEL 1-TIME 1  (1′)
V₀₁=M₀X₀₁+M₁X₁₁+M₂X₂₁CHANNEL 2-TIME 1  (2′)
V₀₂=M₀X₀₂e^j2θ0+M₁₂X₁₂e^jθ1+M₂X₁₂e^jθ2 CHANNEL 2-TIME 2  (3′)
V₀₃=M₀X₀₃e^j2θ0+M₁₃X₁₃e^j2θ1+M₂X₂₃e^j2θ2 CHANNEL 2-TIME 3  (4′)
V₀₄=M₀X₀₄e^j3θ0+M₁₄X₁₄e^j3θ1+M₂X₂₄e^{j3 θ1} CHANNEL 2-TIME 4  (5′)

Taking equations (2′) and (3′) and (4′) and treating M₀X₀ and M₁X₁ and M₂X₂ as the variables and solving the determinant equation for Δ₀ we have: Δ 0 = ⁢ ⅇ jθ 0 1 ⁢ ⅇ jθ 1 1 ⁢ ⅇ jθ 2 1 = ⁢ 1 ⁢ ⅇ jθ 1 ⁢ ⅇ jθ 2 - 1 ⁢ ⅇ jθ 0 ⁢ ⅇ jθ 2 + ⁢ 1 ⁢ ⅇ jθ 0 ⁢ ⅇ jθ 1 ⁢ ⅇ j2θ 0 ⁢ ⅇ j2θ 1 ⁢ ⅇ j2θ 2 ⁢ ⅇ j2θ 1 ⁢ ⅇ j2θ 2 ⁢ ⅇ j2θ 0 ⁢ ⅇ j2θ 2 ⁢ ⅇ j2θ 0 ⁢ ⅇ j2θ 1 = ⁢ ⅇ jθ 1 ⁢ ⅇ j2θ 2 - ⅇ jθ 2 ⁢ ⅇ j2θ 1 - ⅇ jθ 0 ⁢ ⅇ j2θ 2 + ⅇ jθ 2 ⁢ ⅇ j2θ 0 + ⁢ ⅇ jθ 0 ⁢ ⅇ j2θ 1 - ⅇ jθ 1 ⁢ ⅇ j2θ 0 = ⁢ function ⁢   ⁢ of ⁢   ⁢ θ 0 , θ 1 ⁢   ⁢ and ⁢   ⁢ θ 2 ⁢ ⁢ and ⁢   ⁢ solving ⁢   ⁢ for ⁢ ⁢ M 0 ⁢ X 01 _ = M 0 ⁢ X 01 * Δ 0 ⁢ ⁢ M 0 ⁢ X 01 _ = V 02 V 01 ⁢ ⅇ jθ 1 1 ⁢ ⅇ jθ 2 1 = V 01 ⁢ ⅇ jθ 1 ⁢ ⅇ jθ 2 - V 02 ⁢ 1 ⁢ ⁢ 1 ⁢ ⁢ V 03 ⁢ 1 ⁢ ⁢ 1 ⁢ ⁢ V 03 ⁢ ⅇ j2θ 1 ⁢ ⅇ j2θ 2 ⁢ ⅇ j2θ 1 ⁢ ⅇ j2θ 2 ⁢ ⅇ j2θ 1 ⁢ ⅇ j2θ 2 ⁢ ⅇ j2θ 1 ⁢ ⅇ j2θ 2   M 0 ⁢ X 01 _ = ⁢ V 01 ⁡ ( ⅇ jθ 1 ⁢ ⅇ j2θ 2 - ⅇ jθ 2 ⁢ ⅇ j2θ 1 ) - ⁢ V 02 ⁡ ( ⅇ j2θ 2 - ⅇ j2θ 1 ) + V 03 ⁡ ( ⅇ jθ 1 - ⅇ jθ 2 ) = ⁢ function ⁢   ⁢ of ⁢   ⁢ θ 1 ⁢   ⁢ and ⁢   ⁢ θ 2 ⁢ ⁢ and ⁢   ⁢ solving ⁢   ⁢ for ( 6 ) M 1 ⁢ X 11 _ = ⁢ M 1 ⁢ X 11 * Δ 0 = ⁢ V 02 ⁡ ( ⅇ jθ 0 ⁢ ⅇ j2θ 2 - ⅇ jθ 2 ⁢ ⅇ j2θ 0 ) - ⁢ V 03 ⁡ ( ⅇ j2θ 2 - ⅇ j2θ 0 ) + V 04 ⁡ ( ⅇ jθ 2 - ⅇ jθ 0 ) = ⁢ function ⁢   ⁢ of ⁢   ⁢ θ 0 ⁢   ⁢ and ⁢   ⁢ θ 2 ⁢ ⁢ and ⁢   ⁢ solving ⁢   ⁢ for ⁢ ⁢ M 2 ⁢ X 21 _ = ⁢ M 2 ⁢ X 21 * Δ 0 = ⁢ V 02 ⁡ ( ⅇ jθ 0 ⁢ ⅇ j2θ 1 - ⅇ jθ 1 ⁢ ⅇ j2θ 0 ) - ⁢ V 03 ⁡ ( ⅇ j2θ 1 - ⅇ j2θ 0 ) + V 04 ⁡ ( ⅇ jθ 1 - ⅇ jθ 0 ) = ⁢ function ⁢   ⁢ of ⁢   ⁢ θ 0 ⁢   ⁢ and ⁢   ⁢ θ 1 ⁢ ⁢ M 2 ⁢ X 21 _ = ⅇ jθ 0 1 ⁢ ⅇ jθ 1 1 ⁢ V 02 V 02 = V 01 ⁢ ⅇ jθ 0 ⁢ ⅇ jθ 1 - V 02 ⁢ 1 ⁢ ⁢ 1 + V 03 ⁢ 1 ⁢ ⁢ 1 ⁢ ⁢ ⅇ j2θ 0 ⁢ ⅇ j2θ 1 ⁢ V 03 ⁢ ⅇ j2θ 0 ⁢ ⅇ j2θ 1 ⁢ ⅇ j2θ 0 ⁢ ⅇ j2θ 1 ⁢ ⅇ jθ 0 ⁢ ⅇ jθ 1 ( 7 ) M 2 ⁢ X 21 _ = ⁢ V 01 ⁡ ( ⅇ jθ 0 ⁢ ⅇ j2θ 1 - ⅇ jθ 1 ⁢ ⅇ j2θ 0 ) - V 02 ⁢ ( ⅇ j2θ 1 - ⅇ j2θ 0 ) + ⁢ V 03 ⁡ ( ⅇ jθ 1 - ⅇ jθ 0 ) = ⁢ function ⁢   ⁢ of ⁢   ⁢ θ 1 ⁢   ⁢ and ⁢   ⁢ θ 2 ( 8 )
Performing the same operations on equations (3), (4) and (5) with the variables M₀X₀₁ e^jθ0, M₁X₁₁e^jθ1 and M₂X₁₂e^jθ0 and Δ₀ remains the same. The analysis is analogous and the result will be the following: and ⁢   ⁢ solving ⁢   ⁢ for ⁢ ⁢ M 0 ⁢ X 01 ⁢ ⅇ jθ 0 _ = M 0 ⁢ X 01 ⁢ ⅇ jθ 0 * Δ 0   M 0 ⁢ X 01 _ ⁢ ⅇ jθ 0 _ = ⁢ V 02 ⁡ ( ⅇ jθ 1 ⁢ ⅇ j2θ 2 - ⅇ jθ 2 ⁢ ⅇ j2θ 1 ⁢ 1 ) - ⁢ V 03 ⁡ ( ⅇ j2θ 2 - ⅇ j2θ 1 ) + V 04 ⁡ ( ⅇ jθ 2 - ⅇ jθ 1 ) = ⁢ function ⁢   ⁢ of ⁢   ⁢ θ 1 ⁢   ⁢ and ⁢   ⁢ θ 2 ⁢ ⁢ and ⁢   ⁢ solving ⁢   ⁢ for ⁢ ⁢ M 1 ⁢ X 11 ⁢ ⅇ jθ 1 _ = M 1 ⁢ X 11 ⁢ ⅇ jθ 1 * Δ 0 ( 9 ) M 1 ⁢ X 11 _ ⁢ ⅇ jθ 1 _ = ⁢ V 02 ⁡ ( ⅇ jθ 0 ⁢ ⅇ j2θ 2 - ⅇ jθ 2 ⁢ ⅇ j2θ 0 ) - ⁢ V 03 ⁡ ( ⅇ j2θ 2 - ⅇ j2θ 0 ) + V 04 ⁡ ( ⅇ jθ 2 - ⅇ jθ 0 ) = ⁢ function ⁢   ⁢ of ⁢   ⁢ θ 0 ⁢   ⁢ and ⁢   ⁢ θ 2 ⁢ ⁢ and ⁢   ⁢ solving ⁢   ⁢ for ⁢  ⁢ M 2 ⁢ X 21 ⁢ ⅇ jθ 2 _ = M 2 ⁢ X 21 ⁢ ⅇ jθ 2 * Δ 0 ( 10 ) M 2 ⁢ X 21 _ ⁢ ⅇ jθ 2 _ = ⁢ V 02 ⁡ ( ⅇ jθ 0 ⁢ ⅇ j2θ 1 - ⅇ jθ 1 ⁢ ⅇ j2θ 0 ) - ⁢ V 03 ⁡ ( ⅇ j2θ 1 - ⅇ j2θ 0 ) + V 04 ⁡ ( ⅇ jθ 1 - ⅇ jθ 0 ) = ⁢ function ⁢   ⁢ of ⁢   ⁢ θ 0 ⁢   ⁢ and ⁢   ⁢ θ 1 ( 11 )
Taking the equations (9)/(6) and solve we have the following: ⅇ jθ 0 = ⁢ V 03 ⁡ ( ⅇ jθ 1 ⁢ ⅇ j2θ 2 - ⅇ jθ 2 ⁢ ⅇ j2θ 1 ) - V 04 ⁡ ( ⅇ j2θ 2 - ⅇ j2θ 1 ) + ⁢ V 05 ⁡ ( ⅇ j2θ 2 - ⅇ j2θ 1 ) / ⁢ V 02 ⁡ ( ⅇ jθ 1 ⁢ ⅇ j2θ 2 - ⅇ jθ 2 ⁢ ⅇ j2θ 1 ) - V 03 ⁡ ( ⅇ j2θ 2 - ⅇ j2θ 1 ) + ⁢ V 04 ⁡ ( ⅇ j2θ 2 - ⅇ j2θ 1 ) = ⁢ function ⁢   ⁢ of ⁢   ⁢ θ 1 ⁢   ⁢ and ⁢   ⁢ θ 2 ( 12 ″ )
Taking the previous equations (10)/(7) and solve we have the following: ⅇ jθ 1 = ⁢ V 03 ⁡ ( ⅇ jθ 0 ⁢ ⅇ j2θ 2 - ⅇ jθ 2 ⁢ ⅇ j2θ 0 ) - V 04 ⁡ ( ⅇ j2θ 2 - ⅇ j2θ 0 ) + ⁢ V 05 ⁡ ( ⅇ j2θ 2 - ⅇ j2θ 0 ) / ⁢ V 02 ⁡ ( ⅇ jθ 0 ⁢ ⅇ j2θ 2 - ⅇ jθ 2 ⁢ ⅇ j2θ 0 ) - V 03 ⁡ ( ⅇ j2θ 2 - ⅇ j2θ 0 ) + ⁢ V 04 ⁡ ( ⅇ j2θ 2 - ⅇ j2θ 0 ) = ⁢ function ⁢   ⁢ of ⁢   ⁢ θ 0 ⁢   ⁢ and ⁢   ⁢ θ 2 ( 13 ″ )
Taking the previous equation (11)/(8) and solve we have the following: ⅇ jθ 2 = ⁢ V 03 ⁡ ( ⅇ jθ 0 ⁢ ⅇ j2θ 1 - ⅇ jθ 1 ⁢ ⅇ j2θ 0 ) - V 04 ⁡ ( ⅇ j2θ 1 - ⅇ j2θ 0 ) + ⁢ V 05 ⁡ ( ⅇ j2θ 1 - ⅇ j2θ 0 ) / ⁢ V 02 ⁡ ( ⅇ jθ 0 ⁢ ⅇ j2θ 1 - ⅇ jθ 1 ⁢ ⅇ j2θ 0 ) - V 03 ⁡ ( ⅇ j2θ 1 - ⅇ j2θ 0 ) + ⁢ V 04 ⁡ ( ⅇ j2θ 1 - ⅇ j2θ 0 ) = ⁢ function ⁢   ⁢ of ⁢   ⁢ θ 0 ⁢   ⁢ and ⁢   ⁢ θ 2 ( 14 ″ )
Substituting equation (13″) into equation (12″) we have the following:
θ₀ as a function of θ₂
Substituting equation (13″) into equation (14″) we have the following:
θ₀ as a function of θ₂
we have two equations two unknowns and solvable in Φ₀ and Φ₂.
Substituting equation (12″) into equation (13″) we have the following:
θ₁ as a function of θ₂
Substituting equation (12″) into equation (14″) we have the following:
θ₁ as a function of θ₂
we have two equations two unknowns and solvable in Φ₁ as a function of Φ₂.
Substituting equation (14″) into equation (13″) we have the following:
θ₀ as a function of θ₁
Substituting equation (14″) into equation (12″) we have the following:
θ₀ as a function of θ₁
C1—Same as Assist in Section B1 for Assist in Determining Solutions for θ₀, θ₁, θ₂, ETC for More Two Returns Instead of Φ_D0, Φ_D1, Φ_D2, Etc
we have two equations two unknowns and solvable in θ₀ as a function of θ₁ Thus we have solved for θ₂ twice

From a second set of data a small known change in frequency from the first set of data. We assume there will be no change in the channel balancing terms A_M0,Ψ_M0 which are the amplitude and phase term but a known DPCA term ΔK_D0X₀ where X₀ is the unknown change in position of new filter and K_D0 DPCA known constant therefore the term is known. Performing the operations on the second set of data, the solutions are the same for θ₁ and θ₂ and the changes are in X₀₁ to X₀₁ e^j(ΔKD0X0) where the phase term is known and represents the known change in frequency and known position change in peak of filter. The same is for the second and third return X₁₁ to X₁₁e^jΔX1*ΔKD0 and X₂₁ to X₂₁e^jΔX2*ΔKD0 and analogous definitions of terms. The returns change due to frequency change from M₀ to M′₀ where X_F0=M′₀/M₀ and M₁ to M′₁ where X_F1=M′₁/M₁ and solving for M₀X₀₁ and M₀X₀₁e^j(KD0*ΔX0) in the first and second set of data and dividing the terms we get X_F0 e^j(KD0*ΔX0)=M′₀/M₀e^j(KD0*ΔX0). Therefore we have determined the ratio of the returns from which we estimate the position of the peak where is the first return and from that calculate the azimuth of the return. We can analogously perform that for the second and third return. The aforementioned analysis may be applied to many returns per RDB but as the number of returns increase the difficulty of determining the phases of the returns becomes much more difficult up to four or more returns is cumbersome is complicated. More said about that later in the document.

C2—Combining techniques of section III—Two channels “M” pulse system Φ_D technique and section IIII—ΔT technique-DPCA-Two channel “M” pulse system

The two techniques employ the same data and may be processed in any manner to facilitate a solution. The following is a list of common solutions and attributes.

1. Same solutions for all parameters such as the following:

• • a) Φ_{D s}, Φ_{A s} and Φs and M s and X_{F s}, X_{R s} and X_{H s}
  • b) Φ_{D s} have many same solutions in the Φ_D technique
  • c) solutions for position and velocity of respective returns
• 2. Φ_D technique is more effective but requires more storage and processing but requires one less delay in data

3. Accuracy and robustness of solutions are enhanced

4. Appearance of a very practical system

D—Analyzing the relationship (Duality) between one and two pulse systems to that of the one and two channel systems. The one and two channel systems employ many pulses in time which the spectrum (transformed to frequency) are obtained while the one and two pulse systems employ many channels (space elements) which the spectrum (transformed to frequency) are obtained.

1. The many pulse system and the many channel system the detection of clutter is the azimuth of detection while movers have two components one due to there azimuth and the other due to their radial velocity. The total frequency component is the addition of these components.

2. The many pulse system processes all detections of returns in the same RDB and the many channel system processes all detections of returns in the same RAB.

3. The many pulse system processes to determine velocity of returns and then calculates azimuth. The many channel system processes to determine azimuth of returns and then calculates velocity.

4. Calculation of DPCA and channel balancing factors are a duality since one works on space elements transformed while the other works on time elements transformed. Although one works with space but other works with time they have the same form and lend themselves to analogous equations and solutions. The channel balancing terms are in both systems depending on the azimuth of the returns. The DPCA terms depend on respective delays one in time and other in space but are analogous. Consequently they are analyzed in that manner.

5. FIGS. 7 and 8 illustrate in the multipulse systems the number of parameters that could be correlated if n additional sets of data are processed which also applies to the multichannel systems. FIG. 9 illustrates the type of operations to be performed in implementing the multichannel system.

V One Pulse and One Channel Systems

This STAP methodology employs one or more pulse or channel data but one pulse at a time or one channel at a time, process them into its frequency spectrum and consequently localize clutter into its own range azimuth bin (RAB) or range doppler bin (RDB) together with any other returns such as target, thermal noise, and others.

The subsequent processing of each bin will separate out clutter doppler since clutter has zero (0) radial velocity and other returns detected in the same bin will have different velocities. Determining the azimuth and the velocity of the returns will be calculated therefore no additional data is required.

The knowledge aided STAP will be involved to determine from the detected returns which are the targets of interest, sidelobe returns, land sea interface, thermal noise, etc.

The knowledge aided STAP will not be involved in canceling clutter but in the post processing of the returns of interest so they may be detected and there parameters measured and determine the nature of the return.

The following sections will be an analysis of various techniques with there mathematical development to accomplished these ends. It is assumed the channel or time data has been processed by FFT into there individual range azimuth bins (RABs) or (RDBs) respectively where there exist in the cases of interest clutter (0—velocity) and other returns (non “0” velocity). Initially two (2) returns will be developed; it may be clutter and moving target or two moving targets.

More returns detected in one bin will be considered such as three returns, or more.

VA1—Two Returns Only Employing Two Sets of Data Common to all One Channel and One Pulse Systems
V₀₀=M₀+M₁  (1)
V₀₁=M₀e^jΦD0+M₁^jΦD1  (2)
V₀₂=M₀e^j2ΦD0+M₁^j2ΦD1  (3)

VA1—One Channel Many Pulse Odd-Even Data and Apertures

EQUATIONS (1),(2) and (3) and (1′),(2′) and (3′) represented as follows:

(1)-channel 1	-aperture 1-time data	1, 3, 5, - - - , m − 1	odd data
(2)-		3, 5, 7, - - - , m + 1
(3)		5, 7, 9, - - - , m + 3
(1′)	-aperture 2	2, 4, 6, - - - , m	even data
(2′)		4, 6, 8, - - - , m + 2
(3′)		6, 8, 1O, - - - , m + 4

VA2—One Channel Many Pulse No Odd-Even Data and Simultaneous Apertures

EQUATIONS (1), (2) and (3) and (1′), (2′) and (3′) represents as follows:

	(1)-channel 1	-aperture 1-time data	1, 2, 3, - - - , m
	(2)-		1, 2, 3, - - - , m
	(3)		1, 2, 3, - - - , m
	(1′)-	-aperture 2	1, 2, 3, - - - , m
	(2′)		1, 2, 3, - - - , m
	(3′)		1, 2, 3, - - - , m

VA3—One Pulse Many Channels Odd-Even Data and Apertures EQUATIONS (1),(2) and (3) and (1′),(2′) and (3′) represents as follows:

(1)-PULSE 1	-aperture	1, 3, 5, - - - , N − 1	odd data
	1-CHANNEL data
(2)-		3, 5, 7, - - - , N + 1
(3)		5, 7, 9, - - - , N + 3
(1′)	-aperture 2	2, 4, 6, - - - , N	even data
(2′)		4, 6, 8, - - - , N + 2
(3′)		6, 8, 10, - - - , N + 4

VA4—One Pulse Many Channels No Odd-Even Data and Simultaneous Apertures

a) EQUATIONS (1),(2) and (3) and (1′),(2′) and (3′) represents as follows:

	(1)-PULSE 1	-aperture 1-CHANNEL data	1, 2, 3, - - - , N
	(2)-		1, 2, 3, - - - , N
	(3)		1, 2, 3, - - - , N
	(1′)-	-aperture 2	1, 2, 3, - - - , N
	(2′)		1, 2, 3, - - - , N
	(3′)		1, 2, 3, - - - , N

Taking equations (1) and (2) and treating M₀ and M₁ as the variables and solving for M₀ and M₁ we have:
M₀=(V₀₀e^jΦD1−V₀₁)/(e^jΦD1−e^jΦD0)  (5)
M₁=(V₀₁−e^jΦD0V₀₀)/(e^jΦD1−e^jΦD0)  (6)

Taking equations (2) and (3) and solving for M₀ e^jΦD0 and M₁^jΦD1 we have the following:
M₀e^jΦ_D0=(V₀₁e^jΦD1=V₀₂)/(e^{jΦD1−ejΦD0})  (7)
M₁e^jΦD1=(V₀₂−e^jΦD0V₀₁)/(e^jΦD1−e^jΦD0)  (8)
Dividing equations (7)/(5) and equations (8)/(6) we have the following:
e^jΦD0=(V₀₁e^jΦD1−V₀₂)/(V₀₀e^jΦD1−V₀₁)  (9)
e^jΦD1=(V₀₂−e^jΦD0V₀₁)/(V₀₁−e^jΦD0V₀₀)  (10)

Solving equation (9) and (10) for Φ_D0 and Φ_D1 and substituting these values into equation (1) and (2) and determine M₀ and M₁.

Now we have another set of equations for aperture (antenna) 2 formulated a beam width away from the first aperture (it can be any reasonable distance away).
V′₀₀=M′₀+M′₁  (1′)
V′₀₁=M′₀e^jΦ′D0+M′₁jΦ′^D1  (2′)
V′₀₂=M′₀e^j2Φ′D0+M′₁j2Φ′^D1  (3′)

Taking equations (1′), (2′) and (3′) and performing the identical operations as on equations (1), (2) and (3) and solving for Φ′_DO and Φ′_D1 and M′₀+M′₁ and where Φ′_DO=Φ_DO and Φ′_D1=Φ_D1
K_M0e^jα0=M′₀/M₀=(V′₀₀e^jΦ′D1−V′₀₁)/(V₀₀e^jΦ′D1−V₀₁)  (12)
K_M1e^jΦα1=M′₁/M₁=(V′₀₁−V′₀₀e^jΦ′D0)/(V₀₁−V₀₀e^jΦ′D0)  (13)
b) From the data in the significant clutter only area in the antennas we have the equation for clutter only data as follows:

The returns from a prior measurement of the ratio of return output of aperture 2, even data to aperture 1 data, odd data-vs-azimuth or real time measurement of significant clutter only data, their ratio of two apertures-vs-azimuth and compared with ratios calculated in the processing which is the position in the filter where the returns are detected.

The methodology of producing this curve from real time data is as follows:

• • a) chose only relatively large amplitude clutter only data
  • b) determine its azimuth from its measurement and its amplitude and phase ratio of the return of apertures
  • c) measure ratio of ratio of the amplitude and phase from each aperture which is K_M0e^jα0=M′₀/M₀ and K_M1e^jΦα1=M′₁/M₁
  • d) note: clutter only data is at its azimuth position—no radial velocity
  • e) perform this with as much data as to obtain a very good estimate of the curve of c) by any fitting statistical methodology
  • f) thus we have obtained a very good estimate of curve desired

Now having generated from the real time data the curve of ratio of even divided by odd-vs-azimuth the candidate Φ technique is employed where all the possible phases are substituted (which are very limited in number, in equation (12) or (13) and the solutions are compared to the curve for the ratio of aperture data as a function of azimuth and where there is a match this is the solution. The solution for equations (12) and (13) should agree and they are a check on each other. Having attained the solution for azimuth Φ_A and having the solution for Φ_D then Φ=Φ_D−ΦA, the solution for radial velocity of the returns.

From the ratio of the second set of data to the first set of data we obtain K_M0e^jα0 and K_M1e^jα1 which is the ratio of M′₀/M₀=K_M0e^jα0 and M′₁/M₁=K_M1e^jα1 where all other terms are known.

c) A close range is now processed. This is processed to yield a second set of equations and processing similar to first set of data.

Now employing equations (1) and (2) and solving for M₀ and M₁ knowing Φ_A0 and Φ_A1 we are now to find Φ_R0 and Φ_R1 which is the detected position of 0 and 1 return is at its peak in the range bin, are proportional to range of return 0 and return 1 respectively. and M₀ and M₁ and both returns are detected in the same range bin (RB) but the location in that RB is not known. If the location is taken at the center of the filter the error in determination of range is plus or minus a half a RB. If a more accurate determination is desired a point of range close to first filter is created and processed like that of first data set this will be used to a more accurate determination of the range of both returns.

This gives the same Φ_A0 and Φ_A1 and different M₀ and M₁ and the ratios of the M₀ and M₁ should give a good estimate of where the position Φ_R0 and Φ_R1 is detected in the RB. From this an estimate of better range of both returns is determined. To get a more accurate determination, another range or ranges may be processed or a slight change in the range processed and results correlated for best results. The ratio of second set of data for close in range we will obtain a much more accurate range.

The linear array height amplitude and phase changes in the different linear arrays should be the same.

The X_R and X_H should be all equal in the both sets of data and is another determination of the correct results

Up to this point all sections VA1-VA4 have a common set of equations and solutions.

In the one channel system if the time data is delayed a significant time and processed like the first set of data the amplitude and phase change of the range will give the unambiguous velocity from which the ambiguous velocity is calculated and will be a check on the velocity that has been determined. This may be performed as many times as necessary and may eliminate the need for another PRF to increase the ambiguous velocity.

Processing similarly for change in doppler frequency will give the horizontal tangential velocity and for height change in the linear arrays will give vertical tangential velocity.

Thus range, azimuth and radial velocity unambiguously and vertical and horizontal tangential velocity has been attained.

In the one pulse system the precise range and height will be as in the one channel system but precise frequency will not be the same meaning. The significant delay in channel will give these parameters as a function of channel change.

Additional aids in one pulse and one channel systems the attaining the solutions are to employ another delayed set of data and results should agree with the first set of data processed. Further if another close range bin is processed the change in range in both sets of data processed should be the same. This holds for the height and doppler or azimuth change is the same.

VB—Three Returns Only Employing Three Sets of Data Common To all One Channel and One Pulse Systems
V₀₀=M₀+M₁+M₂  (1)
V₀₁=M₀e^jΦD0+M₁^jΦD1+M₂jΦ^D2  (2)
V₀₂=M₀e^j2ΦD0+M₁^j2ΦD1+M₂^j2ΦD2  (3)
V₀₃=M₀e^j3ΦD0+M₁^j3ΦD1+M₂^j3ΦD2  (4)
VB1—One Channel Many Pulse Odd-Even Data and Apertures

EQUATIONS (1),(2),(3) and (4) and (1′),(2′),(3′) and (4′) represented as follows:

(1)-channel 1	-aperture 1-time data	1, 3, 5, - - - , m − 1	odd data
(2)-		3, 5, 7, - - - , m + 1
(3)		5, 7, 9, - - - , m + 3
(4)		7, 9, 11, - - - , m + 5
(1′)	-aperture 2	2, 4, 6, - - - , m	even data
(2′)		4, 6, 8, - - - , m + 2
(3′)		6, 8, 10, - - - , m + 4
(4′)		8, 10, 12, - - - , m + 6

VB2—One Channel Many Pulse No Odd-Even Data and Simultaneous Apertures

EQUATIONS (1), (2), (3) and (4) and (1′), (2′), (3′) and (4′) represented as follows:

	(1)-channel 1	-aperture 1-time data	1, 2, 3, - - - , m
	(2)-		1, 2, 3, - - - , m
	(3)		1, 2, 3, - - - , m
	(4)		1, 2, 3, - - - , m
	(1′)-	-aperture 2	1, 2, 3, - - - , m
	(2′)		1, 2, 3, - - - , m
	(3′)		1, 2, 3, - - - , m
	(4′)		1, 2, 3, - - - , m

VB3—One Pulse Many Channels Odd-Even Data and Apertures

EQUATIONS (1),(2),(3) and (4) and (1′),(2′),(3′) and (4′) represented as follows:

(1)-PULSE 1	-aperture	1, 3, 5, - - - , N − 1	odd data
	1-CHANNEL data
(2)-		3, 5, 7, - - - , N + 1
(3)		5, 7, 9, - - - , N + 3
(4)		7, 9, 11 - - - , N + 5
(1′)	-aperture 2	2, 4, 6, - - - , N	even data
(2′)		4, 6, 8, - - - , N + 2
(3′)		6, 8, 10, - - - , N + 4
(4′)		8, 10, 12 - - - , N + 6

VB4—One Pulse Many Channels No Odd-Even Data and Simultaneous Apertures

EQUATIONS (1),(2),(3) and (4) and (1′),(2′),(3′) and (4′) represented as follows:

	(1)-PULSE 1	-aperture 1-CHANNEL data	1, 2, 3, - - - , N
	(2)-		1, 2, 3, - - - , N
	(3)		1, 2, 3, - - - , N
	(4)		1, 2, 3, - - - , N
	(1′)-	-aperture 2	1, 2, 3, - - - , N
	(2′)		1, 2, 3, - - - , N
	(3′)		1, 2, 3, - - - , N
	(4′)		1, 2, 3, - - - , N

Taking equations (1) and (2) and (3) and treating M₀ and M₁ and M₂ as the variables and solving the determinant equation for Δ₀ we have: Δ 0 = ⅇ jΦ D ⁢   ⁢ 0 1 ⁢ ⅇ jΦ D ⁢   ⁢ 1 1 ⁢ ⅇ jΦ D ⁢   ⁢ 2 1 = 1 ⁢ ⁢ ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⅇ jΦ D ⁢   ⁢ 2 - 1 ⁢ ⁢ ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ jΦ D ⁢   ⁢ 2 + 1 ⁢ ⁢ ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⁢ Δ 0 = ⁢ ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⅇ jΦ D ⁢   ⁢ 2 - ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 + ⁢ ⅇ jΦ D ⁢   ⁢ 2 ⁢ ⅇ j2Φ D ⁢   ⁢ 0 + ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 - ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 0 = ⁢ function ⁢   ⁢ of ⁢   ⁢ Φ D ⁢   ⁢ 0 , Φ D ⁢   ⁢ 1 ⁢   ⁢ and ⁢   ⁢ Φ D ⁢   ⁢ 2 ⁢ ⁢ and ⁢   ⁢ solving ⁢   ⁢ for ⁢ ⁢ M 0 _ = M 0 * Δ 0   M 0 _ = V 01 V 00 ⁢ ⅇ jΦ D ⁢   ⁢ 1 1 ⁢ ⅇ jΦ D ⁢   ⁢ 2 1 = V 00 ⁢ ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⅇ jΦ D ⁢   ⁢ 2 - V 01 ⁢ 1 ⁢ ⁢ 1 + V 02 ⁢ 1 ⁢ ⁢ 1 ⁢ ⁢ V 02 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ V ⁢   ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⁢ ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⅇ jΦ D ⁢   ⁢ 2 ⁢ ⁢ M 0 _ = ⁢ V 00 ⁡ ( ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ jΦ D ⁢   ⁢ 2 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ) - ⁢ V 01 ( ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ j2Φ D ⁢   ⁢ 1 ) + V 02 ⁡ ( ⅇ jΦ D ⁢   ⁢ 2 - ⅇ jΦ D ⁢   ⁢ 1 ) = ⁢ function ⁢   ⁢ of ⁢   ⁢   ⁢ Φ D ⁢   ⁢ 1 ⁢   ⁢ and ⁢   ⁢ Φ D ⁢   ⁢ 2 ⁢ ⁢ and ⁢   ⁢ solving ⁢   ⁢ for ⁢ ⁢ M 1 _ = M 1 * Δ 0 ( 6 ) M 1 _ = ⅇ jΦ D ⁢   ⁢ 0 1 ⁢ V 01 V 00 ⁢ ⅇ jΦ D ⁢   ⁢ 2 1 = V 00 ⁢ ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ jΦ D ⁢   ⁢ 2 - V 01 ⁢ 1 ⁢ ⁢ 1 + V 02 ⁢ 1 ⁢ ⁢ 1 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ V 02 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ⁢ ⁢ ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ jΦ D ⁢   ⁢ 2 ⁢ ⁢ M 1 _ = ⁢ V 01 ⁡ ( ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ jΦ D ⁢   ⁢ 2 ) - ⁢ V 01 ( ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ j2Φ D ⁢   ⁢ 0 ) + V 02 ⁡ ( ⅇ jΦ D ⁢   ⁢ 2 - ⅇ jΦ D ⁢   ⁢ 0 ) = ⁢ function ⁢   ⁢ of ⁢   ⁢ Φ D ⁢   ⁢ 0 ⁢   ⁢ and ⁢   ⁢   ⁢ Φ D ⁢   ⁢ 2 ⁢ ⁢ and ⁢   ⁢ solving ⁢   ⁢ for ⁢ ⁢ M 2 _ = M 2 * Δ 0 ( 7 ) M 2 _ = ⅇ jΦ D ⁢   ⁢ 0 1 ⁢ ⅇ jΦ D ⁢   ⁢ 1 1 ⁢ V 01 V 00 = V 00 ⁢ ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ jΦ D ⁢   ⁢ 1 - V 01 ⁢ 1 ⁢ ⁢ 1 + V 02 ⁢ 1 ⁢ ⁢ 1 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ V 02 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⁢ ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⁢ M 2 _ = ⁢ V 00 ⁡ ( ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 - ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⅇ jΦ D ⁢   ⁢ 0 ) - ⁢ V 01 ( ⅇ j2Φ D ⁢   ⁢ 1 - ⅇ j2Φ D ⁢   ⁢ 0 ) + V 02 ⁡ ( ⅇ jΦ D ⁢   ⁢ 1 - ⅇ jΦ D ⁢   ⁢ 0 ) = ⁢ function ⁢   ⁢ of ⁢   ⁢ Φ D ⁢   ⁢ 0 ⁢   ⁢ and ⁢   ⁢   ⁢ Φ D ⁢   ⁢ 2 ( 8 )

Taking equation (2), (3) and (4) and performing the identical operations as in equation (1), (2) and (3) we have the following: M 0 _ ⁢ ⅇ jΦ D ⁢   ⁢ 0 _ = ⁢ V 01 ⁡ ( ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ jΦ D ⁢   ⁢ 2 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ) - ⁢ V 02 ⁡ ( ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ j2Φ D ⁢   ⁢ 1 ) + V 03 ⁡ ( ⅇ jΦ D ⁢   ⁢ 2 - ⅇ jΦ D ⁢   ⁢ 1 ) = ⁢ function ⁢   ⁢ ( Φ D ⁢   ⁢ 1 ⁢   ⁢ and ⁢   ⁢ Φ D ⁢   ⁢ 2 ) ( 9 ) M 1 _ ⁢ ⅇ jΦ D ⁢   ⁢ 1 _ = ⁢ V 01 ⁡ ( ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ) - ⁢ V 02 ⁡ ( ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ j2Φ D ⁢   ⁢ 0 ) + V 03 ⁡ ( ⅇ jΦ D ⁢   ⁢ 2 - ⅇ jΦ D ⁢   ⁢ 0 ) = ⁢ function ⁢   ⁢ ( Φ D ⁢   ⁢ O ⁢   ⁢ and ⁢   ⁢ Φ D ⁢   ⁢ 1 ) ( 10 ) M 2 _ ⁢ ⅇ jΦ D ⁢   ⁢ 2 _ = ⁢ V 01 ⁡ ( ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 - ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ) - ⁢ V 02 ⁡ ( ⅇ j2Φ D ⁢   ⁢ 1 - ⅇ j2Φ D ⁢   ⁢ 0 ) + V 03 ⁡ ( ⅇ jΦ D ⁢   ⁢ 1 - ⅇ jΦ D ⁢   ⁢ 0 ) = ⁢ function ⁢   ⁢ ( Φ D ⁢   ⁢ 0 ⁢   ⁢ and ⁢   ⁢ Φ D ⁢   ⁢ 1 ) ( 11 )

Dividing equation (9)/(6) and (10)/(7) and (11)/(8) we have the following: ⅇ   ⁢ jΦ D ⁢   ⁢ 0 = ⁢ V 01 ⁡ ( ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ jΦ D ⁢   ⁢ 2 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ) - ⁢ V 02 ⁡ ( ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ j2Φ D ⁢   ⁢ 1 ) + V 03 ⁡ ( ⅇ jΦ D ⁢   ⁢ 2 - ⅇ jΦ D ⁢   ⁢ 1 ) / ⁢ V 00 ⁡ ( ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ jΦ D ⁢   ⁢ 2 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ) - ⁢ V 01 ⁡ ( ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ j2Φ D ⁢   ⁢ 1 ) + V 02 ⁡ ( ⅇ jΦ D ⁢   ⁢ 2 - ⅇ jΦ D ⁢   ⁢ 1 ) = ⁢ function ⁢   ⁢ ( Φ D ⁢   ⁢ 1 , Φ D ⁢   ⁢ 2 ) ( 12 ) ⅇ   ⁢ jΦ D ⁢   ⁢ 1 = ⁢ V 01 ⁡ ( ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ) - ⁢ V 02 ⁡ ( ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ j2Φ D ⁢   ⁢ 0 ) + V 03 ⁡ ( ⅇ jΦ D ⁢   ⁢ 2 - ⅇ jΦ D ⁢   ⁢ 0 ) / ⁢ V 01 ⁡ ( ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ) - ⁢ V 01 ⁡ ( ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ j2Φ D ⁢   ⁢ 0 ) + V 02 ⁡ ( ⅇ jΦ D ⁢   ⁢ 2 - ⅇ jΦ D ⁢   ⁢ 0 ) = ⁢ function ⁢   ⁢ ( Φ D ⁢   ⁢ 0 , Φ D ⁢   ⁢ 2 ) ( 13 ) ⅇ   ⁢ jΦ D ⁢   ⁢ 2 = ⁢ V 01 ⁡ ( ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 - ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ) - ⁢ V 02 ⁡ ( ⅇ j2Φ D ⁢   ⁢ 1 - ⅇ j2Φ D ⁢   ⁢ 0 ) + V 03 ⁡ ( ⅇ jΦ D ⁢   ⁢ 1 - ⅇ jΦ D ⁢   ⁢ 0 ) / ⁢ V 00 ⁡ ( ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 - ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ) - ⁢ V 01 ⁡ ( ⅇ j2Φ D ⁢   ⁢ 1 - ⅇ j2Φ D ⁢   ⁢ 0 ) + V 02 ⁡ ( ⅇ jΦ D ⁢   ⁢ 1 - ⅇ jΦ D ⁢   ⁢ 0 ) = ⁢ function ⁢   ⁢ ( Φ D ⁢   ⁢ 0 , Φ D ⁢   ⁢ 1 ) ( 14 )

The equations for the other data as follows:
V′₀₀=M′₀+M′₁+M′₂  (1′)
V′₀₁=M′₀e^jΦ′D0+M′₁^jΦ′D1+M′₂^j2Φ′D2  (2′)
V′₀₂=M′₀e^j2Φ′D0+M′₁^j2Φ′D1+M′₂^j3Φ′D2  (3′)
V′₀₃=M′₀e^j3Φ′D0+M′₁^j3Φ′D1+M′₂j3Φ′^D2  (4′)

Taking equations (1′ to 4′) and solving identical to equations (1) to (4) we have the following: ⅇ   ⁢ j ⁢   ⁢ Φ D ⁢   ⁢ 0 ′ = ⁢ V 01 ′ ( ⅇ jΦ D ⁢   ⁢ 1 ′ ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ′ - ⅇ jΦ D ⁢   ⁢ 2 ′ ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ′ ) - ⁢ V 02 ′ ( ⅇ j2Φ D ⁢   ⁢ 2 ′ - ⅇ j2Φ D ⁢   ⁢ 1 ′ ) + V 03 ′ ( ⅇ jΦ D ⁢   ⁢ 2 ′ - ⅇ jΦ D ⁢   ⁢ 1 ′ ) / ⁢ V 00 ′ ( ⅇ jΦ D ⁢   ⁢ 1 ′ ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ′ - ⅇ jΦ D ⁢   ⁢ 2 ′ ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ′ ) - ⁢ V 01 ′ ( ⅇ j2Φ D ⁢   ⁢ 2 ′ - ⅇ j2Φ D ⁢   ⁢ 1 ′ ) + V 02 ′ ( ⅇ jΦ D ⁢   ⁢ 2 ′ - ⅇ jΦ D ⁢   ⁢ 1 ′ ) = ⁢ function ⁢   ⁢ ( Φ D ⁢   ⁢ 1 ′ , Φ D ⁢   ⁢ 2 ′ ) ( 12 ′ ) ⅇ   ⁢ jΦ D ⁢   ⁢ 1 ′ = ⁢ V 01 ′ ( ⅇ jΦ D ⁢   ⁢ 0 ′ ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ′ - ⅇ jΦ D ⁢   ⁢ 0 ′ ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ′ ) - ⁢ V 02 ′ ( ⅇ j2Φ D ⁢   ⁢ 2 ′ - ⅇ j2Φ D ⁢   ⁢ 0 ′ ) + V 03 ′ ( ⅇ jΦ D ⁢   ⁢ 2 ′ - ⅇ jΦ D ⁢   ⁢ 0 ′ ) / ⁢ V 01 ′ ( ⅇ jΦ D ⁢   ⁢ 0 ′ ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ′ - ⅇ jΦ D ⁢   ⁢ 0 ′ ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ′ ) - ⁢ V 01 ′ ( ⅇ j2Φ D ⁢   ⁢ 2 ′ - ⅇ j2Φ D ⁢   ⁢ 0 ′ ) + V 02 ′ ( ⅇ jΦ D ⁢   ⁢ 2 ′ - ⅇ jΦ D ⁢   ⁢ 0 ′ ) = ⁢ function ⁢   ⁢ ( Φ D ⁢   ⁢ 0 ′ , Φ D ⁢   ⁢ 2 ′ ) ( 13 ′ ) ⅇ   ⁢ jΦ D ⁢   ⁢ 2 ′ = ⁢ V 01 ′ ( ⅇ jΦ D ⁢   ⁢ 0 ′ ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ′ - ⅇ jΦ D ⁢   ⁢ 1 ′ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ′ ) - ⁢ V 02 ′ ( ⅇ j2Φ D ⁢   ⁢ 1 ′ - ⅇ j2Φ D ⁢   ⁢ 0 ′ ) + V 03 ′ ( ⅇ jΦ D ⁢   ⁢ 1 ′ - ⅇ jΦ D ⁢   ⁢ 0 ′ ) / ⁢ V 00 ′ ( ⅇ jΦ D ⁢   ⁢ 0 ′ ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ′ - ⅇ jΦ D ⁢   ⁢ 1 ′ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ′ ) - ⁢ V 01 ′ ( ⅇ j2Φ D ⁢   ⁢ 1 ′ - ⅇ j2Φ D ⁢   ⁢ 0 ′ ) + V 02 ′ ( ⅇ jΦ D ⁢   ⁢ 1 ′ - ⅇ jΦ D ⁢   ⁢ 0 ′ ) = ⁢ function ⁢   ⁢ ( Φ D ⁢   ⁢ 0 ′ , Φ D ⁢   ⁢ 1 ′ ) ⁢ ⁢ where ⁢ ⁢ Φ D ⁢   ⁢ 0 = Φ D ⁢   ⁢ 0 ′ ; Φ D ⁢   ⁢ 1 = Φ D ⁢   ⁢ 1 ′ ⁢   ⁢ and ⁢   ⁢ Φ D ⁢   ⁢ 2 = Φ D ⁢   ⁢ 2 ′ ( 14 ′ )

Taking equations (12) to (14) and equations (12′) to (14′) and solving in both sets of equations for Φ_D0, Φ_D1 and Φ_D2 solution which should be equal is the following procedure:

Substituting equation (12) or (12′) into equation (13) or (13′) we have the following:

Φ_D0 as a function of Φ_D2

Substituting equation (13) or (13′) into equation (14) or (14′) we have the following:

Φ_D0 as a function of Φ_D2

Substituting equation (14) or (14′) into equation (12) or (12′) we have the following: Φ_D0 as a function of Φ_D1

substituting equation (14) or (14′) into equation (13) or (13′) we have the following:

Φ_D0 as a function of Φ_D1

Similarly we have two equations two unknowns and solvable in Φ_D0 and Φ_D2 and similarly for Φ_D1 and Φ_D2.

Thus we have solved for Φ_D1, Φ_D2 and Φ_D3 twice.

From the second set of data a small frequency change from the first set of data this gives the same Φ_D1, Φ_D2 and Φ_D3 and different M₀, M₁ and M₂ and the ratios of the M′₀/M₀,M′₁/M₁ and M′₂/M₂ should give a good check of where the position Φ_D0 and Φ_D1 and Φ_D2 is detected at that RDB or RAB and a check on the solutions determined. From this the azimuth of the returns are determined.

A methodology in simplifying solutions where there are three or more returns detected in RDB or RAB. Determine from processing adjacent RDB or RAB detection of returns that will be detected in the processed bin. Then by employing candidate solution methodology, that is substituting all possible solutions which are very limited in number, and knowing one or more solutions. The solutions must have the known solution as one of its solutions. This makes it much simpler solution and more robust and accurate

Calculating from real time data a curve of the ratio of amplitude of returns at the two different apertures-vs-azimuth. Chose relatively significant clutter only detected in same bin in both apertures.

The methodology of producing this curve from real time data is as follows:

• • a) chose only relatively large amplitude clutter only data.
  • b) determine its azimuth from its measurement and its amplitude and phase of the return in each aperture.
  • c) measure ratio of ratio of the amplitude and phase from each aperture which is K_M0e^jα0=M′₀/M₀ and K_M1e^jΦα1=M′₁/M₁ etc.
  • d) perform this with as much data as to obtain a very good estimate of the curve by any fitting statistical methodology.
  • e) thus we have obtained a very good estimate of curve desired.

Now having generated from the real time data the curve of ratio of even divided by odd-vs-azimuth the candidate Φ technique is employed where all the possible phases are substituted (which are very limited in number, in equations and the solutions are compared to the curve for the ratio of aperture data as a function of azimuth and where there is a match this is the solution. The solution for equations should agree and they are a check on each other. Having attained the solution for azimuth, Φ_A, and having the solution for Φ_D then Φ=Φ_D−ΦA, the solution for radial velocity of the returns.

Determine its azimuth position and measure its ratio between apertures. Perform this for all clutter that meets the criteria and obtain a best statistical estimate of the said curve.

Dealing with three unknown returns or more an aid in finding the Φ_{D S} will be very helpful and obtain a more robust solutions. There are not many cases where there will be three or more returns are detected in the same RDB or RAB but in those cases the aid in processing first if there is clutter or other return detected close to processed bin to be detected in processed bin also. To continue to find as many possible solutions from all the close bins processed. If so you know at least one of the solutions assume you that returns solution is known and a great deal more robust and accurate.

Employing the possible candidate solutions, only a restricted number of solutions are possible, with the known solution or solutions, vary the other candidate solutions until the solution is

M′₀/M₀=K_M0e^jα0 or M′₁/M₁=K_M1e^jα1 or M′₂/M₂=K_M2e^jα2 etc.

The agreement with curve of ratio of returns ratio-vs-azimuth and the azimuth of each return has been determined. Since

Φ₀=Φ_D0−Φ_A0 and Φ₁=Φ_D1−Φ_A1 and Φ₂=Φ_D2−Φ_A2 etc. we solve for all returns radial velocity.

From the ratio of the second set of data to the first set of data we obtain XF0 and XF1 and XF2 from which is the ratio of

M′₀/M₀=X_F0 and M′₁/M₁=X_F1 and M′₂/M₂=X_F2 where all other terms are known. From the previous determinations of the Φ_D0 and Φ_D1 and Φ_D2 which is the position in the filter where the returns are detected at there peak in the RDB or RAB where Φ_0A is the phase of the return proportional to the azimuth of the return, and Φ_D0 is the phase of the return proportional to the peak of the return, Φ₀, is the phase of the return proportional to the velocity of the return.

In the one channel system:

A small change in doppler bin may be taken and its determined XF0, XF1 and XF2 which determines where the peak of the returns in doppler, this does not help in the evaluation in azimuth.

A small change in range bin may be taken and we determine XR0, XR1 and XR2 which determines where the peak of the returns in range, this does not help in the evaluation in azimuth. Resolving velocity ambiguity is taking a meaning delay in time and processing again. Thus we can determine the peak of each return in range and azimuth to obtain the maximum amplitude for each return for further use.

Other linear arrays are processed may be taken and its determine XH0, XH1 and XH2 which determines where the peak of the returns in height, this does not help in the evaluation in azimuth.

Resolving velocity ambiguity with the taking of meaningful delay in time and processing again. Thus we can determine the peak of each return in range and azimuth to obtain the maximum amplitude for each return for further use. The change in amplitude and phase of the range bin in conjunction with a delay in time gives an accurate determination of radial velocity which will resolve velocity ambiguity. This will also be for doppler change give the horizontal tangential velocity and with an analogous technique in a planar array giving vertical tangential velocity therefore an estimate of total velocity and estimate of pointing angle of the return. Which is the following: The previous analysis was with two returns possible per RAB processed though three (3) or more returns per RAB may be processed and determine the solution

From the analysis for the one channel system there is determined all the parameters of the returns the accurate position in range, azimuth and velocity and radial velocity, horizontal and vertical tangential velocity, mover, noise, and other returns may be detected and thresholded for importance and later processed to determine if sidelobes, multipath targets, etc.

One Pulse System

The analysis indicates the same solutions for range, velocity, azimuth and precise range and height are obtained but not the other parameters.

Obtaining curve of ratio of aperture 2 divided by aperture 1 M′/M=|K_M|e^jα-vs-azimuth as detailed in previous cases, we perform the candidate solution technique we substitute each candidate solution for right side of equations above and when the right solution is entered the left side will equal the correct Km and phase. This performed for all three equations and the solutions should correlate. Employ analogous techniques as in other three return cases in the one pulse technique to find all the same solutions.

RECAPPING—We have determined all Φs, in the two return and three return case and as consequently it may be performed for many Φs. The significance of this development is as each RDB that is processed clutter, target, noise, and other returns may be detected and thresholded for importance and later processed to determine if sidelobes, multipath targets, etc.

Employ analogous techniques as other three return and three sets of data cases to find all the same solutions.

VC—One Pulse—One Aperture System—ΔC Technique

The analysis may be performed with a one antenna transmit and many (channel) receive system. This system is called ΔC technique with one aperture formulated where the data will be delayed as desired to solve the problem. We will consider two returns clutter and mover and the channel data has been spectrum processed into its individual RABs and each will be treated as follows:

Referenced previous delayed one data point, each channel the data point is delayed it is multiplied by a suitable weighting function and its spectrum is obtained with such as FFT. In processing a particular RAB we may have the following:
V₀₀=M₀+M₁PULSE 1-APERTURE 1-CHANNEL 1-N  (1)
V₀₁=M₀e^jΦD0+M₁^jΦD1 PULSE 1-APERTURE 1-CHANNEL 2-N+1-DELAYED ONE CHANNEL  (2)
V₀₂=M₀e^j2ΦD0+M₁^j2ΦD1 PULSE 1-APERTURE 1-CHANNEL 3-N+2-DELAYED TWO CHANNELS  (3)
V₀₃=M₀e^j3ΦD0+M₁^j3ΦD1 PULSE 1-APERTURE 1 CHANNEL 4-N+3-DELAYED THREE CHANNELS  (4)
Above equations are for two returns where
V₀₀—is the return in the RAB being processed at delayed data 1 at PULSE 1 V₀₁—is the return in the RAB being processed at delayed data 2 at PULSE 1
V₀₂—is the return in the RAB being processed at delayed data 3 at PULSE 1
V₀₃—is the return in the RAB being processed at delayed data 4 at PULSE 1
M₀—is the first return vector
M₁—is the second return vector
Φ_D0 —is the phase of the first return where the phase is proportional to the azimuth plus velocity of the return
Φ_D1—is the phase of the second return where the phase is proportional to the azimuth plus velocity of the return

Taking equations (1) and (2) and treating M₀ and M₁ as the variables and solving for M₀ and M₁ we have:
M₀=(V₀₀e^jΦD1−V₀₁)/(e^jΦD1−e^jΦD0)  (5)
M₁=(V₀₀−e^jΦD0V₀₁)/(e^jΦD1−e^jΦD0)  (6)
Taking equations (2) and (3) and treating M₀e^jΦD0 and M₁e^jΦD1 as the variables and solving for M₀e^jΦD0 and M₁e^jΦD1 we have:
M₀e^jΦD0=(V₀₁e^jΦD1−V₀₂)/(e^jΦD1−e^jΦD0)  (5′)
M₁e^jΦD1=(V₀₁−e^jΦD0V₀₂)/(e^jΦD1−e^jΦD0)  (6′)
Equation (5′)/Equation (5) and (6′)/(6) are the following:
e^jΦD0=(V₀₁e^jΦD1−V₀₂)/V₀₀e^jΦD1−V₀₁)  (7′)
e^jΦD1=(V₀₁−e^jΦD0V₀₂)/(V₀₀−e^jΦD0V₀₁)  (8′)
Equation (7′) or Equation (8′) is easily solved for Φ_D0 and Φ_D1 which are proportional to the azimuth plus velocity of return 0 and return 1 respectively.
Now employing equations (1) and (2) and solving for M₀ and M₁ knowing Φ_D0 and Φ_D1 we are now to find Φ_A0 and Φ_A1 which are proportional to total phase of the azimuth return 0 and return 1 respectively and M₀ and M₁ and both returns are detected in the same RAB.
Now we have another set of equations for another close azimuth bin from the first azimuth bin processed.
V′₀₀=M′₀+M′₁ PULSE 1-AZIMUTH BIN 2-CHANNEL 1-N  (1′)
V′₀₁=M′₀e^jΦ′D0+M′₁^jΦ′D1 PULSE 1-AZIMUTH BIN 2-CHANNEL 2-N+1-DELAYED ONE CHANNEL  (2′)
V′₀₂=M′₀e^j2Φ′D0+M′₁j^2Φ′D1 PULSE 1-AZIMUTH BIN 2-CHANNEL 3-N+2-DELAYED TWO CHANNELS  (3′)
V′₀₃=M₀e^j3Φ′D0+M′₁^j3Φ′D1 PULSE 1-AZIMUTH BIN 2-CHANNEL 4-N+3-DELAYED THREE CHANNELS  (4′)
This gives the same Φ_A0 and Φ_A1 and different M₀ and M₁ and the ratios of the M₀ and M₁ should give a good estimate of where the position Φ_D0 and Φ_D1 is detected at in the RAB. Φ_D0 and Φ_D1 is the peak position of return 0 and 1 are detected in the RDB processed.

From the ratio of the second set of data to the first set of data we obtain |A_Z0|and |A_Z1|which is the ratio of M′₀/M₀=|A_Z0|and M′₁/M₁=|A_Z1|where all other terms are known.

A small change in the azimuth bin may be taken and we determine |A_Z0|and |A_Z1|which determines where the peak of the returns is in azimuth, this helps in the determination in tangential velocity. If attaining tangential velocity is desired then the taking of meaningful delay in time and processing again. Thus we can determine the peak of each return in azimuth to obtain the Φ_A0 and Φ_A1 of each return for further use in determining the radial velocity Φ₀=Φ_D0−Φ_A0 and Φ₁=Φ_D1−Φ_A1 change in amplitude and phase of the azimuth bin in conjunction with a delay in time gives an accurate determination of radial velocity. This will also be for and an analogous technique in a planar array giving vertical tangential velocity therefore an estimate of total velocity and estimate of pointing angle of the return which is the following: The previous analysis was with two returns possible per RAB processed though three (3) or more returns per RAB may be processed and determine the solution

From the previous analysis for the one pulse system there is determined all the parameters of the returns the accurate position in range, azimuth and height and ambiguous radial velocity.

VD—One Channel—Multiple Time Data-Change in Time-One Aperture

The system is a transmission from a fixed transmission array and one or more receive antennas This STAP methodology process the data into its frequency spectrum and consequently localize clutter into its own range doppler bin (RDB) together with any other returns such as target, thermal noise, and others.

The subsequent processing of each RDB will separate out clutter doppler since clutter has zero (0) radial velocity and other returns in the same (RDB) will have different velocities. From determining the velocity of the returns the azimuth will be calculated therefore no additional data is required.

The knowledge aided STAP will be involved to determine from the detected returns which are targets of interest, sidelobe returns, land sea interface, thermal noise, etc.

The knowledge aided STAP will be not involved in canceling clutter but in the post processing of the returns of interest so they may be detected and there parameters measured and determine the nature of the return.

Two returns in one RDB will be considered initially and such as three moving returns, or more will be dealt with as analogously as other techniques.
V₀₀=M₀+M₁ TIME DATA  (1)
V₀₁=M₀e^jΦD0+M₁^jΦD1 TIME DATA DELAYED ONE TIME  (2)
V₀₂=M₀e^j2ΦD0+M₁^j2ΦD1 DELAYED TWO TIMES  (3)
V₀₃=M₀e^j3ΦD0+M₁^j3ΦD1 DELAYED THREE TIMES  (4)
Above equations are for two returns where
V00—is the return in the RDB being processed at data 1
V₀₁—is the return in the RDB being processed at data 2
V₀₂—is the return in the RDB being processed at data 3
V₀₃—is the return in the RDB being processed at data 4
M₀—is the first return vector.
M₁—is the second return vector.
Φ_D0—is the phase of the first return where the phase is proportional to the azimuth plus velocity of the return.
Φ_D1—is the phase of the second return where the phase is proportional to the azimuth plus velocity of the return.

Taking equations (1) and (2) and treating M₀ and M₁ as the variables and solving for M₀ and M₁ we have:
M₀=(V₀₀e^jΦD1−V₀₁)/(e^jΦD1−e^jΦD0)  (5)
M₁=(V₀₀−e^jΦD0(V₀₁)/(e^jΦD1−e^jΦD0)  (6)
Taking equations (2′) and (3′) and treating M₀e^jΦD0 and M₁e^jΦD1 as the variables and solving for M₀ e^jΦD0 and M₁e^jΦD1 we have:
M₀e^jΦD0=(V₀₁e^jΦD1−V₀₂)/(e^jΦD1−e^jΦD0)  (5′)
M₁e^jΦD1=(V₀₁−e^jΦD0V₀₂)/(e^jΦD1−e^jΦD0)  (6′)
Equation (5′)/Equation (5) and (6′)/(6) are the following:
e^jθD0=(V₀₁e^jΦD1−V₀₂)/V₀₀e^jΦD1−V₀₁)  (7′)
e^jΦD1=(V₀₁−e^jΦD0V₀₂)/(V₀₀−e^jΦD0V₀₁)  (8′)

Equation (7′) or Equation (8′) is easily solved for Φ_D0 and Φ_D1 which are proportional to the azimuth plus velocity of return 0 and return 1 respectively. A close range is processed. This is processed to yield a second set of equations and processing similar to first set of data.
V′₀₀=M′₀+M′₁ NO DELAY IN DATA  (1′)
V′₀₁=M′₀e^jΦ0+M′₁^jΦ′D1 DELAYED ONE TIME  (2′)
V′₀₂=M′₀e^j2Φ′D0+M′₁^j2Φ′D1 DELAYED THREE TIMES  (3′)
V′₀₃=M′₀e^j3Φ′D0+M′₁^j3Φ′D1 DELAYED FOUR TIMES  (4′)

Now employing equations (1′) to (4′) and solving in same manner as equation (1) to (4) for M′₀, M′₁ and Φ′_D0, Φ′_D1 in which Φ′_D0=Φ_D0 and Φ′_D1=Φ_D1 we are now to find M′₀/M₀=X_RO and M′₁/M₁=X_RO which is the detected position of 0 and 1 return gives estimate of peak in the range bin, are proportional to range of return 0 and return 1 respectively and M₀ and M₁ and both returns are detected in the same range bin (RB).

To get a more accurate determination, another range or ranges may be processed or a slight change in the range processed and results correlated for best results. The ratio of second set of data for close in range we will obtain a much more accurate range. Now we have another set of equations after a significant delay in time and perform the same operations as first set of data and determine the amplitude and phase change of the determination of the range bin data will give an estimate of the unambiguous velocity of the returns and from this the known peak of the returns the azimuth of the returns are calculated.

VE—Combination of multi pulse techniques where processing significant time later (multi-aperture and ΔR techniques)

The two techniques may employ the same data since in multi aperture technique each aperture data the ΔR techniques may be implemented and may be processed in any manner to correlate with other solutions. The following is a list of common solutions.

1. Same solutions for all parameters such as the following:

• • a) Φ_{D s},Φ_{A s} and Φ_s and M s and X_{F s}, X_{R s} and X_{H s}
  • b) solutions for position and velocity of respective returns

2. Accuracy and robustness of solutions are enhanced

VF-Combination of multi channel technique (multi-aperture and ΔZ—change in azimuth-techniques)

The two techniques may employ the same data since in multi aperture technique each aperture data the ΔZ techniques may be implemented and may be processed in any manner to facilitate a solution. The following is a list of common solutions.

2. Same solutions for all parameters such as the following:

• • a) Φ_{D s}, Φ_{A s} and Φ_s and M s and X_{F s}, X_{R s} and X_{H s}
  • b) solutions for position and velocity of respective returns

3. Accuracy and robustness of solutions are enhanced

From the previous analysis for the one pulse system and that for the one channel system there is determined all the parameters of the returns the accurate position in range, azimuth and velocity and unambiguous radial velocity, horizontal and vertical tangential velocity, mover, noise, and other returns may be detected and thresholded for importance and later processed to determine if sidelobes, multipath targets, etc.

VG—Correlation Factors

1—Time delay as many pulses and process data and determines all Φs.

2—Additional time delay data processed again and all Φs should agree

3—Other RDB processed have same return data related to each such as mover should agree

4—Other techniques as to be shown later in document and results should agree

VH—The aforementioned system has many advantages but the outstanding are as the following

1—Channel matching is not a problem since only one channel is employed

2—Dwell Time is reduced dramatically due to the minimum of only one pulse required limited by the processor to perform the necessary functions.

3—Returns and clutter do not compete with each other in the resulting processing and therefore clutter and returns are thresholded separately and returns are ideally are competing with white noise only and make for an excellent return ratio to noise.

4—One pulse system is very simple system-less storage-less processing-more hardware due to RF front end and A/D converter required for every channel or groups of channels.

5—May be applied to one pulse or two or more pulses to give more than one solution and they should be very close to each other. Each processed as one pulse system and should have same results and correlated.

6—The significance of the ability to process many returns in same RDB and determining there amplitude and phases and radial velocity gives the ability to separate clutter and other returns such as bona fide movers, moving clutter, multi path returns, etc. Data base of returns and the mathematical basis and knowledge aided information would aid in categorizing these returns.

7—If more than one pulse implemented than the results of each pulse should be the same or close to the same. If a different transmission frequency is accompanied by another pulse they can be close in time and process as another one pulse system without interfering with each other.

8—If a number of pulses and a number of channels are employed with the delta T and delta C technique and correlated would make for excellent results.

9. FIG. 10 illustrates the detection of two returns in the same bin.

VI—A Two PULSES at a Time-“N”-Channels of Data-Two Returns-Φ_D Technique

The analysis may be performed with a one antenna transmit and two pulse system. This system is called ΔC methodology where the data will be delayed one or more channel increments in pulse 2 as required for a solution. We will consider two returns clutter and target and the “N” channel data (channel data) has been spectrum processed into its individual range azimuth bins (RABs) and each will be treated as follows: Each set of data, channel data point is delayed it is multiplied by a suitable weighting function and its spectrum is obtained with such as FFT. In processing a particular RAB where we have two returns we have the following equations:
V₀₀=M₀+M₁ PULSE 1 CHANNEL data 1-N Delay 0  (1)
V₀₁=M₀e^jΦD0+M₁e^jΦD1 PULSE 1 CHANNEL data 2-N+1 Delay 1  (2)
V₀₂=M₀e^j2ΦD0+M₁e^j2ΦD1 PULSE 1 CHANNEL data 3-N+2 Delay 2  (3)
V₀₃=M₀e^j3ΦD0+M₁e^j3ΦD1 PULSE 1 CHANNEL data 4-N+3 Delay 3  (4)
Above equations are for two returns where
V₀₀—is the return in the RAB being processed at time 1
V₀₁—is the return in the RAB being processed at time 2
V₀₂—is the return in the RAB being processed at time 3
V₀₃—is the return in the RAB being processed at time 4
M₀—is the first return vector
M₁—is the second return vector
Φ_D0—is the phase of the first return where the phase is proportional to the phase due to radial velocity plus that due to the azimuth of the return.
Φ_D1—is the phase of the second return where the phase is proportional to the radial velocity plus that due to the azimuth of the return

It is noted with each delay in channel data the vectors of the returns phase is increased proportional to the delay which represents the phase of the return proportional to the velocity plus that due to its azimuth in the antenna beam. Zero velocity returns such as clutter will have phase shift equal to zero due to its velocity but one due to its azimuth position in the antenna beam and other returns will have phase shifts directly proportional to their radial velocity and one due to its azimuth position in the main beam. When returns are detected in the same RAB the sum of their phases (Φ_D0 or Φ_D1) (frequency) are in the same RAB and it is on this basis the returns are analyzed, processed and separated out.

Taking equations (1) and (2) and treating M₀ and M₁ as the variables and solving for M₀ and M₁ we have:
M₀=(V₀₀e^jΦD1−V₀₁)/(e^jΦD1−e^jΦD0)  (1′)
M₁=(V₀₀−V₀₁e^jΦD0)/(e^jΦD1−e^jΦD0)  (2′)
Taking equations (2) and (3) and treating M₀e^jΦD0 and M₁e^jΦD1 as the variables and solving for M₀e^jΦD0 and M₁e^jΦD1 we have:
M₀e^jΦD0=(V₀₁e^jΦD1−V₀₂)/(e^jΦD1−e^jθD0)  (1″)
M₁e^jΦD1=(V₀₁−V₀₂e^jΦD0)/(e^jΦD1−e^jΦD0)  (2″)
Equation (2″)/Equation (2′) or Equation (1″)/Equation (1′) are the following:
e^jΦD0=(V₀₁e^jΦD1−V₀₂)/(V₀₀e^jΦD1−V₀₁)  (3′)
e^jΦD1=(V₀₂−V₀₁e^jΦD0)/(V₀₁−V₀₀e^jΦD0)  (4′)
Equation (3′) or Equation (4′) is easily solved for Φ_D0 and Φ_D1 which are proportional to the total phase of return 0 and return 1 respectively. If return “Mo” is clutter then Φ₀=0 corresponding to clutter having zero (0) velocity. Now employing equations (1) and (2) and solving for M₀ and M₁ knowing Φ_D0 and Φ_D1 we are now to find Φ₀ and Φ₁ which are proportional to velocity of return 0 and return 1 respectively. M₀ and M₁ and both returns that are detected in the same RAB.

Having the second channel data and performing the same operations as in channel 1 and the equations are as follows:
V′₀₀=M₀X₀₁+M₁X₁₁ PULSE 2 CHANNEL data 1-N Delay 0  (1′)
V′₀₁=M₀X₀₁e^jΦ′D0+M₁X₁₁e^jΦ′D1 PULSE 2 CHANNEL data 2-N+1 Delay 1  (2′)
V′₀₂=M₀X₀₁e^j2Φ′D0+M₁X₁₁e^j2Φ′D1 PULSE 2 CHANNEL data 3-N+2 Delay 2  (3′)
V′₀₃=M₀X₀₁e^j3Φ′D0+M₁X₁₁e^j3Φ′D1 PULSE 2 CHANNEL data 4-N+3 Delay 3  (4′)
X₀₁=e^jDΦA0e^jKD0X0/W_M0=|1/A_M0|e^{j(KD0X0−ΨM0)} where D=0X₀₁=e^jKD0X0/W_M0=|1/A_M0|e^{j(KD0X0−ΨM0)}
X₁₁=e^jDΦA1e^jKD0X1/W_M1=|1/A_M1|e^{j(KD0X1−ΨM1)} where D=0X₁₁=e^jKD0X1/W_M1=|1/A_M1|e^{j(KD0X1−ΨM1)}

Solving equations (1′) to (4′) in the same manner as equations (1) to (4) we solve for Φ′_D0 and Φ′^D1 and M₀X₀₁ and M₁X₁₁. Φ′_D0 and Φ′^D1 Solution should be the same as for Φ_D0 and Φ_D1 since in pulse 2 we have the same returns detected in the same RAB with the same velocity components.

solution for M₀X₀₁ and M₁X₁₁ in pulse 2/solving for M₀ and M₁ in pulse 1 yields X₀₁ and X₁₁

Having the second pulse data delayed and performing the same operations as in channel 1 and 2 and the equations are as follows:
V″₀₀=M₀X₀₂+M₁X₁₂ PULSE 2 CHANNEL data 2-N+1 Delay 1  (1″)
V″₀₁=M₀X₀₂e^jΦ″D0+M₁X₁₂e^jΦ″D1 PULSE 2 CHANNEL data 3-N+2 Delay 2  (2″)
V″₀₂=M₀X₀₂e^j2Φ″D0+M₁X₁₂e^j2Φ″D1 PULSE 2 CHANNEL data 4-N+3 Delay 3  (3″)
V″₀₃=M₀X₀₂e^j3Φ″D0+M₁X₁₂e^j3Φ″D1 PULSE 2 CHANNEL data 5-N+4 Delay 4  (4″)
X₀₂=X₀₁e^{j(ΦA0+ΔKD0X0)}=X₀₁e^j(γ0): X₁₂=X₁₁e^{j(ΦA1+ΔKD0X1)}=X₁₁e^j(γ1)

Rewriting equations (1″) to (4″) we have the following:
V″₀₀=M₀X₀₁e^jγ0+M₁X₁₂e^jγ1 PULSE 2 CHANNEL data 2-N+1 Delay 1  (1″)
V″₀₁=M₀X₀₁e^jγ0e^jΦ″D0+M₁X₁₂e^jγ1e^jΦ″D1 PULSE 2 CHANNEL data 3-N+2 Delay 2  (2″)
V″₀₂=M₀X₀₁e^jγ0e^j2Φ″D0+M₁X₁₂e^jγ1e^j2Φ′D1 PULSE 2 CHANNEL data 4-N+3 Delay 3  (3″)
V″₀₃=M₀X₀₁e^jγ0e^j3Φ″D0+M₁X₁₂e^jγ1e^j3Φ″D1 PULSE 2 CHANNEL data 5-N+4 Delay 4  (4″)
Solving equations (1″) to (4″) the same manner as equations (1) to (4) we solve for Φ′_D0 and Φ′_D1 and M₀X₀₁ and M₁X₁₁. Φ″_D0 and Φ″_D1 Solution should be the same.
solution for M₀X₀₁e^jγ0 and M₁X₁₁e^jγ1 in channel 2 delayed/solving for M₀X₀₁ and M₁X₁₁ in channel 2 yields e^jγ0 and e^jγ1 and γ₀=Φ_A0+ΔK_D0X₀ where X₀ is unknown and γ₀ is known and therefore Φ_A0 has to be determined
γ₁=Φ_A1+ΔK_D0X₁ where X₁ is unknown and γ₁ are known and therefore Φ_A1 has to be determined₁
Solving for phase proportional to velocity in both returns we have the following:

Finding Φ_D0 and Φ_D1 as in the channel technique and finding

Φ_O=Φ_D0−Φ_A0 and Φ₁=Φ_D1−Φ_A1

Definition of terms not defined previously:

ALL “V” TERMS ARE MEASURED TERMS.—

X01—CHANNEL 2 TERM THAT makes relates channel 2 to channel 1 for return 1

X02—CHANNEL 2 TERM THAT makes relates channel 2 to channel 1 for return 2

ΔK_D0—The difference factor for different delays for return 1 and 2

X₀—The azimuth position in filter factor for return 1 γ₀—The difference in angle between different delayed data of return 1

γ₁—The difference in angle between different delayed data of return 2

Φ_AO—phase proportional to azimuth of return 1

Φ_A1—phase proportional to azimuth of return 2

A_M0—Amplitude balancing term between channel 1 and 2 for return 1

A_M1—Amplitude balancing term between channel 1 and 2 for return 2

Ψ_M0—Phase balancing term between channel 1 and 2 for return 1

Ψ_M1—Phase balancing term between channel 1 and 2 for return 2

Comments and observations on technique

1—All solutions

Φ_D0, Φ′_D0, Φ″_D0 should be equal and Φ_D1, Φ′_D1, Φ″_D1 should be equal

2—Solving for M0 and M1 by this approach solves for the location of their peaks therefore they have a phase shift equal to zero at this point.

3—To solve for the channel balancing terms three sets of equations are required but for solving for velocity and azimuth only last two sets are required.

4—Correlate with other ΔC technique in the following manner:

• • a) Same solution
  • b) All variables are the same value such as M0, M1, ETC
  • c) ΔF, ΔR and ΔH Results should correlate
    Analogously a small change in range bin may be taken and we determine X_R0 and X_R1 which determines where the peak of the returns in range, this does not help in the evaluation in azimuth. Resolving velocity ambiguity with the taking of meaningful delay in time and processing again. Thus we can determine the peak of each return in range and azimuth to obtain the maximum amplitude for each return for further use. The change in amplitude and phase of the range bin in conjunction with a delay in time gives an accurate determination of velocity which will resolve velocity ambiguity.

The change in amplitude and phase of the doppler bin in conjunction with a delay in time gives an accurate determination of horizontal tangential velocity.

The change in amplitude and phase in the different linear arrays of the doppler bin in conjunction with a delay in time gives an accurate determination of vertical tangential velocity.

Thus we have obtained the three dimensional position and velocity of all returns.

B. Two pulse “N” channel data in time-three returns-Φ_D technique

The previous analysis was for two returns possible per RAB processed; this will be for three (3) returns per RAB.
V₀₀=M₀+M₁+M₂ PULSE 1 CHANNEL data 1-N Delay 0  (1)
V₀₁=M₀e^jΦD0+M₁e^jΦD1+M₂e^jΦD2 PULSE 1 CHANNEL data 2-N+1 Delay 1  (2)
V₀₂=M₀e^j2ΦD0+M₁e^j2ΦD1+M₂e^j2ΦD2 PULSE 1 CHANNEL data 3-N+2 Delay 2  (3)
V₀₃=M₀e^j3ΦD0+M₁e^j3ΦD1+M₂e^j3ΦD2PULSE 1 CHANNEL data 4-N+3 Delay 3  (4)
All terms previously defined except the following:
M—Third return
Φ_D2—phase proportional to radial velocity plus azimuth of third return
Φ_A2—phase proportional to azimuth of third return
Φ₂—phase proportional to radial velocity of third return
Taking equations (1) and (2) and (3) and treating M₀ and M₁ and M₂ as the variables and solving the determinant equation for Δ₀ we have: Δ 0 = ⅇ jΦ D ⁢   ⁢ 0 1 ⁢ ⅇ jΦ D ⁢   ⁢ 1 1 ⁢ ⅇ jΦ D ⁢   ⁢ 2 1 = 1 ⁢ ⁢ ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⅇ jΦ D ⁢   ⁢ 2 - 1 ⁢ ⁢ ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ jΦ D ⁢   ⁢ 2 + 1 ⁢ ⁢ ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⁢ Δ 0 = ⁢ ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⅇ jΦ D ⁢   ⁢ 2 - ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 + ⁢ ⅇ jΦ D ⁢   ⁢ 2 ⁢ ⅇ j2Φ D ⁢   ⁢ 0 + ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 - ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 0 = ⁢ function ⁢   ⁢ of ⁢   ⁢ Φ D ⁢   ⁢ 0 , Φ D ⁢   ⁢ 1 ⁢   ⁢ and ⁢   ⁢ Φ D ⁢   ⁢ 2 ⁢ ⁢ and ⁢   ⁢ solving ⁢   ⁢ for ⁢ ⁢ M 0 _ = M 0 * Δ 0   M 0 _ = V 01 V 00 ⁢ ⅇ jΦ D ⁢   ⁢ 1 1 ⁢ ⅇ jΦ D ⁢   ⁢ 2 1 = V 00 ⁢ ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⅇ jΦ D ⁢   ⁢ 2 - V 01 ⁢ 1 ⁢ ⁢ 1 + V 02 ⁢ 1 ⁢ ⁢ 1 ⁢ ⁢ V 02 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ V ⁢   ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⁢ ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⅇ jΦ D ⁢   ⁢ 2 ⁢ ⁢ M 0 _ = ⁢ V 00 ⁡ ( ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ jΦ D ⁢   ⁢ 2 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ) - ⁢ V 01 ( ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ j2Φ D ⁢   ⁢ 1 ) + V 02 ⁡ ( ⅇ jΦ D ⁢   ⁢ 2 - ⅇ jΦ D ⁢   ⁢ 1 ) = ⁢ function ⁢   ⁢ of ⁢   ⁢   ⁢ Φ D ⁢   ⁢ 1 ⁢   ⁢ and ⁢   ⁢ Φ D ⁢   ⁢ 2 ⁢ ⁢ and ⁢   ⁢ solving ⁢   ⁢ for ⁢ ⁢ M 1 _ = M 1 * Δ 0 ( 6 ) M 1 _ = ⅇ jΦ D ⁢   ⁢ 0 1 ⁢ V 01 V 00 ⁢ ⅇ jΦ D ⁢   ⁢ 2 1 = V 00 ⁢ ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ jΦ D ⁢   ⁢ 2 - V 01 ⁢ 1 ⁢ ⁢ 1 + V 02 ⁢ 1 ⁢ ⁢ 1 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ V 02 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ⁢ ⁢ ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ jΦ D ⁢   ⁢ 2 ⁢ ⁢ M 1 _ = ⁢ V 01 ⁡ ( ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ jΦ D ⁢   ⁢ 2 ) - ⁢ V 01 ( ⅇ j2Φ D ⁢   ⁢ 2 - ⅇ j2Φ D ⁢   ⁢ 0 ) + V 02 ⁡ ( ⅇ jΦ D ⁢   ⁢ 2 - ⅇ jΦ D ⁢   ⁢ 0 ) = ⁢ function ⁢   ⁢ of ⁢   ⁢ Φ D ⁢   ⁢ 0 ⁢   ⁢ and ⁢   ⁢   ⁢ Φ D ⁢   ⁢ 2 ⁢ ⁢ and ⁢   ⁢ solving ⁢   ⁢ for ⁢ ⁢ M 2 _ = M 2 * Δ 0 ( 7 ) M 2 _ = ⅇ jΦ D ⁢   ⁢ 0 1 ⁢ ⅇ jΦ D ⁢   ⁢ 1 1 ⁢ V 01 V 00 = V 00 ⁢ ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ jΦ D ⁢   ⁢ 1 - V 01 ⁢ 1 ⁢ ⁢ 1 + V 02 ⁢ 1 ⁢ ⁢ 1 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ V 02 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⁢ ⅇ j2Φ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⁢ ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ jΦ D ⁢   ⁢ 1 ⁢ ⁢ M 2 _ = ⁢ V 00 ⁡ ( ⅇ jΦ D ⁢   ⁢ 0 ⁢ ⅇ j2Φ D ⁢   ⁢ 1 - ⅇ j2Φ D ⁢   ⁢ 1 ⁢ ⅇ jΦ D ⁢   ⁢ 0 ) - ⁢ V 01 ( ⅇ j2Φ D ⁢   ⁢ 1 - ⅇ j2Φ D ⁢   ⁢ 0 ) + V 02 ⁡ ( ⅇ jΦ D ⁢   ⁢ 1 - ⅇ jΦ D ⁢   ⁢ 0 ) = ⁢ function ⁢   ⁢ of ⁢   ⁢ Φ D ⁢   ⁢ 0 ⁢   ⁢ and ⁢   ⁢   ⁢ Φ D ⁢   ⁢ 1 ( 8 )
Solving for Φ_D0, Φ_D1 and Φ_D2 and substituting these values in equations (1), (2) and (3) we determine M₀, M₁ and M₂.
Taking equations (2), (3) and (4) and treating M₀e^jDΦ0, M₁e^jDΦ1 and M₂e^jDΦ2 as the variables and solving the determinant equation for Δ₀ is the same and performing the same operations as with the first set of equations we have the following:
M₀e^jΦD0=function of (Φ_D1,Φ_D2)  ( 6)
M₁e^jΦD1=function of (Φ_D0,Φ_D2)  ( 7)
M2e^jΦD2=function of (Φ_D0,Φ_D1)  ( 8)
Equation ( 6)/(6), ( 7)/(7) and ( 8)/(8)
( 6)/(6)=e^jΦD0=function (Φ_D1,Φ_D2)
( 7)/(7)=e^jΦD1=function (Φ₀,Φ_D2)
( 8)/(8)=e^jΦD2=function (Φ_D0,Φ_D1)
Solving for Φ_D0,Φ_D1 and Φ_D2 and substituting these values in equations (2), (3) and (4) we determine M₀e^jDΦ0, M₁e^jDΦ1 and M₂e^jDΦ2.
V′₀₀=M₀X₀₁+M₁X₁₁+M₂X₂₁ PULSE 2 CHANNEL data 1-N Delay 0  (1)
V′₀₁=M₀X₀₁e^jΦ′D0+M₁X₁₁e^jΦ′D1+M₂X₂₁e^jΦ′D2 PULSE 2 CHANNEL data 2-N+1 Delay 1  (2)
V′₀₂=M₀X₀₁e^j2Φ′D0+M₁X₁₁e^j2Φ′D1+M₂X₂₁e^j2Φ′D2 PULSE 2 CHANNEL data 3-N+2 Delay 2  (3′)
V′₀₃=M₀X₀₁e^j3Φ′D0+M₁X₁₁e^j3Φ′D1+M₂X₂₁e^j3Φ′D2 PULSE 2 CHANNEL data 4-N+3 Delay 3  (4)
X₀₁=e^jDΦA0e^jKD0X0/W_M0=|1/A_M0|e^{j(KD0X0−ΨM0)} where D=0X₀₁=e^jKD0X0/W_M0=|1/A_M0|e^{j(KD0X0−ΨM0)}
X₁₁=e^jDΦA1e^jKD0X1/W_M1=|1/A_M1|e^{j(KD0X1−ΨM1)} where D=0X₁₁=e^jKD0X1/W_M1=|1/A_M1|e^{j(KD0X1−ΨM1)}
X₂₁=e^jDΦA2e^jKD0X2/W_M2=|1/A_M2|e^{j(KD0X2−ΨM2)} where D=0X₂₁=e^jKD0X2/W_M2=|1/A_M2|e^{j(KD0X2−ΨM2)}

Performing the same operations on equations (1′), (2′) and (3′) with the variables M₀X₀₁, M₁X₁₁ and M₂X₂₁ and Δ₀ remains the same. The analysis is analogous and the result is the following: and ⁢   ⁢ solving ⁢   ⁢ for ⁢   ⁢ M 0 ⁢ X 01 _ = M 0 ⁢ X 01 * Δ 0   M 0 ⁢ X 01 _ = ⁢ V 01 ⁡ ( ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 1 ′ ⁢ ⅇ j ⁢   ⁢ 2 ⁢   ⁢ Φ ⁢   D ⁢   ⁢ 2 ′ - ⅇ j ⁢   ⁢ 2 ⁢   ⁢ Φ D ⁢   ⁢ 1 ′ ⁢ ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 2 ′ ) - ⁢ V 02 ⁡ ( ⅇ j ⁢   ⁢ 2 ′ ⁢ Φ D ⁢   ⁢ 1 - ⅇ j ⁢   ⁢ 2 ⁢ Φ D ⁢   ⁢ 2 ′ ) + V 03 ⁡ ( ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 1 ′ - ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 2 ′ ) = ⁢ function ⁢   ⁢ of ⁢   ⁢ Φ D ⁢   ⁢ 1 ′ ⁢   ⁢ and ⁢   ⁢ Φ D ⁢   ⁢ 2 ′ ⁢ ⁢ and ⁢   ⁢ solving ⁢   ⁢ for ⁢   ⁢ M 1 ⁢ X 11 _ = M 1 ⁢ X 11 * Δ 0 ( 6 ′ ) V 01 ⁡ ( ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 0 ′ ⁢ ⅇ j ⁢   ⁢ 2 ⁢   ⁢ Φ ⁢   D ⁢   ⁢ 2 ′ - ⅇ j ⁢   ⁢ 2 ⁢   ⁢ Φ D ⁢   ⁢ 0 ′ ⁢ ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 2 ′ ) - V 02 ⁡ ( ⅇ j ⁢   ⁢ 2 ′ ⁢ Φ D ⁢   ⁢ 0 - ⅇ j ⁢   ⁢ 2 ⁢ Φ D ⁢   ⁢ 2 ′ ) + V 03 ⁡ ( ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 0 ′ - ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 2 ′ ) = function ⁢   ⁢ of ⁢   ⁢ Φ D ⁢   ⁢ 0 ′ ⁢   ⁢ and ⁢   ⁢ Φ D ⁢   ⁢ 2 ′ ⁢ ⁢ and ⁢   ⁢ solving ⁢   ⁢ for ⁢   ⁢ M 2 ⁢ X 21 _ = M 2 ⁢ X 21 * Δ 0 ( 7 ′ ) M 2 ⁢ X 21 _ = V 01 ⁡ ( ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 0 ′ ⁢ ⅇ j ⁢   ⁢ 2 ⁢   ⁢ Φ ⁢   D ⁢   ⁢ 1 ′ - ⅇ j ⁢   ⁢ 2 ⁢   ⁢ Φ D ⁢   ⁢ 0 ′ ⁢ ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 1 ′ ) - V 02 ⁡ ( ⅇ j ⁢   ⁢ 2 ′ ⁢ Φ D ⁢   ⁢ 1 - ⅇ j ⁢   ⁢ 2 ⁢ Φ D ⁢   ⁢ 0 ′ ) + V 03 ⁡ ( ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 1 ′ - ⅇ j ⁢   ⁢ Φ D ⁢   ⁢ 0 ′ ) = function ⁢   ⁢ of ⁢   ⁢ Φ D ⁢   ⁢ 0 ′ ⁢   ⁢ and ⁢   ⁢ Φ D ⁢   ⁢ 1 ′ ( 8 ′ )
Taking equation (6′)/(6) we have the following:
X₀₁=M₀X₀₁/M₀

Taking equation (7′)/(7) we have the following:
X₁₁=M₁X₁₁/M₁
Taking equation (8′)/(8) we have the following:
X₂₁=M₂X₂₁/M₂
Performing the same operations on equations (2′), (3′) and (4′) with the variables M₀X₀₁, M₁X₁₁ and M₂X₂₁ and Δ₀ remains the same. The analysis is analogous and the result is the following:
M₀e^jΦD0=function of (Φ_D1,Φ_D2)  ( 6′)
M₁e^jΦD1=function of (Φ_D0,Φ_D2)  ( 7)
M₂e^jΦD2=function of (Φ_D0,Φ_D1)  ( 8′)
Equation ( 6)/(6′), ( 7)/(7′) and ( 8′)/(8′)
( 6′)/(6′)=e^jΦD0=function (Φ_D1,Φ_D2)
( 7′)/(7)=e^jΦD1=function (Φ_D0,Φ_D2)
( 8′)/(8′)=e^jΦD2=function (Φ_D0,Φ_D1)
Solving for Φ_D0, Φ_D1 and Φ_D2 and substituting these values in equations (2), (3) and (4) we determine M₀X₀₁, M₁X₀₂ and M₂X₀₃.

Taking the next set of equations as follows:
V″₀₀=M₀X₀₂+M₁X₁₂+M₂X₂₂ PULSE 2 CHANNEL data 2-N+1 Delay 1  (1″)
V″₀₁=M₀X₀₂e^jΦ″D0+M₁X₁₂e^jΦ″D1+M₂X₂₂e^jΦ″D2 PULSE 2CHANNEL data 3-N+2 Delay 2  (2″)
V″₀₂=M₀X₀₂e^j2Φ″D0+M₁X₁₂e^j2Φ″D1+M₂X₂₂e^j2Φ″D2 PULSE 2CHANNEL data 4-N+3 Delay 3  (3″)
V″₀₃=M₀X₀₂e^j3Φ″D0+M₁X₁₂e^j3Φ″D1+M₂X₂₂e^j3Φ″D2 PULSE 2CHANNEL data 5-N+4 Delay 4  (4″)
X₀₂=X₀₁e^{j(ΦA0+ΔKD0X0)}=X₀₁e^j(γ0):X₁₂=X₁₁e^{j(ΦA1+ΔKD0X0)}=X₁₁e^j(γ1):X₂₂=X₂₁e^{j(ΦA2+ΔKD0X0)}=X₂₁e^j(γ0)
Rewriting equations (1″ to 4″) we have the following:
V″₀₀=M₀X₀₁e^jγ0+M₁X₁₁e^jγ1+M₂X₁₂e^jγ2 PULSE 2 CHANNEL data 2-N+1 Delay 1  (1″)
V″₀₁=M₀X₀₁e^jγ0e^jΦ″D0+M₁X₁₁e^jγ1e^jΦ″D1+M₂X₁₂e^jγ2e^jΦ″D2 PULSE 2 CHANNEL data 3-N+2 Delay 2  (2″)
V″₀₂=M₀X₀₁e^jγ0e^j2Φ″D0+M₁X₁₁e^jγ1e^j2Φ″D1+M₂X₁₂e^jγ2e^j2Φ″D2 PULSE 2 CHANNEL data 4-N+3 Delay 3  (3″)
V″₀₃=M₀X₀₁e^jγ0e^j3Φ″D0+M₁X₁₁e^jγ1e^j3Φ″D1+M₂X₁₂e^jγ2e^j3Φ″D2 PULSE 2 CHANNEL data 5-N+4 Delay 4  (4″)
Performing the same operations on equations (1″), (2″) and (3″) with the variables M₀X₀₁e^jγ0, M₁X₁₁e^jγ1 and M₂X₂₁e^jγ2 and Δ₀ remains the same. The analysis is analogous and the result is the following: and ⁢   ⁢ solving ⁢   ⁢ for ⁢ ⁢   ⁢ M 0 ⁢ X 01 _ ⁢ ⅇ jΥ 0 _ = M 0 ⁢ X 01 ⁢ ⅇ jΥ 0 * Δ 0 ⁢ ⁢ M 0 ⁢ X 01 _ ⁢ ⅇ jΥ 0 _ = ⁢ V 02 ″ ⁡ ( ⅇ jΦ D ⁢   ⁢ 1 ′ ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ′ - ⅇ j2Φ D ⁢   ⁢ 1 ′ ⁢ ⅇ jΦ D ⁢   ⁢ 2 ′ ) - ⁢ V 03 ″ ⁡ ( ⅇ j2 ′ ⁢ Φ D ⁢   ⁢ 1 - ⅇ j2Φ D ⁢   ⁢ 2 ′ ) + V 04 ″ ⁡ ( ⅇ jΦ D ⁢   ⁢ 1 ′ - ⅇ jΦ D ⁢   ⁢ 2 ′ ) = ⁢ function ⁢   ⁢ of ⁢   ⁢ Φ D ⁢   ⁢ 1 ″ ⁢   ⁢ and ⁢   ⁢ Φ D ⁢   ⁢ 2 ″ ⁢ ⁢ and ⁢   ⁢ solving ⁢   ⁢ for ⁢ ⁢ M 1 ⁢ X 11 _ ⁢ ⅇ jΥ 1 _ = M 1 ⁢ X 11 ⁢ ⅇ jΥ 1 * Δ 0 ( 6 ″ ) M 1 ⁢ X 11 _ ⁢ ⅇ jΥ 1 _ = ⁢ V 02 ⁡ ( ⅇ jΦ D ⁢   ⁢ 0 ″ ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ″ - ⅇ j2Φ D ⁢   ⁢ 0 ″ ⁢ ⅇ j2Φ D ⁢   ⁢ 2 ″ ) - ⁢ V 03 ⁡ ( ⅇ j2 ″ ⁢ Φ D ⁢   ⁢ 0 - ⅇ j2Φ D ⁢   ⁢ 2 ″ ) + V 04 ⁡ ( ⅇ jΦ D ⁢   ⁢ 0 ″ - ⅇ jΦ D ⁢   ⁢ 2 ″ ) = ⁢ function ⁢   ⁢ of ⁢   ⁢ Φ D ⁢   ⁢ 0 ′ ⁢   ⁢ and ⁢   ⁢ Φ D ⁢   ⁢ 2 ′ ⁢ ⁢ and ⁢   ⁢ solving ⁢   ⁢ for ⁢ ⁢ M 2 ⁢ X 21 _ ⁢ ⅇ jΥ 2 _ = M 2 ⁢ X 21 ⁢ ⅇ jΥ 2 * Δ 0 ( 7 ″ ) M 2 ⁢ X _ 21 ⁢ ⅇ jΥ 2 _ = ⁢ V 02 ( ⅇ jΦ D ⁢   ⁢ 0 ″ ⁢ ⅇ j2Φ D ⁢   ⁢ 1 ″ - ⅇ j2Φ D ⁢   ⁢ 0 ″ ⁢ ⅇ jΦ D ⁢   ⁢ 1 ″ ) - ⁢ V 03 ( ⅇ j2 ′ ⁢ Φ D ⁢   ⁢ 1 ″ - ⅇ j2Φ D ⁢   ⁢ 0 ″ ) + V 04 ⁡ ( ⅇ jΦ D ⁢   ⁢ 1 ″ - ⅇ jΦ D ⁢   ⁢ 0 ″ ) = ⁢ function ⁢   ⁢ of ⁢   ⁢ Φ D ⁢   ⁢ 0 ″ ⁢   ⁢ and ⁢   ⁢ Φ D ⁢   ⁢ 1 ″ ( 8 ″ )
Taking equation (6″)/(6′) we have the following:
e^jγ0=M₀X₀₁e^jγ0/M₀X₀₁=V″₀₂(e^jΦ′D1e^j2Φ′D2−e^j2Φ′D1e^jΦ′D2)−V″₀₃(e^j2′ΦD1−e^j2Φ′D2)+V″₀₄(e^jΦ′D1−jΦ′^D2)/V₀₁(e^jΦ′D1e^j2Φ′D2−e^j2Φ′D1e^jΦ′D2)−V₀₂(e^j2′ΦD1−e^j2Φ′D2)+V₀₃(e^jΦ′D1−e^jΦ′D2)  (12)
Taking equation (7″)/(7′) we have the following:
e^jγ1=M₁X₁₁e^jγ1/M₁X₁₁=V₀₂(e^jΦ″D0e^j2Φ″D2−e^j2Φ″D0e^jΦ″D2)−V₀₃(e^j2″ΦD0−e^j2Φ″D2)+V₀₄(e^jΦ″D0−jΦ″^D2)/V₀₁(e^jΦ′D0e^j2Φ′D2−e^j2Φ′D0e^jΦ′D2)−V₀₂(e^j2′ΦD0−e^j2Φ′D2)+V₀₃(e^jΦ′D0−e^jΦ′D2)  (13)
Taking equation (8″)/(8′) we have the following:
e^jγ2=M₂X₂₁e^jγ2/M₂X₂₁=V₀₂(e^jΦ″D0e^j2Φ″D1−e^j2Φ″D0e^jΦ″D1)−V₀₃(e^j2″ΦD1−e^j2Φ″D0)+V₀₄(e^jΦ″D1−jΦ″^D0)/V₀₁(e^jΦ′D0e^j2Φ′D2−e^j2Φ′D0e^jΦ′D2)−V₀₂(e^j2′ΦD0−e^j2Φ′D2)+V₀₃(e^jΦ′D0−e^jΦ′D2)  (14)
Performing the same operations on equations (2″), (3″) and (4″) with the variables M₀X₀₁e^jθ0, M₁X₁₁e^jθ1 and M₁X₂₁e^jθ2 and Δ₀ remains the same
Substituting equation (13) into equation (12) we have the following:
The analysis is analogous and the result is the following:
M₀e^jΦD0=function of (Φ_D1,Φ_D2)  ( 6′)
M₁e^jΦD1=function of (Φ_D0,Φ_D2)  ( 7′)
M₂e^jΦD2=function of (Φ₀,Φ_D1)  ( 8″)
Equation ( 6′)/(6″), ( 7′)/(7″) and ( 8′)/(8−)
( 6″)/(6″)=e^jΦD0=function (Φ_D1,Φ_D2)
( 7″)/(7″)=e^jΦD1=function (Φ_D0,Φ_D2)
( 8″)/(8″)=e^jΦD2=function (Φ_D0,Φ_D1)
Solving for Φ_D0, Φ_D1 and Φ_D2 and substituting these values in equations (2), (3) and (4) we determine M₀X₀₀e^jγ0, M₁X₀₁e^jγ1 and M₂X₀₂e^jγ2 and we have determined γ₁, γ₂ and γ₃ and γ₀=Φ_A0+ΔK_D0X₀:γ₁=Φ_A1+ΔK_D0X₁:γ₂=Φ_A2+ΔK_D0X₂ where everything is known except Φ_A0, Φ_A1 and Φ_A2 are easily determined and having to determine Φ_D0 and Φ_D1 and Φ_D2 as in the two return case since Φ₀=Φ_D0−Φ_A0: Φ₁=Φ_D1−Φ_A1:Φ₂=Φ_D2−Φ_A2 the velocity of the returns have been attained.
e^jγ0=M₀X₀₀e^jγ0/M₀X₀₀
e^jγ1=M₁X₀₁e^jγ1/M₁X₀₁
e^jγ2=M₂X₀₂e^jγ2/M₂X₀₂
Substituting equation (13) into equation (12) we have the following:
Same solution as in two channel section BI for Φ₀, Φ₁ and Φ₂ and for the Assist in Determining Solutions when the Number of Returns are Three or More
M₀, M₁ and M₂ and the ratios of the M′₀/M₀,M′₁/M₁ and M′₂/M₂ should give a good estimate of where the position Φ_D0 and Φ_D1 and Φ_D2 is detected at that RAB. From this an estimate of velocity of the returns are determined.
From the ratio of the second set of data to the first set of data we obtain XF0 and XF1 and XF2 from which is the ratio of M′₀/M₀=X_F0 and M′₁/M₁=X_F1 and M′₂/M₂=X_F2 where all other terms are known. From the previous determinations of the estimate Φ_D0 and Φ_D1 and Φ_D2 which is the position in the filter where the returns are detected at there peak in the RAB. This gives a good estimate of the azimuth of the returns. To get more accurate determinations other close frequency points to initially processed data are processed.

From XF0 an estimate of where the returns are detected at there peak in the RDB. From the following equation Φ₀=Φ_D0−Φ_A0 where Φ_0A is the phase of the return proportional to the azimuth of the return, and Φ_D0 is the phase of the return proportional to the peak of the return, Φ₀

is the phase of the return proportional to the velocity of the return.

Similarly this is performed for XF1 and XF2 hence finding the azimuth of the second and third return.

Analogously a small change in range bin may be taken and we determine XR0, XR1 and XR2 which determines where the peak of the returns in range, this does not help in the evaluation in azimuth. Resolving velocity ambiguity with the taking of meaning delay in time and processing again. Thus we can determine the peak of each return in range and azimuth to obtain the maximum amplitude for each return for further use. A more accurate determination of Φ_D0, Φ_D1 and Φ_D2 is determined by taking the four sets of equations and employing the candidate Φ technique substituting all possible solutions which are restricted to the values that can be in only one range azimuth bin (RAB) but for when greater accuracy is required the number of candidate solutions increase. The candidate solutions should be very close in value for all four sets of data giving very robust and accurate solutions which determines all the parameters of the returns.

Also another correlating and checking operation is to repeat the processing with a close frequency and the results should be very close plus obtaining and obtaining change in return vector (XF). This would be the same for all sets of four sets of data and also would be a check on the Φ_D0, Φ_D1 and Φ_D2 solutions.

Analogously this would process a close sample in the range direction and also for processing other linear arrays.

RECAPPING—We have determined all Φs, in the two return and three return case and as consequently it may be performed for many Φs. The significance of this development is as each RAB that is processed clutter, target, noise, and other returns may be detected and thresholded for importance and later processed to determine if sidelobes, multipath targets, etc.

CORRELATION FACTORS

1—Time delay as many pulses at a time as the number of returns and process data determine all the Φs.

2—Additional channel delay processed again and all Φs should agree

3—Other RAB processed have same return data related to each such as mover should agree

4—IF more than two pulses other dual pulses is processed and results should agree.

5—Other techniques as to be shown later in document and results should agree

6—If a planar array is implemented all other linear arrays should obtain the same results and height and vertical tangential velocity obtained.

The aforementioned system has many advantages such as the following:

1—No clutter cancellation of any kind is required therefore as follows:

• • a) no clutter covariance matrix
  • b) no training data
  • c) no special clutter knowledge required

2—No channel matching required

3—Returns and clutter do not compete with each other in there detection and therefore clutter and returns are thresholded separately and returns are ideally are competing with white noise only and make for an excellent return ratio to noise.

4—Disadvantage many channels-radar receiver front end and a/d converter for each channel

5—Full transmit and receive antenna employed with there full antenna gains

• • a) smaller antenna sidelobes
  • b) full antenna gain
  • c) narrow clutter band width

6—May be applied to two pulses and two pulses or three pulses. Each dual pulse processed as two pulse system and should have same results and correlated.

7—Correlation factors as stated in previous paragraph.

8—The significance of the ability to process many returns in same RAB and determining there amplitude and phases and radial velocity gives the ability to separate clutter and either returns such as bona fide targets, moving clutter, multi path returns, etc. Knowledge aided information would aid in categorizing these returns.

VII—A Dual Pulse at a Time Many Channel System-Second Pulse Channel Delay-Delta C-DPCA Like Operation

As in the two channel-many pulse system the two pulse many channel system described previously as the relation between them as a duality this is also true in this case and will be illustrated here for the two return case only. The dual pulse many channel system is a duality with the dual channel many pulse system.

Same comments as at end of two return case.

We will now analyze the various DPCA type delay in the ΔC methodologies, where only two pulses will be employed. The second pulse will have a number of DPCA type delays ΔC and starting at zero (“0”) and going up to two (2). It does not have to start at zero; it may start at any convenient delay and continue to as many as required. Only two returns will be considered in this analysis.

Mathematic Development (Two Returns)
V₀₀=M₀+M₁ PULSE 1—  (1)
V₀₁=M₀X₀₁+M₁X₁₁ PULSE 2-ΔC DELAY 0  (2)
V₀₂=M₀X₀₂e^jΦA0+M₁X₁₂e^jΦA1 PULSE 2-ΔC DELAY 1  (3)
V₀₃=M₀X₀₃e^j2ΦA0+M₁X₁₃e^j2ΦA1 PULSE 2-ΔC DELAY 2  (4)
V₀₄=M₀X₀₄e^j3ΦA0+M₁X₁₄e^j3ΦA1 PULSE 2-ΔC DELAY 3  (5)

The following is the further definition of Xs in light of the different DPCA delays.
X₀₁=e^jDΦA0/W_M0 at D=0:A₀=1/W_M0=1/A_M0e^{−j(ΨM0+KD0X0)}
X₀₁=A₀
X₀₂=A₀e^{j(ΦA0+ΔKD0X0)}=A₀e^jγ0
X₀₃=A₀e^j2γ0
X₀₄=A₀e^j3γ0
Also in similar manner
X₁₁=e^jDΦA1/W_M1 at D=0:A₁=1/W_M1=1/A_M1e^{−j(ΨM1+KDOX1)}
X₁₁=A₁
X₁₂=A₁e^{j(ΦA1+ΔKDOX1)}=A₁e^jγ1
X₁₃=A₁e^j2γ1
X₁₄=A₁e^j3γ1
We will now rewrite the five equations above with it incorporated.
V₀₀=M₀+M₁ PULSE 1-TIME 1  (1′)
V₀₁=M₀A₀+M₁A₁ PULSE 2-ΔC-DPCA DELAY 0  (2′)
V₀₂=M₀A₀e^jγ0+M₁A₁e^jγ1 PULSE 2-ΔC-DPCA DELAY 1  (3′)
V₀₃=M₀A₀e^j2γ0+M₁A₁e^j2γ1 PULSE 2-ΔC-DPCA DELAY 2  (4′)
V₀₄=M₀A₀e^j3γ0+M₁A₁e^j3γ1 PULSE 2-ΔC-DPCA DELAY 3  (5′)
Employing equations (2′) and (3′) and the variables MoAo and M1A1 and solving for γ₀ and γ₁ we have the following:
M₀A₀=(V₀₁e^jγ1−V₀₂)/(e^jγ1−e^jγ0)  (2″)
M₁A₁=(V₀₂−V₀₁e^jγ0)/(e^jθ1−e^jγ0)  (3″)
Employing equations (3′) and (4′) and the variables MoAo e^jγ0 and M1A1 e^jγ1 and solving for γ₀ and γ₁ we have the following:
M₀A₀e^jγ0=(V₀₂e^jγ1−V₀₃)/(e^jγ1−e^jγ0)  (2′″)
M₁A₁e^jγ1=(V₀₃−V₀₂e^jγ0)/(e^jγ1−e^jγ0)  (3′″)
Next take equation (2′″)/(2″) and equation (3′″)/(3″) we have the following:
e^jγ0=(V₀₂e^jγ1−V₀₃)/(V₀₁e^jγ1−V₀₂)  (6′)
e^jγ1=(V₀₃−V₀₂e^jγ0)/(V₀₂−V₀₁e^jγ0)  (7′)
Equation (6′) or equation (7′) may be solved for γ₀ and γ₁ and consequently Φ_A0 and Φ_A1 and at this point the velocity may be determined with maximum error of a half of a azimuth bin since the returns peak within a halve of a doppler bin where detected. processing like the first frequency.
Taking equation (2″) and solving for MoAo since everything is known on right side of equation
To attain much greater accuracy the returns position of there peaks will be endeavored to be found. This will be performed by inserting to zero fill and processing to attain a frequency close to first frequency and taking equivalent equation in second frequency and solving for M′oA′o since everything is known on right side of this equation.
The ratio of M′o/Mo equals XF0 which is the response to return in the RDB and gives an estimate where the peak of the return is in that RAB.

The ratio of A′o/Ao equals the difference between where frequencies are processed in that RDB. The actual distance is known.
A₀=1/W_M0=1/A_M0e^{−j(ΨM0+KD0X0)}
A′₀=1/W_M0=1/A_M0e^{−j(ΨM0+KD0(X0+ΔX0))}
Solving equation (2″)=K₀e^jβ0
Solving equivalent equation (2′″)=K′₀e^jβ′0
M′₀/M₀=X_F0
A′₀M′₀/A₀M₀=X_F0e^jΔKCMΔxo
Therefore |X_F0|=|K′₀/K₀|and angle of (M′₀/M₀)=β′₀/β₀−ΔK_D0Δ_X0

We this ratio the peak of first return is estimated (Φ_D0) and from the equation Φ_D0 and Φ₀ where the azimuth of the return is calculated from the known or estimated Φ_D0 and Φ₀. If a better estimate of the peak of first return more samples of frequencies may be taken. The best accuracy obtained by getting the frequency sample at the peak of the return.

The analogous procedure would be undertaken for the second return.

This technique is the ΔC DPCA various delays plus and if additional accuracy is required or desired the addition samples at frequencies close to frequency processed is processed. It does not require any channel balancing and DPCA special processing and is very accurate.

Similarly another technique as in two channel technique to determine azimuths is to obtain radial velocities in the two pulse technique. Also similarly the channel balancing terms may be obtained.

B—Similar comments apply as in the two channels at a time since they are a duality of each other. The exchanging of ΦS With Φ_As and solution thereof. The three return system is analogous of the other three return systems related to their related two return systems.

C—Advantages and Disadvantages of System

The aforementioned system has many advantages as the other systems but the outstanding are as the following

1—Channel matching is a problem but utilizing many channels reduces the effect.

2—Dwell time is reduced drastically due to the minimum of only two pulses required limited by the processor to perform the necessary functions.

3—Returns and clutter do not compete with each other in the resulting processing and therefore clutter and returns are thresholded separately and returns are ideally are competing with white noise only and make for an excellent return ratio to noise.

4—Very simple system-less storage-less processing-more hardware due to RF front end and A/D converter required for every channel.

5—May be applied to two or more pulses to give more than one solution and they should be very close to each other. Each processed as two pulse system and should have same results and correlated.

6—The significance of the ability to process many returns in same RDB and determining there amplitude and phases and radial velocity gives the ability to separate clutter and other returns such as bona fide movers, moving clutter, multi path returns, etc. Data base of returns and the mathematical basis and knowledge aided information would aid in categorizing these returns.

7—If more than two pulses are implemented than the results of each pulse should be the same or close to the same

V₀₀—OUTPUT PULSE 1—ΔC=0

V₀₁—OUTPUT CHANNEL 2—ΔC DPCA DELAY 0

V₀₂—OUTPUT CHANNEL 2—ΔC DPCA DELAY 1

V₀₃—OUTPUT CHANNEL 2—ΔC DPCA DELAY 2

V₀₄—OUTPUT CHANNEL 2—ΔC DPCA DELAY 3

X₀₁—return equalizer between PULSE 1 and 2 for return 0-ΔC-DPCA delay 0

X₀₂—return equalizer between PULSE 1 and 2 for return 0-ΔC-DPCA delay 1

X₀₂—return equalizer between PULSE 1 and 2 for return O-ΔC-DPCA delay 2

X₁₁—return equalizer between PULSE 1 and 2 for return 1-ΔC-DPCA delay 0

X₁₂—return equalizer between PULSE 1 and 2 for return 1-ΔC-DPCA delay 1

X₁₂—return equalizer between PULSE 1 and 2 for return 1-ΔC-DPCA delay 2

W_MO—RETURN 0 EQUALIZER BETWEEN PULSE 1 AND 2

W_M1—RETURN 1 EQUALIZER BETWEEN PULSE 1 AND 2

A_MO—AMPLITUDE 0 EQUALIZER BETWEEN PULSE 1 AND 2

A_M1—AMPLITUDE I EQUALIZER BETWEEN PULSE 1 AND 2

Ψ_MO—PHASE 0 EQUALIZER BETWEEN PULSE 1 AND 2

Ψ_M1—PHASE 1 EQUALIZER BETWEEN PULSE 1 AND 2

W_MO—RETURN 0 EQUALIZER BETWEEN PULSE 1 AND 2

W_M1—RETURN 1 EQUALIZER BETWEEN PULSE 1 AND 2

A_MO—AMPLITUDE 0 EQUALIZER BETWEEN PULSE 1 AND 2

A_M1—AMPLITUDE 1 EQUALIZER BETWEEN PULSE 1 AND 2

Ψ_MO—PHASE 0 EQUALIZER BETWEEN PULSE 1 AND 2

Ψ_M1—PHASE 1 EQUALIZER BETWEEN PULSE 1 AND 2

KDO—ΔC DPCA constant that makes returns equal

Ao—constant term of return o

A1—constant term of return 1

As can be observed there is a direct duality between two channels many pulse system and the two pulse many channel system. In the two channel many pulse system you are solving for the radial velocity of the returns (φ_s) and calculate the azimuths (φ_{A s}) and θ_s are the sum of φs and DPCA terms. All other terms are the same.

In the two pulse many channel system you are solving for the azimuth (φ_{A s}) of the returns (φ_s) and calculated the velocity and. γ_s are the sum of φ_{A s} and dpca terms.

Therefore the equations, etc may represent both systems with this noted such as in the claims.

VI—B Three returns are very analogous to other three return techniques

VI—C Combined techniques of section III and IV

D—Combining techniques of section III—Two pulses “N” channel system Φ_D technique and section IIII—ΔC technique—Two pulse “N” channel system

The two techniques employ the same data and may be processed in any manner to facilitate a solution. The following is a list of common solutions and attributes.

• • 1. Same solutions for all parameters such as the following:
    • a) Φ_{D s}, Φ_{A s} and Φ_s and M_S and X_{F s}, X_{R s} and X_{H s}
    • b) Φ_{D s} have many same solutions in the Φ_D technique
    • c) solutions for position and velocity of respective returns
  • 2. (DD technique is more effective but requires more storage and processing but requires one less delay in data
  • 3. Accuracy and robustness of solutions are enhanced
  • 4. Appearance of a very practical system
    VIII—A—ONE CHANNEL SYSTEM—DELT R technique where the same as two channel system but only employing one channel and when Φ_D processing significant time later with the delta R technique the change in amplitude and phase shift of the return will determine the velocity of the return and consequently the azimuth of the return. This system requires a relatively high resolution in range to determine low velocity of returns.

B—One Channel or One Pulse Methodology for Detecting Ships Over Water where the Following:

1—Ship return much larger than water returns the velocity and azimuth of the ship is determined if there is only one return and range, azimuth and velocity is easily determined.

2—Black hole and/or shadow technique to determine range, azimuth and velocity.

3. By delta R technique of section VIII A

4. By one channel or one pulse techniques in the disclosure

5. Combination of any of the techniques that is compatible

IX—ACCOMPANYING TECHNIQUES are illustrated for the many pulse techniques in FIG. 6 which also applies to many channel techniques and also illustrated in FIG. 11 employing groups of data and interlaced data.

A1—ΔI Technique with dual channel system. This technique breaks up the received set of data into two or more interleaved data sets and process each data set as initially independent and the results should agree. The difference being that there is a loss of coherent gain proportional to the number of interlaced sets of data but that may be offset by non-coherently integrated between sets of interlaced data. Also the bandwidth of the filters is increased proportional to the number of interlaced sets of data which may be an advantage or disadvantage according to what is being processed. The processing of the interlaced data sets as well as the full data set may get the advantage of each. In the interlaced data there is a known relationship between each set of data.

Comment:

1. Interleave processing allows wider doppler bin proportional to the number of interleaved sets of data, if that to be effective.

2. Same results for all interleaved data and correlate results

3. Amplitude and phase relationship known between interleaved data sets.

4. Affects on returns of larger doppler bins

• • a) Movers have a bandwidth of 16 hz, accordingly this will determine how many detections per doppler bin.
  • b) Clutter has a bandwidth of the doppler bin.
  • c) Affect on other returns such as jamming, noise antenna sidelobes, other.
  • e) Other.

5. It applies to all systems.

6. The resulting solutions are increased proportional to the number of interleaved sets of data.

7. The doppler filter width and the ambiguous range is increased proportional to the number of interleaved data sets. The ambiguous velocity decreases proportional to the number of interleaved data sets

A2—ΔI Technique with ΔC for single pulse system or with dual pulse system. This technique breaks up the received set of data into two or more interleaved data sets and process each data set as initially independent with the ΔC and/or ΔC delay technique and the results should agree. The difference being that there is a loss of coherent gain proportional to the number of interlaced sets of data but that may be offset by non-coherently integrated between sets of interlaced data. Also the bandwidth of the filters is increased proportional to the number of interlaced sets of data which may be an advantage or disadvantage according to what is being processed. The processing of the interlaced data sets as well as the full data set may obtain the advantage of each. In the interlaced data there is a known relationship between each set of data.

Comment:

1. Interleave processing allows wider azimuth bin proportional to the number of interleaved sets of data, if that to be effective.

2. Same results for all interleaved data and correlate results

3. Amplitude and phase relationship between interleaved data sets.

4. Affects on returns of larger azimuth bins

• • a) Movers have a bandwidth of approximately 16 hz, accordingly this will determine how many detections per azimuth bin.
  • b) Clutter has a bandwidth of the azimuth bin.
  • c) Affect on other returns such as jamming, noise antenna sidelobes, other.
  • e) Other.

5. It Applies to Pulse Systems

6. The resulting solutions are increased proportional to the number of interleaved sets of data.

7. The azimuth filter width and the ambiguous azimuth is increased proportional to the number of interleaved data sets. The ambiguous azimuth decreases proportional to the number of interleaved data sets

B1. ΔI+ΔA Technique for dual channel system

This is similar to the ΔI technique but with each interlaced data set the receive antenna is moved approximately by the antenna beam width divided by the number of interlaced data sets. This technique breaks up the received set of data into two or more interleaved data sets and process each data set as initially independent and the results should agree. The difference being that there is a loss of coherent gain proportional to the number of interlaced sets of data but that may be offset by non-coherently integrated between sets of interlaced data. Also the bandwidth of the filters is increased proportional to the number of interlaced sets of data which may be an advantage or disadvantage according to what is being processed. In the interlaced data there is a known relationship between each set of data. Due to the receive arrays movement between each set of interlaced data set the ratio of the amplitude of the returns between data sets is determined by receive antenna arrays which if known will tell very accurately the returns azimuth and must correlate with results of processing each data set independently.

To attain curve of the ratio between apertures-vs-azimuth of return process significant clutter only data and obtain said curve and employ this in the above technique.

Comment: SAME AS I EXCEPT FOR THE FOLLOWING

1. Amplitude ratio between interleaved data with simultaneously aperture change determines the azimuth of the return

B2. ΔI+ΔA Technique with ΔC for dual pulse systems. This is similar to the ΔI technique but with each interlaced data set the receive antenna is moved approximately by the antenna b beam width divided by the number of interlaced data sets. This technique breaks up the received set of data into two or more interleaved data sets and process each data set as initially independent with the Φ_D and/or ΔC delay technique and the results should agree. The difference being that there is a loss of coherent gain proportional to the number of interlaced sets of data but that may be offset by non-coherently integrated between sets of interlaced data. Also the bandwidth of the filters is increased proportional to the number of interlaced sets of data which may be an advantage or disadvantage according to what is being processed. In the interlaced data there is a known relationship between each set of data. Due to the receive arrays movement between each set of interlaced data set the ratio of the amplitude of the returns between data sets is determined by receive antenna arrays which if known will tell very accurately the returns azimuth and must correlate with results of processing each data set independently.

To attain curve of the ratio between apertures-vs-azimuth of return process significant clutter only data and obtain said curve and employ this in the above technique.

Comment: SAME AS I EXCEPT FOR THE FOLLOWING

1. Amplitude ratio between interleaved data determines the azimuth of the return

2. For Two Pulse System

C1—ΔG technique with dual channel system Φ_D and/or ΔT technique. The technique involves taking the “M” data points and processing them in groups (two, three or more groups). Each group is processed independently and solutions should be very close to the same for each group and also for processing the full set of data. To maintain gain, the full set of data is processed. As the number of data points increases the gain increases and the doppler bin narrows proportional to the size of the data group. The amplitude and phase shift between groups is dependent on the number of data points in a group. Non-coherently processing the groups of data will gain most of the coherent gain of the full set of data. Groups of data applications are implemented mainly on high speed targets which may travel out of the range gate due to their velocity.

Comment:

1. Wider doppler bin with decreasing number of data points in a group.

2. Same results for all groups of data.

3. Amplitude and phase relationship determinable between groups.

4. Same as 4 for delta I.

5. The ambiguous range and velocity is the same but doppler bin widens proportional to size of group.

C2—ΔG technique with ΔC for dual pulse system Φ_D and/or ΔC technique. The technique involves taking the “N” data points and processing them in groups (two, three or more groups). Each group is processed independently and solutions should be very close to the same for each group and also for processing the full set of data. To maintain gain, the full set of data is processed. As the number of data increases the gain increases and the azimuth bin narrows proportional to the size of the data group. The amplitude and phase shift between groups is dependent on the number of data in a group.

Comment:

1. Wider azimuth bin with decreasing number of data points in a group.

2. Same results for all groups of data.

3. Amplitude and phase relationship determinable between groups.

4. Same as 4 for delta I.

5. The ambiguous range and velocity is the same and azimuth bin widens proportional to size of group.

D1—ΔG+ΔA technique for dual channel system Φ_D and/or ΔT technique. Same as K1 but with each group of data the antenna is moved a portion of the bandwidth of the antenna divided by the number of groups. Each group is initially processed independently and result close to same for each group. The amplitude ratio between groups of data is the ratio of the antenna curve at each group position. The antenna curves ratio-vs-azimuth is measured with significant large clutter only data as stated for all change in aperture data. This is performed for all groups.

Comments:

1. same as C1 except for amplitude ratio-vs-azimuth

D2—ΔG+ΔA technique dual pulse system Φ_D and/or ΔC technique. Same as K2 but with each group of data the antenna is moved a portion of the bandwidth of the antenna divided by the number of groups. Each group is initially processed independently and result close to same for each group. The amplitude ratio between groups of data is the ratio of the antenna curve at each group position. The antenna curves ratio-vs-azimuth is measured with significant large clutter only data. The antenna curves ratio-vs-azimuth is measured with significant large clutter only data as stated for all change in aperture data. This is performed for all groups.

Comments:

1. same as C2 except for amplitude ratio-vs-azimuth

D2—ΔG+ΔA technique dual pulse system Φ_D and/or ΔC Two or more groups of data within each group of data having two or more interleaved set of data. Each interleaved set of data within a group of data is processed independently and the solution should be close to the same

Comments:

1. Interleaved data within each group comments are same delta I plus delta A comments

2 Group data has same comments as delta G comments

3. Amplitude and phase relationship between interleaved set of data are determinable

4. Amplitude and phase relationship between groups of data are determinable

E1—ΔI+ΔG—

Two or more groups of data with each group of data having two or more interleaved sets of data. Each interleaved set of data within a group of data is processed independently and the solution should be close to the same

Comments:

1. Interleaved data within each group comments are same delta I plus delta A comments

2 Group data has same comments as delta G comments

3. Amplitude and phase relationship between interleaved set of data are determinable

4. Amplitude and phase relationship between groups of data are determinable

E2—ΔI+ΔG—

Two or more groups of data with each group of data having two or more interleaved sets of data. Each interleaved set of data within a group of data is processed independently and the solution should be close to the same

Comments:

1. Interleaved data within each group comments are same delta I plus delta A comments

2. Group data has same comments as delta G comments

3. Amplitude and phase relationship between interleaved set of data are determinable

4. Amplitude and phase relationship between groups of data are determinable

F1—ΔI+ΔA+ΔG Technique for dual channel system Φ_D and/or ΔT

Same as ΔI+ΔA with the additional factor assuming the same number of interleaved sets of data per group and the aperture change per interleaved sets of data, then the ratio of the output between interleaved sets of data is the same. Each interleaved set of data within a group of data is processed independently and the solution should be close to the same.

F2—ΔI+ΔA+ΔG Technique dual pulse system Φ_D and/or ΔC. Same as ΔI+ΔA with the additional factor assuming the same number of interleaved sets of data per group and the aperture change per interleaved sets of data, then the ratio of the output between interleaved sets of data is the same. Each interleaved set of data within a group of data is processed independently and the solution should be close to the same.

G—Comments on all Developed Techniques

1—If employing the basic system as two receive channels where these two receive channels (or may be summed as one receive channel) there is the ability to process the sum channels and the two receive channels as independent systems. The sum system may be processed as one channel system and the two individual channel or pulse systems as separate system.

H—.Special Application to the Overocean Implementation

Detection of ships and location of its black hole and shadow to determine the ships range, velocity and azimuth and ship classification

H1—Overall detection of the ship at a particular azimuth (proportion to phase shift of where in the radar antenna beam the ship is detected), but when the ship has a relative radial velocity to the radar the phase shift of the ship adds to that due to its azimuth position relative to the boresight of the main beam of the radar, (which has motion compensation for the boresight of the antenna which is zero velocity) such as the following:
Φ_D1=Φ_A1+Φ₁
where Φ_D1 is the phase shift where the ship is detected at its peak Φ_A1 is the phase shift due to the azimuth position. Φ₁ is the phase shift due to its radial velocity.
The ship is detected at Φ_D1, its phase shift due to its radial velocity, Φ₁, plus that due to its azimuth position, Φ_A1, in the main beam of the antenna.
In FIG. 1 that is 1a, b, c, d and 5 is the place where the ship is detected, but the actual position of the ship is 4. The shading of the ship shows this position. The radar illuminates this position, but due to the ships radial velocity it appears at 5. It leaves a lack of detection or black hole where it was at 4. The area 3 is also a lack of detection and this is the lined area, this is the ship blocking radar waves from illuminating the sea behind the ship. These phenomena will assist in the detection of the ship and the determination of its range, azimuth and radial velocity and horizontal tangential and vertical velocity and other ship parameters. Thus, we have a ship shaded area where the ship is, and a lined area where the shadow of the ship that tells us the azimuth (Φ_A1) position, when we detect this lack of signal compared to the sea clutter next to it. The lack of signal should be about comparable to the thermal noise level of the radar. While the sea clutter will have a level in general considerably higher than the thermal noise. This indicates the azimuth position and may be correlated with the previous determination where the azimuth, velocity and range are attained.
The area where there is a lack of detection at that azimuth is the outline of the ship and its shadow.
The area where there is a detection of the ship is the outline of the ship without the shadow.

In theory, if we have very fine range and doppler resolution, this area is defined very well, but we live in the real world and there is a limit to the resolution attained in range and azimuth and will limit our accuracy in attaining the parameters of the ship. FIG. 3 illustrates how the ships height profile would be attained. D being the height of ship and E is he airborne radar and A the height of the radar, B is the slant range of the radar to the ship and F is the angle of the ships shadow with the sea. The solution by simple geometry for height of the ship D.
D≈C*sin F where sin F≈A/(B+C)
By this method knowing “C” slant range of shadow (where lack of detection ends)(D+C)−D=V′_IN=M₁e^jNΦ1 C. This method can be performed across the width of the ship to give its outline.
These measurements are all a function of the attainable range and doppler resolution.

2. The aforementioned analysis may be performed with a one antenna transmit and receive system as developed in the disclosure. If we have only one or the ship return much greater than clutter (ocean) return we have the following:
V₀₁/V₀₀=e^jΦ1

but with the two channel DPCA system with the ΔT technique we have the following:
V₀₁/V₀₀=e^jΦ1/W_M1 where W_M1=e^{j(ΨM1−KD1X1)}

and this determines Φ₁ is the phase shift due to its radial velocity.

The lack of detection that is lower than the ocean return defines the azimuth position of the ship. It allows the measurement of ship parameters as explained previously illustrated in FIG. 5.

3. The aforementioned analysis may be applied for a one transmit or a two or more receive antenna system. Each receive antenna may be treated as independent receive system and treated and analyzed as that so we have two or more independent looks at the ship and the results correlated.

4. The aforementioned analysis may be applied for a one transmit and two receive antenna system with the two or more receive antenna utilizing DPCA techniques to find the range, radial velocity and azimuth independent of the shadow and black hole technique and then correlated with other techniques. Illustrated in FIG. 5.

All techniques or best technique for application may be performed and correlated to obtain the best results.

5. Applying to all systems, if a signal of the ship is sufficient to obtain desired information with the required accuracy:

a) Estimate of the parameters of the ship.

b) Estimate of the radial velocity, azimuth and range.

6. Applying to all systems, if we obtain multiple looks at the ship we will determine the unambiguous radial velocity and tangential velocity and greater accuracy in determining the ship parameters, range, velocity and azimuth.

Multilooks is defined as a look with data point 1 to N and delay the data a portion of the N point such as N/4 and adding N/4 points at the end and performing the same operations. This will result in the increased capability as stated.

When the data is delayed and reprocessed as the first set of data the ships radial motion will be measured by the number of range bins or part of a bin traveled in this time (ΔR/ΔT) gives the true velocity of the ship and will resolve the unambiguous velocity, if any, of the ship without resorting to another PRF. The tangential velocity will also be determined which could not be determined before as a measure of (ΔD/ΔT) doppler bins moved in the time difference. Hence, the total velocity of the ship is determined, not only the radial velocity, the ratio of the radial velocity to the tangential velocity which will give the angle the ship is pointing.

Multilooks of the ships will result in better parameter estimation of the ship and estimation of the range, velocity and azimuth and better estimate of the wake determination and bow wave parameters which will aide in determining all the parameters of the ship.

Isodop correction for the velocity of the ship, focusing the array may be performed to enhance the accuracy of the system.

Motion compensation relative to the boresight of the antenna is assumed.

7. Increased range and doppler resolution-VS-decreased spacing of doppler and range bins with same resolution applied to all systems.

a) Increased range resolution produces smaller range bins but with increased band width of the radar and increased storage for the increased number of cells per given range swath. The processing is increased due to range bin number increasing. The number of range bins increased is independent of the dwell time required for processing.

Increasing the doppler resolution, the number of data points have to be increased this results in increased time on target (ship) called dwell time and more storage and processing is required to handle the increased number of data points. This is a great disadvantage since it restricts the number of tasks to perform in a limited amount of time.

There is an increase the band width of the radar and in storage, processing and dwell time with an increase resolution of range and doppler.

a) A decrease in the range bin spacing is performed by increasing the number of samples, but with the same band width of the radar. This results in an increase in the number of samples (storage) per given range swath and also a proportional increase in processing required with no increase in resolution of range bins, but are more closely spaced samples of range.

A decrease in the doppler bin spacing is performed by adding zeros to the number of data samples stored. When this is processed as if the total number of data points includes all the zeros, the decrease in doppler bin spacing results are proportional to the number of zeros added. There is no additional storage, additional zeros does not constitute additional storage. The processing increased proportional to the additional doppler bins produced.

There is an increase in the number or range bins processed and stored, but no increase in storage due of input data required, but an increase in processing required, but most important no increase in dwell time required.

c) The results are with increased resolution in range and doppler has the disadvantage of increased bandwidth required of the radar and increased dwell time.

On the other hand increasing the number of samples per range bin (oversampling in range), the resolution of range is larger but the affect of lower spacing will give the affect of increased resolution of range without the higher bandwidth. The disadvantage is the larger range bin; this may be ameliorated by the previous discussion.

Adding zeros to the number of data points has the affect of the doppler resolution is the same, but the filters are spaced closer together. For example, if doubling the number of data points by adding an equal number of zeros, then the spacing of the doppler filter is halved. Since there is half the data points, the dwell time is halved. This is analogous to over sampling in range.

The affect of closer spacing of the filters gives increased resolution of azimuth without the increased dwell time. The disadvantage as in over sampling in range there is a larger doppler bin, thus may be ameliorated by previous discussion.

The trade off may be obtain doppler and range resolution as required to detect the ship, but the over sample in range and add zeros in doppler for the increased capability in range and azimuth.

8. Add ISAR processing for further classification of the ship when the ΔT parameters are processed.

9. Additional aides in detection and measurement of ship parameters, especially taken at different times within the same dwell time.

a) Measurement of sea clutter all around the point where the ship is detected.

b) Indicates the sea state conditions.

Determines the parameters of the ship, as well as radial and tangential velocity, especially when motion compensation is performed, as well as the array is focused and there are included isodop corrections.

c) Around the azimuth determination of ship

• • 1) Another measure of the azimuth determination.
  • 2) Another measure of the parameters of the ship.
  • 3) Another measure of the radial velocity and tangential velocity.
  • 4) Ship direction.
  • 5) Bow wave detection, its velocity and heading of ship and radial and tangential velocity of the ship will give another indication of ship parameters and sea clutter around ship.
  • 6) Wake determination will give another indication similar to 5.
  • 7) Measure sea clutter around the indication of azimuth of ship will give additional correlation of all parameters of the ship.

d) Sea state in general

• • 1) Will give sea state conditions.
  • 2) Locations and detections of high and low detections. (shadows of the sea state).
  • 3) May have more than one clutter detection per doppler bin.

10. Problem Areas

• • 1) High clutter sea states are a big problem area and challenge to perform meaningful operations but are feasible.
  • 2) Long dwell time to perform accurate determinations. This is the reason for decreased spacing of doppler bins and increased sampling per range bin might ameliorate that condition.
  • 3) Long range makes things very difficult.
  • 4) Performing surveillance and tracking at the same time as classification of ships.

11. Operate without change in time operation for the surveillance mode.

12. Additional techniques are to deal with the case where sea clutter is significant in value to ship return.

13. Two dimensional array may be employed, as well as a one dimensional array.

14. Combined with overland patent pending of Dual Synthetic Aperture Radar System (DSARS) Ser. No. 10/14,156, by Thomas J. Cataldo with overseas capability and employing the phase corrections and phase coefficient if necessary as explained in DSARS patent.

15. The mode of operation depends on many factors such as range, surveillance, tracking or spot light operation, sea state, etc.

16. Surveillance mode may be combined with spot image mode.

• • Wake and bow wave signatures of ships signatures as a function of their velocity and direction of the ship in help in classification of ships.

17. Surveillance mode could be a one antenna transmit and receive system.

18. High sea state conditions create shadow conditions that could be employed for better processing.

19. Surveillance plus spot image mode may be combined to reduce dwell time and obtain maximum information per unit time.

20. This technique may be extended to space borne operations

21—With the one pulse or one channel technique an electronic scanned array that may be rotated 360 degrees to cover all angles.

22—With the one pulse technique an electronic scanned array does not require motion compensation or phase compensation for boresight of transmitted antenna since only one pulse is required.

Block Diagram of System

This is a simplified system indicating the basic data is received from the radar in digital form and stored and processed in any or combinations of the techniques described in the disclosure. The data is spectrally processed and the detection of the ship is performed together with the detection of the shadow and black hole to determine the ships radial velocity, azimuth and range, as well as the measurement of the ship parameters to classify the ship with as much accuracy as possible.

The delay data is processed to determine radial velocity unambiguously and to determine the horizontal and vertical tangential velocity and the ship parameters more accurately.

I—Special Section on Detection of High Speed Targets

The detection of high speed planes and missiles is very important in the presence of clutter. Processing in groups of data enables the detection and measuring its velocity and azimuth without the target passing thru range bin then and non-coherently integrating from range to range bin. Employing groups of data keeps the target in the range bin long enough to detect its radial velocity and use delta R to determine its unambiguous velocity. In this respect the range bin has to be made wide enough to be able to have the mover a significant time in the range bin. Results may be correlated between range bins.

Being detected in the various range bins at different times to be able to employ the association techniques of tracking the detections thru the range bins and doppler bins with time. This will measure radial velocity unambiguously and tangential velocity.

The processing by groups enables that high speed movers will not travel thru the range bin in the time it is in the group. The integration between groups will give a significant increase in gain for detection.

The processing of each group together with association processing will give the unambiguously radial velocity and the azimuth of the mover.

J—Special Section on Foliage Penetration to Detect Movers and Measure there Velocity and Azimuth Accurately

Foliage penetration radar has to be low in transmission frequency enough to see thru foliage. The ground and large trees will give a significant clutter return and wind blowing foliage a small but significant velocity to clutter and render clutter cancellation techniques ineffective. The USS system will separate out all returns and identify the mover and clutter returns.

K—Special Section on Stealthy Targets

STEALTHY TARGETS PRESENT A SPECIAL CHALLENGE since they have small radar cross sections and with the common stap systems especially difficult because they are embedded in clutter and the common stap systems find it very difficult to differentiate them from clutter residue. the uss has less difficulty clutter and target are measured differently and they don't compete with each other. Consequently very low returns can be detected and there velocity, azimuth and range determined.