Apparatus and method for echo indicator generation

Description

This application claims the benefit of a provisional U.S. application, U.S. Ser. No. 60/562,484 filed Apr. 15, 2004.

BACKGROUND

The invention relates to channel response estimation, and in particular, to echo indicator generation for generating an echo indicator representing echo levels in a multipath channel.

In DTV applications, a television broadcasting station inserts a sync signal in front of signals transmitted in units of horizontal lines. FIG. 1 shows the format of transmitted data sequence 100 consisting of consecutive data segments 106. One data segment 106 corresponds to one horizontal line of the DTV and is made by 832 symbols having a segment sync signal 102 followed by an 828 data symbols 104. In this example, the segment sync signal 102 is a predetermined 4-symbol pattern given by “+5 −5 −5 +5”. And the data symbols carries the television signal and generally having random signal levels, e.g. {±1, ±3, ±5, ±7} in DTV applications. Typically, the segment sync signal is used for the synchronization of the data segments at the receiver side. The operation is referred to as “segment synchronization”, based on which the data symbols can be correctly located in time domain and then retrieved. After segment synchronization, the data sequence is typically input to an adaptive equalizer.

As is well known, in addition to being corrupted by noise, the transmitted signal is also subject to channel distortions and especially the distortions caused by multipath interference. The adaptive equalizer is used to compensate for these transmission channel effects. As is well known in the art, the convergence speed of the adaptive equalizer depends on the initial setting of the equalizer coefficients as well as the step size value used in the coefficient adaptation. If an echo indicator, which approximately indicates the height and location of each echo in a multipath channel, can be obtained, the convergence speed of the adaptive equalizer can thus be enhanced according to the information of the echo indicator.

SUMMARY

An embodiment of the invention provides an echo indicator generation method, generating an echo indicator to indicate echo levels in a channel response of a transmission channel based on an input signal. The input signal comprises segment sync patterns repeating at intervals transmitted over the transmission channel. The input signal is filtered by a match filter having a filter response matched to the segment sync pattern to output a first signal. Thereafter, the first signal is periodically averaged to output a second signal. An echo indicator representing the echo levels of the transmission channel response, is generated based on the second signal. Thus the echo level of the channel response is observable.

Another embodiment of the invention provides an echo indicator generator comprising a match filter, an average unit, and a quantizer. The match filter has a filter response matched to the segment sync pattern for filtering the input signal to output a first signal. The average unit periodically averages the first signal to output a second signal. The quantizer generates an echo indicator representing the echo levels in the transmission channel response based on the second signal.

BRIEF DESCRIPTION OF THE DRAWINGS

The following detailed description, given by way of example and not intended to limit the invention solely to the embodiments described herein, will best be understood in conjunction with the accompanying drawings, in which:

FIG. 1 shows the format of transmitted data sequence 100 consisting of consecutive data segments 106;

FIG. 2 shows an echo indicator generator based on an embodiment of the invention;

FIG. 3a shows the output signal x₅[n] of the average unit 206;

FIG. 3b shows the content in the delay line 2066, which keeps the recent 832 symbols of the signal x₅[n].

FIG. 4 shows a graph of the strength of the signal x₇[n], |x₇[n]|, versus the time index n; and

FIG. 5 is a flowchart of echo indicator generation according to an embodiment of the invention.

DETAILED DESCRIPTION

FIG. 2 shows an echo indicator generator based on an embodiment of the invention. The echo indicator generator 200 comprises a match filter 204, an average unit 206, a post filter 208, and a quantizer 212. The signal received at the echo indicator generator 200, which is delivered from the transmitter side (not shown) over a transmission channel, can be equivalently expressed as,
x₁[n]=x₀[n]{circle over (×)}h[n]  (1)

• • where x₀[n] denotes the transmitted signal at the transmitter side, h[n] denotes the channel response. The match filter 204 having a filter response of h₁[n] for matching the sync pattern {+5, −5, −5, +5} in the received signal x₁[n], thus a filtered signal x₂[n] is generated according to
    x₂[n]=x₁[n]{circle over (×)}h₁[n]  (2)
  • where the filter response h₁[n] typically equals to {+1, −1, −1, +1}. The filtered signal x₂[n] is then input to an average unit 206 for periodically averaging the filtered signal x₂[n].

The average unit 206 comprises a first multiplier 2062, an adder 2064, a delay line 2066, and a second multiplier 2068. The first multiplier 2062 multiplies the filtered signal x₂[n] with a preset attenuation factor c_a, which is preferably a positive real number close to zero. The adder 2064 sums up the output signal x₃[n] of the first multiplier 2062 and the output signal x₆[n] of the second multiplier 2068, and generates the signal x₄[n], which is fed to the delay line 2066. The delay line 2066 comprises 832 delay cells. The output signal of the delay line 2066 is denoted by x₅ [n]. The second multiplier 2068 multiplies the signal X₅[n] with a preset number (1−c_a) to generate the signal x₆[n]. In other words, the average unit 206 performs a function given by x 5 ⁡ [ n ] = c a · x 2 ⁡ [ n ] + ( 1 - c a ) · x 5 ⁡ [ n - 832 ] = ∑ k = 0 ∞ ⁢ c a · ( 1 - c a ) k · x 2 ⁡ [ n - 832 · k ] ( 3 )

This recursive integration given in (3) is usually referred to as a running average algorithm in a periodic way. Note that the sync signal in each data segment are identical, therefore these sync signal will be summed together by the average unit in a constructive way. On the other hand, the data symbols in each data segment are random and they will cancel each other after being processedby the average unit 206. As a result, the average unit 206 acts like a low pass filter to pass the sync signal while filter out the data symbols.

From (1), (2) and (3), we have x 5 ⁡ [ n ] = ⁢ ∑ k = 0 ∞ ⁢ c a · ( 1 - c a ) k · x 2 ⁡ [ n - 832 · k ] = ⁢ ∑ k = 0 ∞ ⁢ c a · ( 1 - c a ) k · ( x 0 ⁡ [ n - 832 · k ] ⊗ h ⁡ [ n ] ⊗ h 1 ⁡ [ n ] ) = ⁢ h ⁡ [ n ] ⊗ ∑ k = 0 ∞ ⁢ c a · ( 1 - c a ) k · ( x 0 ⁡ [ n - 832 · k ] ⊗ h 1 ⁡ [ n ] ) ( 4 )

Since the filter response h₁[n] of the match filter 204 is set to match the sync signal, the output signal x₂ [n] of the match filter 204 will have a local maximal peak, which corresponds to an echo in the channel response, occurs when the sync signal is matched. FIG. 3a shows the output signal x₅[n] of the average unit 206. FIG. 3b shows the content in the delay line 2066, which keeps the recent 832 symbols of the signal x₅[n]. Ideally, a local maximal peak 32 can be found every 832 symbols. In general, if the channel response contains N echoes, then N local optimal peaks of the output signal x₅[n] during a segment period can be found correspondingly.

Preferably, but optionally, a post filter 208 is used to sharpen the local maximal peaks 32 in signal x₅[n] such that the local maximal peaks can more easily and clearly been identified. The filter response of the post filter 208 is chosen such that the resultant output obtained by letting a sync signal being filtered by the matched filter 204 and the post filter 208, will resemble an impulse signal. As an example, the filter response of the post filter 208 is set to be {1, −0.25, 4, −0.25, 1}.

By utilizing the fact that each echo in the channel response will produce a corresponding local maximal peak in the signal x₇[n], which is the output signal of the post filter 208, the quantizer 212 receives and quantizes strength of the signal X₇ [n] to obtain the echo indicator xE[n].

FIG. 4 shows a graph of the strength of the signal x₇[n], |x₇[n]|, versus the time index n. The domain of the |x₇[n]| is divided into K non-overlapping regions, denoted by region #0˜K-1, where K is a positive integer. The region number of the value of |x₇[n]| is thus reported as the echo indicator x_E[n] Preferably, the K regions is obtained by dividing the range of (0, x_M), where the x_M is the maximal peak within one period of data segment. In other words, the quantizing levels are chosen to be relative to the value of x_M.

As an application, the echo indicator x_E[n] can be applied to an adaptive equalizer. The equalizer may adopts different step size for the coefficient adaptation to be proportional to corresponding echo level in the echo indicator x_E[n]. In this application, a segment sync signal, which indicates the time point at which the first echo occurs within one segment period, would be needed for correctly associating the echo indicator with the coefficients of the adaptive equalizer. The method of generating segment peak detector is a prior art and is not described herein. In such an application, the convergence speed of the adaptive equalizer can be increased while the convergence error is reduced.

FIG. 5 is a flowchart of echo indicator generation according to an embodiment of the invention. In step 504, a match filter 204 having a filter response h₁[n] matched to the sync pattern is used to filter the input signal x₁[n], thus a filtered signal x₂[n] is generated. In step 506, the filtered signal x₂[n] is input to an average unit 206 for periodically averaging the filtered signal x₂[n], thereby generating a signal x₅[n]. In step 508, which is a preferable but optional, the post filter 208 receives the signal x₅[n] for sharpening the local maximal peaks in signal x₅[n] and generates a signal x₇[n]. In step 510, the signal x₇[n] is then output to the quantizer 212 to obtain the echo indicator x_E[n].

While the invention has been described by way of example and in terms of the preferred embodiment, it is to be understood that the invention is not limited thereto. To the contrary, it is intended to cover various modifications and similar arrangements (as would be apparent to those skilled in the art) Therefore, the scope of the appended claims should be accorded the broadest interpretation so as to encompass all such modifications and similar arrangements.