Techniques for energy saving and smart echo cancellation

Abstract

The invention relates to an apparatus for determining filter coefficients of an echo compensation filter, comprising a FIR filter having a set of M+1 filter coefficients for filtering values of a reference signal to obtain a first filtered reference signal. The apparatus comprises an interval selector for selecting a value of the reference signal associated with a k-th time interval of an echo impulse response having N time intervals. The apparatus comprises M delay elements arranged after the interval selector for successively delaying the selected values of the reference signal. The apparatus comprises M+1 error estimators. A first error estimator is coupled to an output of the interval selector. Each of the remaining M error estimators is coupled to an output of a respective delay element. The M+1 error estimators are for determining the M+1 filter coefficients such that a measure of a deviation between the first filtered reference signal and the received signal is minimized.

Description

TECHNICAL FIELD

[0001] The present invention relates to techniques for energy saving and smart echo cancellation, in particular with on-board diagnostic capabilities, for example for automotive, industrial and consumer electronics. The present invention relates in particular to an apparatus and a method for determining filter coefficients of an echo cancellation filter for reducing an echo signal in a communication channel having a long impulse response extending over N time intervals. BACKGROUND

[0002] Wired based high speed serial communication PHYs (Physical Interfaces) 110, 120 exchange data over a cable 130, the maximum length of which is defined by a standard (e.g. IEEE 802.3cy) as shown in Figure 1 The length of the cable 130 can vary from a fraction of a meter to 100m and more. The cable 130 is usually divided into different segments and connected by means of inline connectors 131 as shown in Figure 1 When the impedance changes at the connection points 131, 132 between the cable and the connectors, the transmitted signal 112 is reflected back as an echo signal 123. In a full duplex communication system, the reflected echo signal 123 is superimposed on top of the received signal 122. The reflected echo signal 123 needs to be cancelled at the receiver.

[0003] Finite Impulse Response (FIR) based digital echo cancellation uses a signal processing algorithm to remove the reflected signal 123 from the received signal 122. The number of taps of the FIR filter depends on the length of the cable 130 and the baud rate. The channel behavior varies over time, so an adaptive cancellation is needed. With increasing data rates, the number of FIR filter taps for echo cancellation explodes. For example, for 100BASE-T1 and 1000BASE-T1, echo cancellation can still be handled. However, for 10G / 25GBASE-T1, the digital echo cancellation can be very complex, because an echo cancellation implementation with about 5600 taps is a very complex task, especially when using multiple echo taps equal to the full length of the echo response. In this case, the implementation becomes more complex and very power consuming. SUMMARY

[0004] The present invention provides a solution for significantly reducing echo signals in a communication channel (e.g. a wired based high speed serial communication channel as shown in Figure 1 to overcome the above mentioned problems.

[0005] In particular, the present invention provides the concept of significantly reducing the number of echo taps compared to echo cancellation covering the full echo response.

[0006] The above and other objects are achieved by the features of the independent claims. Further implementation forms are apparent from the dependent claims, the description and the figures.

[0007] The focus of the present invention is a technique that reduces the need to cover the full window length echo taps for echo cancellation, thereby reducing power consumption, complexity, and silicon area usage. The overall echo response is divided into multiple segments, and an echo cancellation block is designed to cover the first segment. At the beginning of the communication, the echo of the first segment is estimated and cancelled. Once the coefficients reach their steady state values, the coefficients are frozen, i.e., kept constant, and the process moves to the second segment to detect whether there is any significant echo to cancel. Similarly, the process repeats from the first segment to the last segment and again. The coefficients are updated according to known adaptive algorithms, e.g., least mean square (LMS) algorithm. The second segment does not have to be adjacent to the first segment in the echo response; the second segment can be any segment in the echo response other than the first segment.

[0008] This technique significantly reduces the number of echo taps compared to the number of echo taps needed when covering the full echo response. The new technique disclosed below can maintain the same performance compared to full echo response compensation.

[0009] The solution presented in the present invention is based on the following points: (1) the ability to detect and cancel the echo without using full length echo taps on the signal processing block. (2) applying a technique to reuse a limited number of echo taps to detect and cancel the echo propagating on the echo response. (3) applying a technique to decide whether the detected echo needs to be cancelled or can be left as uncancelled. (4) a method to move the window of the FIR filter from one segment to another once the coefficients reach their steady state values.

[0010] The concepts described in the present invention can be applied in automotive applications using wired-based data communication links, such as Ethernet cables defined by 1000BASE-T1, 10GBASE-T1, and future 25GBASE-T1. Since future autonomous vehicles need more and more bandwidth to reliably transmit sensor data to the central processing unit, the techniques according to the present invention can be advantageously applied to improve the efficiency of data transmission.

[0011] In addition to automotive, the techniques described herein can also be applied in industrial and automation applications as well as consumer electronics.

[0012] To describe the present invention in detail, the following terms, abbreviations, and notations are used:

[0013] PHY physical layer device or physical interface

[0014] MII media independent interface

[0015] DAC digital-to-analog converter

[0016] LMS least mean square

[0017] FIR finite impulse response

[0018] According to a first aspect, the invention relates to an apparatus for determining filter coefficients of an echo compensation filter for reducing an echo signal in a communication channel based on a received signal, the received signal being a receivable version of a known reference signal transmittable on the communication channel, the receivable version of the reference signal being affected by the echo signal, wherein an impulse response of the communication channel extends over N time intervals, the apparatus comprising: a finite impulse response filter having a set of M+1 filter coefficients, the finite impulse response filter being configured to filter values of the reference signal using the M+1 filter coefficients to obtain a first filtered reference signal; an interval selector configured to select a value of the reference signal associated with a k-th time interval of the N time intervals; M delay elements configured to successively delay the selected value of the reference signal, the M delay elements being arranged after the interval selector; M+1 error estimators, wherein a first error estimator is coupled to an output of the interval selector, and wherein each of the remaining M error estimators is coupled to an output of a respective delay element, wherein the M+1 error estimators are configured to determine the M+1 filter coefficients such that a measure of a deviation between the first filtered reference signal and the received signal is minimized.

[0019] M can be any integer. A time interval can comprise, for example, a number of sample values corresponding to the number M+1 of filter coefficients. The number N of time intervals can be any integer, for example greater than 1. The k-th time interval can be, for example, any one of the N time intervals selected by a controller.

[0020] Such an apparatus can reduce echo cancellation complexity due to interval selection by the interval selector, thereby reducing power consumption and reducing the silicon area used.

[0021] When covering a full echo response, the apparatus significantly reduces the number of echo taps compared to the number of echo taps required. The same performance can be maintained when using the disclosed apparatus compared to full echo response compensation.

[0022] For example, the apparatus can be advantageously applied in wired full-duplex chip-to-chip communication.

[0023] In an exemplary implementation of the apparatus, the M+1 error estimators are mean square error estimators, in particular least mean square error estimators or least mean square estimators.

[0024] This provides the advantage that well-known iterative procedures can be applied to update the filter coefficients. Such iterative procedures are readily available in software libraries. These procedures guarantee a fast convergence of the filter coefficients even in time-variant environments.

[0025] In an exemplary implementation of the apparatus, the measure of the deviation between the first filtered reference signal and the received signal is a mean squared deviation between the value of the first filtered reference signal and the value of the received signal.

[0026] This provides the advantage that the error signal is minimized. The filter represents an accurate model of the communication channel.

[0027] In an exemplary implementation of the apparatus, the apparatus comprises a memory for storing the M+1 filter coefficients of the value of the reference signal selected by the interval selector.

[0028] This provides the advantage that the actual filter coefficients processed for the respective time interval of the impulse response can be stored in the memory for later use. Thus, a sequential determination of all filter coefficients of the full impulse response of the communication channel can be performed without increasing the computational complexity.

[0029] In an exemplary implementation of the apparatus, the memory is for storing the M+1 filter coefficients of each of the N time intervals applied by the interval selector.

[0030] By storing the filter coefficients of each time interval, all filter coefficients of the full impulse response of the communication channel can be stored in the memory and improve the compensation accuracy when filtering with the complete set of filter coefficients. However, the chip size can be designed to process only one or two time intervals of the complete impulse response at a time.

[0031] In an exemplary implementation of the apparatus, the memory is for storing the M+1 filter coefficients together with the value of the k-th time interval applied by the interval selector.

[0032] This provides the advantage that each set of M+1 filter coefficients is stored in the correct order. This simplifies the access to the filter coefficients by the full impulse response compensation filter later on.

[0033] In an exemplary implementation of the apparatus, the M+1 error estimators are for determining the M+1 filter coefficients for which the measure of the deviation between the first filtered reference signal and the received signal is below a predefined threshold.

[0034] This provides the advantage that the filtering accuracy can be adjusted or preconfigured. By defining the threshold, the convergence time of the filter can be flexibly adjusted.

[0035] In an exemplary implementation of the apparatus, the apparatus comprises a subtracter for determining a measure of the deviation by subtracting the first filtered reference signal from a digital representation of the received signal.

[0036] This provides the advantage that the error signal can easily be determined by simply subtracting the filtered signal (i.e. the first filtered reference signal) from the digital received signal (i.e. the digital representation of the received signal). The subtraction is performed in the digital domain.

[0037] In an exemplary implementation of the apparatus, the apparatus comprises an echo compensation filter having a set of N times (M+1) filter coefficients, the echo compensation filter being configured to filter a value of a transmit signal transmittable over the communication channel by using the M+1 filter coefficients of each of the N time intervals.

[0038] This provides the advantage that a different filter set of M+1 filter coefficients determined for each time interval selected by the interval selector can be used for compensating the full echo impulse response. Thus, echoes at different time interval positions in the impulse response of the communication channel can be compensated.

[0039] In an exemplary implementation of the apparatus, the apparatus comprises a controller configured to send a selection signal to the interval selector for selecting a value of the k-th time interval applied by the interval selector.

[0040] This provides the advantage that the interval selector can be flexibly controlled by the controller. For example, a sequential filtering of different time intervals or time periods can be performed or it can be preferred to filter a special time interval (e.g. a time interval in which an echo is known to be located).

[0041] In an exemplary implementation of the apparatus, the apparatus comprises a plurality of spare resources for increasing the number of filter coefficients of the finite impulse response filter.

[0042] This provides the advantage that the precision of the FIR filter can be flexibly adjusted. For example, if a high precision of the echo compensation is required, the number of filter coefficients can be increased by additionally using the spare resources. If a low precision is required, less filter coefficients than the above-mentioned M+1 filter coefficients can be used by allocating some of the M+1 filter coefficients as spare resources.

[0043] In an exemplary implementation of the apparatus, the spare resources comprise at least one additional filter coefficient; the finite impulse response filter is configured to filter a value of the reference signal using the M+1 filter coefficients and the at least one additional filter coefficient to obtain the first filtered reference signal.

[0044] This provides the advantage that by using filter coefficients larger than M+1 for filtering, a higher precision of the filtering process can be obtained.

[0045] In an exemplary implementation of the apparatus, the apparatus comprises a combining element for combining the finite impulse response filter with the plurality of spare resources.

[0046] This provides the advantage that by the combining element, the spare resources can be easily and efficiently added to the existing M+1 filter coefficients.

[0047] In an exemplary implementation of the apparatus, the controller is configured to send an enabling signal to the plurality of spare resources and the combining element, the enabling signal being configured to enable the plurality of spare resources and to enable the combining element.

[0048] This provides the advantage that the usage of the spare resources can be efficiently controlled by the controller.

[0049] In an exemplary implementation of the apparatus, the apparatus comprises an analog front end configured to receive the receive signal, and an analog-to-digital converter configured to convert the receive signal processed by the analog front end into a digital representation.

[0050] This provides the advantage that the receive signal can be efficiently converted into a digital representation for combination with the digital signal output of the digital FIR filter, i.e. the first filtered reference signal.

[0051] According to a second aspect, the present application relates to a method for determining filter coefficients of an echo compensation filter for reducing an echo signal in a communication channel based on a receive signal, the receive signal being a receivable version of a known reference signal transmittable on the communication channel, the receivable version of the reference signal being affected by the echo signal, wherein an impulse response of the communication channel extends over N time intervals, the method comprising filtering values of the reference signal by using a finite impulse response filter having a set of M+1 filter coefficients to obtain a first filtered reference signal, selecting, by a interval selector, a value of the reference signal associated with a k-th time interval of the N time intervals, successively delaying, by M delay elements, the selected value of the reference signal, the M delay elements being arranged after the interval selector, determining, by M+1 error estimators, the M+1 filter coefficients to minimize a measure of a deviation between the first filtered reference signal and the receive signal, wherein a first error estimator is coupled to an output of the interval selector, and wherein each of the remaining M error estimators is coupled to an output of a respective delay element.

[0052] This method provides the same advantages as described above for the apparatus of the first aspect. Namely, the method can advantageously reduce the echo cancellation complexity due to the interval selection of the interval selector. When covering the full echo response, the method significantly reduces the number of echo taps compared to the number of echo taps required. The same performance can be maintained when using the disclosed method compared to full echo response compensation.

[0053] According to a third aspect, the application relates to a computer program product comprising computer executable code or computer executable instructions, which when executed by at least one computer, cause the at least one computer to perform the method according to the second aspect.

[0054] The computer program product can comprise a non-transitory computer readable storage medium having stored the program code for use by or in connection with an instruction execution system, apparatus, or device.

[0055] The computer program product can be run on a computer, e.g. on a processor or controller of a communication system using a wired based data communication link 130 as shown. Figure 1

[0056] Using such a computer program product improves the efficiency of the echo compensation due to the reduced number of echo taps. BRIEF DESCRIPTION OF DRAWINGS

[0057] Other embodiments of the application will be described in connection with the following drawings.

[0058] Figure 1 A schematic diagram showing the source of the echo signal 123 in the wired communication system 100 is shown.

[0059] Figure 2 An exemplary impulse response 200 is shown, which has unpredictable echoes 221, 222 due to the age of the cable and connectors.

[0060] Figure 3 A block diagram of an apparatus 300 for determining filter coefficients of an echo compensation filter having short echo tap lengths according to the application is shown.

[0061] Figure 4 A block diagram of an apparatus 400 for determining filter coefficients of an echo compensation filter according to the application is shown, which covers all segments of the echo response using segment shifting techniques.

[0062] Figure 5 A schematic diagram of a method 500 for determining filter coefficients of an echo compensation filter according to the application is shown. DETAILED DESCRIPTION​

[0063] In the following detailed description, reference is made to the accompanying drawings, which form a part hereof, and in which are shown by way of illustration specific aspects in which the application can be practiced. It is to be understood that other aspects can be utilized and structural or logical changes can be made without departing from the scope of the present application. The following detailed description, therefore, is not to be taken in a limiting sense, and the scope of the present application is defined by the appended claims.

[0064] It should be understood that the remarks regarding the method described apply analogously to a device or system for performing the method or, vice versa, that the remarks regarding a device or system apply analogously to the method performed by the device or system. For example, if a specific method step is described, a corresponding device can include a unit for performing the described method step, even if such a unit is not explicitly described or shown in the figures. Further, it should be understood that features of the various exemplary aspects described herein can be combined with each other, unless explicitly stated otherwise.

[0065] The devices and methods described herein can be used for transmitting and / or receiving data over a wired-based serial data communication link, e.g. according to IEEE 802.3, in particular IEEE 802.3cy for wired Ethernet. The original Ethernet used coaxial cables as a shared medium, while newer Ethernet variants use twisted pair and fiber optic links in conjunction with switches. The Ethernet standards include several wiring and signaling variants of the OSI physical layer (PHY) used with Ethernet.

[0066] Figure 1 A schematic diagram showing the source of the echo signal 123 in a wired communication system 100.

[0067] Two wired-based high-speed serial communication PHYs 110, 120 exchange data over a wired-based data communication link 130. Each PHY 110, 120 can transmit and / or receive data. The length of the cable 130 can vary from a fraction of a meter to a maximum length defined by a standard, e.g. IEEE 802.3, in particular IEEE 802.3cy. The length of the cable 130 varies from a fraction of a meter to 100 m and more. The cable 130 is usually divided into different segments and connected by inline connectors 131, 132. When the impedance changes at the connection points 131, 132 between the cable and the connector, the transmitted signal 112 is reflected back as an echo signal 123. In a full-duplex communication system, the reflected echo signal 123 is superimposed on top of the received signal 122. The reflected echo signal 123 needs to be cancelled at the receiver. The echo signal 123 can be significant and adds significantly to complexity and power consumption.

[0068] For example, the line impedance 135 of the cable 130 left side (i.e. for the PHY 110) can be 100 Ohm. For example, the line impedance 136 of the cable 130 right side (i.e. for the PHY 120) can be 100 Ohm. In one example, the wired data communication link 130 can comprise a shielded twisted pair cable.

[0069] Figure 2 An exemplary impulse response 200 is shown which has unpredictable echoes 221, 222 due to the age of the cable 130 and connectors 131, 132 as shown in Figure 1

[0070] In addition to 10 Gbit / s Ethernet PHYs, the cable return loss quality is expected to improve. Since the quality of the cables and connectors used in 25 Gbit / s technology is expected to be better, it can not be necessary to use full length echo cancellation in terms of the number of echo taps at the receiver. Instead, only the relevant window length 210 of the echo needs to be cancelled, the rest can be left as noise, as shown in Figure 2 However, in a real use case, the cable or connectors can degrade over time and can lead to unpredictable echoes 221, 222, as shown in Figure 2 This will lead to a decrease in SNR and most likely to an unstable link due to high bit error rate or the link can even drop. For example, during installation of the cable in a car, the return loss specification can pass, but no one can guarantee the lifetime of the car.

[0071] The present invention proposes a new concept of echo cancellation with reduced number of taps, as described in the following with respect to Figure 3 and Figure 4 This concept supports to compensate for echoes that can occur during the lifetime of a car. It is understood that the concept described in the present invention is not limited to the car use case. The concept can also be applied to other use cases, for example consumer electronics or automation, etc.

[0072] Figure 3 A block diagram of an apparatus 300 for determining filter coefficients of an echo compensation filter with a short echo tap length according to the present invention is shown.

[0073] The apparatus 300 can determine filter coefficients of an echo compensation filter for reducing an echo signal 123 in a communication channel 130 based on a received signal 122, for example, as shown in Figure 1 The received signal 122 can be a receivable version of a known reference signal 301 that can be transmitted on the communication channel 130. This receivable version of the reference signal 301 will be affected by the echo signal 123 or Figure 1 Figure 2 ​​the impact of the illustrated echo signals 221, 222. The impulse response 304 of the communication channel 130 extends over N time intervals 342a-d, as Figure 3 as illustrated in the upper part of Fig. 3.

[0074] The apparatus 300 comprises finite impulse response filters 311a-c, 312a-d, 314a-c with a set of M+1 filter coefficients 313a-d. The finite impulse response filters are configured to filter values of the reference signal 301 using the M+1 filter coefficients 313a-d to obtain a first filtered reference signal 315.

[0075] The apparatus 300 comprises an interval selector 340 configured to select a value of the reference signal 301 associated with a k-th time interval of the N time intervals 342a-d.

[0076] The apparatus 300 comprises M delay elements 321a-c configured to sequentially delay the selected values of the reference signal 301. The M delay elements 321a-c are arranged after the interval selector 340, as Figure 3 illustrated.

[0077] The apparatus 300 comprises M+1 error estimators 331a-d. A first error estimator 331a is coupled to an output of the interval selector 340. Each of the remaining M error estimators 331b-d is coupled to an output of a respective delay element 321a-c. The M+1 error estimators 331a-d are configured to determine the M+1 filter coefficients 313a-d such that a measure of a deviation between the first filtered reference signal 315 and the received signal 122 is minimized.

[0078] M can be any integer. A time interval can comprise, for example, a number of sample values corresponding to the number M+1 of filter coefficients. The number N of time intervals can be any integer, for example greater than 1. The k-th time interval can be, for example, any one of the N time intervals selected by a controller.

[0079] The M+1 error estimators 331a-d can be mean square error estimators, for example least means square (LMS) error estimators or least mean square estimators.

[0080] The measure of a deviation between the first filtered reference signal 315 and the received signal 122 can be a mean square deviation between values of the first filtered reference signal 315 and the received signal 122.

[0081] The apparatus 300 can comprise a memory 370 configured to store the M+1 filter coefficients 313a-d of the values of the reference signal 301 selected by the interval selector 340.

[0082] The memory 370 can be used to store the M+1 filter coefficients 313a-d for each of the N time intervals 342a-d applied by the interval selector 340.

[0083] The memory 370 can be used to store the M+1 filter coefficients 313a-d together with the value of the k-th time interval applied by the interval selector 340.

[0084] The M+1 error estimators 331a-d can be used to determine that the measure of the deviation between the first filtered reference signal 315 and the received signal 122 is below a predefined threshold for the M+1 filter coefficients 313a-d.

[0085] The apparatus 300 can comprise a subtractor 330 for determining the measure of the deviation by subtracting the first filtered reference signal 315 from the digital representation 302 of the received signal 122.

[0086] The apparatus 300 can further comprise an echo compensation filter (not shown in the figure) having a set of N times (M+1) filter coefficients. The echo compensation filter can be used to filter a value of a transmit signal transmittable on the communication channel 130 by using the M+1 filter coefficients 313a-d for each of the N time intervals 342a-d. Figure 3

[0087] The apparatus 300 can comprise a controller 350 for sending a selection signal 351 to the interval selector 340 for selecting the value of the k-th time interval applied by the interval selector 340. By means of the controller 350, the respective time intervals or time segments 342a, 342b, 342c, 342d of the channel impulse response 304 can be adjusted.

[0088] The apparatus 300 can comprise a plurality of spare resources 371, e.g. storage units, for increasing the number of filter coefficients of the finite impulse response filters 311a-c, 312a-d, 314a-c.

[0089] The spare resources 371 can comprise one or more additional filter coefficients 313e, 313f. The finite impulse response filters 311a-c, 312a-d, 314a-c can be used to filter a value of the reference signal 301 using the M+1 filter coefficients 313a-d and at least one additional filter coefficient 313e, 313f to obtain the first filtered reference signal 315.

[0090] ​The apparatus 300 can also include a combining element 372 for combining the finite impulse response filters 311a-c, 312a-d, 314a-c with the plurality of spare resources 371. The combining element 372 can be a logic circuit for adding the spare resources to the FIR filters 311a-c, 312a-d, 314a-c and / or the delay elements 321a-c.

[0091] The controller 350 can be configured to send an enable signal 352 to the plurality of spare resources 371 and the combining element 372. The enable signal 352 can be configured to enable the plurality of spare resources 371 and the combining element 372.

[0092] The apparatus 300 can also include an analog front end 381 for receiving the received signal 122 and an analog-to-digital converter 382 for converting the received signal 122 processed by the analog front end 381 into a digital representation 302.

[0093] The apparatus 300 can detect and cancel echoes without using full-length echo taps at the signal processing block. The controller 350 can provide a smart echo cancellation with on-board diagnostic capability by controlling the behavior of the echo cancellation. Specifically, the controller 350 can control the reuse of a limited number of echo taps to detect and cancel echoes propagating on the echo response. The controller 350 can decide which of the detected echoes needs to be canceled and which of the detected echoes can be kept as uncanceled echoes. Once the coefficients reach their steady-state values, the controller 350 can move the window of the FIR filters from one segment to another, e.g., as shown in more detail in Figure 4 .

[0094] Below, exemplary functions of the apparatus 300 shown in Figure 3 are described. The apparatus 300 can perform the following operations:

[0095] As shown in Figure 3 , the complete impulse response 304 is divided into several segments: Seg-1 (342a), Seg-2 (342b), Seg-3 (342c)... Seg-n (342d).

[0096] The number of echo tap allocations is reduced and M taps are allocated for each segment.

[0097] The upper M number of delay elements 311a, 311b, 311c are used to compute the echoes to be canceled, while the lower M number of delay elements 321a, 321b, 321c are used to compute the coefficients, and these computed coefficients are stored in the memory 370, as shown in Figure 3 .

[0098] The control block 350 controls the checking of the computed coefficients and the selection of the number of delays (351) to move the FIR filter to a specific segment. If additional resources are needed to compute the echo cancellation control block, the resources 371 that can be used as accelerators are enabled (352).

[0099] C i is the current value of the coefficient, C i+1 is the new value of the computed coefficient.

[0100] Let the algorithms 331a, 331b, 331c, 331d compute the coefficients of a segment until they reach a saturation value. A typical algorithm is the commonly used least mean square (LMS).

[0101] The coefficient saturation is detected according to a sensitivity analysis, for example dC / dt = ~ 0.

[0102] The index and the value of each coefficient are stored into the memory 370.

[0103] These stored values are used to successively cancel the echo of each received symbol of a segment.

[0104] The term Z -M , Z -2M and Z -kM is a set of delays that help to move the FIR filter from one segment to another segment as shown in Figure 4 .

[0105] All the delays Z -M , Z -2M and Z -kM can be filled from the beginning.

[0106] Once the first segment 342a reaches its steady state value, the FIR filter can be moved to the second segment Seg-2 using the switches 401 as shown in Figure 4 . The segments to be processed do not have to be in the same order as shown in Figure 4 . Any other sequence can also be used, for example Seg-1, Seg-N, Seg-2, Seg-N-1; or Seg-1, Seg-N, Seg-3, Seg-2, etc.

[0107] The adaptive algorithms 331a, 331b, 331c, 331d readjust the gains and reach the steady state relatively early.

[0108] Figure 4 A block diagram of an apparatus 400 for determining filter coefficients of an echo compensation filter according to the present application is shown, which covers all segments of the echo response using a segment shifting technique.

[0109] The apparatus 400 corresponds toFigure 3 The apparatus 300 shown in Fig. 3, wherein the interval selector 340 and the memory 370 are shown in more detail to illustrate the segment shifting technique for all segments, i.e. time intervals 342a, 342b, 342c, 342d, of the echo response 304.

[0110] The interval selector 340 can be controlled by the controller 350 using the switch 401 to switch the respective segment, i.e. time interval 342a, 342b, 342c, 342d, of the echo response 304. Depending on the selected value k, the reference signal 301 is delayed by 0, M, 2M,... kM samples.

[0111] The FIR filter comprises M delay elements 321a, 321b, 321c, 321d and M+1 multipliers 312a, 312b, 312c, 312d for applying the respective filter coefficients stored in the respective memory cells 373a, 373b, 373c, 373d for each of the time intervals 342a, 342b, 342c, 342d selected by the controller 350.

[0112] It is noted that in Figure 4 the FIR filter is exemplarily represented by four blocks of delay elements Z -1 and a corresponding number of filter coefficients, whereas in Figure 3 the FIR filter is exemplarily represented by three blocks of delay elements Z -1 and a corresponding number of filter coefficients. In both figures, the length of the FIR filter corresponds to M+1 to cover the respective time segments 342a, 342b, 342c, 342d of the echo impulse response 304 as shown in Figure 3 and Figure 4 .

[0113] When the next time interval is selected by the switch 401 controlled by the controller 350, the next set of memory cells 374a, 374b, 374c, 374d is selected for the FIR filtering as shown in Figure 4 .

[0114] The FIR filter provides a first filtered reference signal 315 subtracted from a digital representation of the received signal 122 comprising the echo signal 123 to obtain an echo-free signal 303, also denoted as error signal in the following, which is used by the adaptation algorithm for adjusting the filter coefficients. The received signal 122 is passed through an analog front end 381 and an ADC 382 to obtain the digital received signal 302.

[0115] When applying the LMS algorithm as an example for adjusting the filter coefficients, the steepest descent is determined to find the filter weights Ci, Ci+1 that minimize the cost function.

[0116] The LMS algorithm for an Mthorder filter can be summarized as:

[0117] Parameters: M = filter order

[0118] μ = step size

[0119] Initialization: C0= 0 (M)

[0120] Calculation: for n = 0, 1, 2,...

[0121] X(n) = [x(n), x(n-1),..., x(n-M+1)] T

[0122] e(n) = d(n) - C H (n) x(n)

[0123] C n+1 = C n + μ e*(n) x(n).

[0124] In this representation, X(n) corresponds to the reference signal 301, e(n) corresponds to the error signal 303 (or echo-free signal), and d(n) corresponds to the digital representation 302 of the received signal 122.

[0125] Any other algorithm can be used instead of using the LMS algorithm, such as normalized LMS, RLS, fast transversal filter algorithm, LFTF algorithm, etc.

[0126] Figure 5 A schematic representation of a method 500 for determining filter coefficients of an echo compensation filter according to the present application is shown.

[0127] The method 500 determines filter coefficients of an echo compensation filter for reducing an echo signal 123 in a communication channel 130 based on a received signal 122, as Figures 1 to 4 shown. The received signal 122 is a receivable version of a known reference signal 301 that can be transmitted on the communication channel 130. The receivable version of the reference signal 301 is affected by the echo signal 123. The impulse response 304 of the communication channel 130 extends over N time intervals 342a-d, as Figure 3 and Figure 4 shown.

[0128] The method 500 comprises filtering (501) values of the reference signal 301 by using a finite impulse response filter 311a-c, 312a-d, 314a-c having a set of M+1 filter coefficients 313a-d to obtain a first filtered reference signal 315, e.g. as Figure 3 and Figure 4 shown.

[0129] The method 500 comprises selecting 502, by the interval selector 340, a value of the reference signal 301 associated with the k-th time interval of the N time intervals 342a-d, e.g. as shown in Figure 3 and Figure 4 .

[0130] The method 500 comprises successively delaying 503, by the M delay elements 321a-c, the selected value of the reference signal 301, the M delay elements 321a-c being arranged after the interval selector 340, e.g. as shown in Figure 3 and Figure 4 .

[0131] The method 500 comprises determining 504, by the M+1 error estimators 331a-d, M+1 filter coefficients 313a-d to minimize a measure of a deviation between the first filtered reference signal 315 and the received signal 122, wherein the first error estimator 331a is coupled to the output of the interval selector 340, and wherein each of the remaining M error estimators 331b-d is coupled to the output of a respective delay element 321a-c, e.g. as shown in Figure 3 and Figure 4 .

[0132] The present application also supports a computer program product comprising computer-executable code or computer-executable instructions that, when executed, cause at least one computer to perform the execution steps and the calculation steps described herein, in particular the above-described methods and processes. The computer program product can comprise a readable non-transitory storage medium storing the program code for use by the computer. The program code can perform the processing steps and the calculation steps described herein, in particular the above-described methods and processes.

[0133] Although a particular feature or aspect of the present application can have been disclosed with respect to only one of several implementations, other features and aspects can be combined with respect to other implementations. Moreover, with respect to the use of terminology, for example, "comprising", "having", "with", or any other word(s) connoting inclusion, the use of such terminology is either intended to be representative of a possible inclusion of one or more elements or components, or exclusion of any element or component, unless otherwise indicated. Also, the use of terminology such as "example" or "exemplary" is intended to convey that a particular feature or aspect is an example and not necessarily the best or only example. The use of terms such as "coupled" or "connected" or similar terms is not intended to mean that two elements or components are directly connected or physically touch each other unless otherwise indicated. It is to be understood that the terms "coupled" or "connected" or similar terms can be used to generally refer to two or more elements or components that co-operate or interact with each other, whether or not the two elements or components are in direct physical or electrical contact with each other.

[0134] While particular aspects have been illustrated and described herein, it will be appreciated that various alternative and / or equivalent implementations can be made by those skilled in the art without departing from the scope of the present disclosure. This application is intended to cover any and all modifications or variations in the particular aspects discussed herein.

[0135] Although elements in the claims have been presented in a certain order, the order is not used as a criterion that does not affect other ways apart from the order. Unless the claim recitations imply a particular order for implementing portions or all of the elements, the elements do not have to be implemented in the particular order.

[0136] From the foregoing, it will be appreciated that numerous alternative arrangements can be devised which, although not explicitly described herein, embody the principles of the application and are intended to fall within the scope of the application. Accordingly, it will be understood that the application is not limited to the embodiments described herein, but is intended to cover any and all modifications of the application within the scope of the claims.

Claims

1. An apparatus (300) characterized by, For determining filter coefficients of an echo compensation filter for reducing an echo signal (123) in a communication channel (130) based on a received signal (122), which is a receivable version of a known reference signal (301) transmittable on the communication channel (130), which receivable version of the reference signal (301) is affected by the echo signal (123), wherein an impulse response (304) of the communication channel (130) extends over N time intervals (342a-d), the apparatus (300) comprises: a finite impulse response filter (311a-c, 312a-d, 314a-c) having a set of M+1 filter coefficients (313a-d) for filtering values of the reference signal (301) using the M+1 filter coefficients (313a-d) to obtain a first filtered reference signal (315); an interval selector (340) for selecting a value of the reference signal (301) associated with a k-th time interval of the N time intervals (342a-d); M delay elements (321a-c) for sequentially delaying the selected values of the reference signal (301), the M delay elements (321a-c) being arranged after the interval selector (340); M+1 error estimators (331a-d), wherein a first error estimator (331a) is coupled to an output of the interval selector (340), and wherein each of the remaining M error estimators (331b-d) is coupled to an output of a respective delay element (321a-c), wherein the M+1 error estimators (331a-d) are for determining the M+1 filter coefficients (313a-d) such that a measure of a deviation between the first filtered reference signal (315) and the received signal (122) is minimized.

2. The apparatus (300) according to claim 1, characterized in that the M+1 error estimators (331a-d) are mean square error estimators, in particular least mean square error estimators or least mean square estimators.

3. The apparatus (300) according to claim 1, characterized in that the measure of the deviation between the first filtered reference signal (315) and the received signal (122) is a mean square deviation between values of the first filtered reference signal (315) and values of the received signal (122).

4. The apparatus (300) according to any one of claims 1 to 3, characterized in that comprises: a memory (370) for storing the M+1 filter coefficients (313a-d) of the values of the reference signal (301) selected by the interval selector (340).

5. The apparatus (300) according to claim 4, characterized in that the memory (370) is for storing the M+1 filter coefficients (313a-d) of each time interval of the N time intervals (342a-d) applied by the interval selector (340).

6. The apparatus (300) according to claim 4, characterized in that the M+1 error estimators (331a-d) are mean square error estimators, in particular least mean square error estimators or least mean square estimators. the measure of the deviation between the first filtered reference signal (315) and the received signal (122) is a mean square deviation between values of the first filtered reference signal (315) and values of the received signal (122). comprises: a memory (370) for storing the M+1 filter coefficients (313a-d) of the values of the reference signal (301) selected by the interval selector (340). The memory (370) is configured to store the M+1 filter coefficients (313a-d) and the value of the kth time interval applied by the interval selector (340).

7. The apparatus (300) according to any one of claims 1 to 3 or 5 to 6, characterized in that The M+1 error estimators (331a-d) are configured to determine that the measure of the deviation between the first filtered reference signal (315) and the received signal (122) is below a predefined threshold for the M+1 filter coefficients (313a-d).

8. The apparatus (300) according to claim 7, characterized by comprising: a subtracter (330) configured to determine the measure of the deviation by subtracting the first filtered reference signal (315) from a digital representation (302) of the received signal (122).

9. The apparatus (300) according to any one of claims 1 to 3, 5 to 6, or 8, characterized by, comprising: an echo compensation filter having a set of N times (M+1) filter coefficients, the echo compensation filter being configured to filter a value of a transmit signal transmittable on the communication channel (130) by using the M+1 filter coefficients (313a-d) of each of the N time intervals (342a-d).

10. The apparatus (300) according to any one of claims 1 to 3, 5 to 6, or 8, characterized by, comprising: a controller (350) configured to send a selection signal (351) to the interval selector (340) for selecting the value of the kth time interval applied by the interval selector (340).

11. The apparatus (300) according to claim 10, characterized in that comprising: a plurality of spare resources (371) for increasing the number of filter coefficients of the finite impulse response filter (311a-c, 312a-d, 314a-c).

12. The apparatus (300) according to claim 11, characterized in that the spare resources (371) comprise at least one additional filter coefficient (313e, 313f); the finite impulse response filter (311a-c, 312a-d, 314a-c) is configured to filter the value of the reference signal (301) using the M+1 filter coefficients (313a-d) and the at least one additional filter coefficient (313e, 313f) to obtain the first filtered reference signal (315).

13. The apparatus (300) according to claim 11 or 12, characterized by comprising: a combination element (372) configured to combine the finite impulse response filter (311a-c, 312a-d, 314a-c) with the plurality of spare resources (371).

14. The apparatus (300) according to claim 13, characterized in that the controller (350) is configured to send an enable signal (352) to the plurality of spare resources (371) and to the combination element (372), the enable signal (352) being configured to enable the plurality of spare resources (371) and to enable the combination element (372).

15. The apparatus (300) according to any one of claims 1 to 3, 5 to 6, 8, 11 to 12, or 14, characterized by, comprising: an analog front end (381) configured to receive the received signal (122); an analog-to-digital converter (382) configured to convert the received signal (122) processed by the analog front end (381) into a digital representation (302).

16. A communication method (500), characterized by, For determining filter coefficients of an echo compensation filter for reducing an echo signal (123) in a communication channel (130) based on a received signal (122), which is a receivable version of a known reference signal (301) transmittable on the communication channel (130), which receivable version of the reference signal (301) is affected by the echo signal (123), wherein an impulse response (304) of the communication channel (130) extends over N time intervals (342a-d), the method (500) comprises: filtering (501) values of the reference signal (301) by using a finite impulse response filter (311a-c, 312a-d, 314a-c) having a set of M+1 filter coefficients (313a-d) to obtain a first filtered reference signal (315); selecting (502), by an interval selector (340), a value of the reference signal (301) associated with a k-th time interval of the N time intervals (342a-d); delaying (503), by M delay elements (321a-c), the selected value of the reference signal (301) in sequence, the M delay elements (321a-c) being arranged after the interval selector (340); determining (504), by M+1 error estimators (331a-d), the M+1 filter coefficients (313a-d) to minimize a measure of a deviation between the first filtered reference signal (315) and the received signal (122), wherein a first error estimator (331a) is coupled to an output of the interval selector (340), and wherein each of the remaining M error estimators (331b-d) is coupled to an output of a respective delay element (321a-c).

Images