Decision feedback equalizer and related control method

Abstract

Embodiments of the present disclosure relate to decision feedback equalizers and related control methods. A decision feedback equalizer, for generating a decision output signal based on an input signal, comprises: a feedforward equalizer, a feedback equalizer, and a weight coefficient control unit. The feedforward equalizer comprises a plurality of tapped delay lines and is controlled by a set of first weight coefficients. The feedback equalizer comprises a plurality of tapped delay lines and is controlled by a set of second weight coefficients. The weight coefficient control unit is used to selectively adjust at least one of the set of first weight coefficients and to determine a set of first boundary values ​​for at least one of the set of second weight coefficients. When at least one of the set of second weight coefficients does not exceed the set of first boundary values, the weight coefficient control unit increments at least one of the set of first weight coefficients.

Description

Technical Field

[0001] The present invention relates to a decision feedback equalizer, and in particular to a decision feedback equalizer and a related control method for avoiding error propagation by periodically adjusting the weight coefficient of the equalizer. Background Art

[0002] When signals are transmitted through a channel, they are often distorted by the channel's time dispersion effect. This is primarily due to the fact that when the channel's frequency response exhibits non-constant amplitude and nonlinear phase, the signal's amplitude and phase are distorted by the channel response, leading to inter-symbol interference (ISI), which prevents the receiver from correctly identifying the signal.

[0003] A decision feedback equalizer (DFE) can be used to eliminate the aforementioned channel distortion. The DFE comprises a feed-forward equalizer, a decision unit, and a feedback equalizer. The feed-forward equalizer, also known as a zero-forcing equalizer, ensures that the impulse responses of the channel and the equalizer, after convolution, have a single value. The advantage of the feed-forward equalizer lies in its simple architecture, but its disadvantage is that it amplifies noise, leading to decision errors. Therefore, a feedback equalizer is needed to address this issue. The feedback equalizer uses the detected symbol d0 as input, assuming that the symbol detected by the DFE is correct, thereby eliminating channel ISI. Therefore, the feedback equalizer does not amplify noise.

[0004] Generally speaking, decision feedback equalizers (DFEQs) rely on the least mean square (LMS) algorithm to determine the weight coefficients for each tapped delay line in the feedforward and feedback equalizers. By repeatedly adjusting the weight coefficients of the feedforward and feedback equalizers, the algorithm approaches the least squares solution found by the LMS algorithm, thereby converging to a high signal-to-noise ratio. However, in certain situations, when a decision unit in a DFEQ makes a decision error, this error is input into the feedback equalizer. This decision error is also fed back to the entire DFEQ via the output of the feedback equalizer. If the error is large, it can cause a large error loop between the feedback equalizer and the decision unit, a condition known as error propagation. When the error propagation is severe, the entire DFEQ system will fail to converge, leading to a systemic failure. Summary of the Invention

[0005] In order to avoid the occurrence of error propagation, the present invention provides a mechanism for controlling a decision feedback equalizer. In the control mechanism of the present invention, the weight coefficients in the feedback equalizer are limited in the early stage of convergence of the least mean square algorithm, thereby limiting the weight energy of the feedback equalizer. In this way, it is possible to avoid the error propagation caused by the error being amplified when a decision error occurs, thereby preventing the least mean square algorithm from converging. Moreover, the decision error in the early stage of convergence is more significant than when the convergence is stable, so suppressing the weight energy of the feedback equalizer helps to improve the stability in the early stage of convergence. Furthermore, when the least mean square algorithm tends to converge stably, the control mechanism of the present invention relaxes the restrictions on the weight coefficients of the feedback equalizer, thereby gradually allowing the weight coefficients to approach the least square solution in the stage of stable convergence, thereby improving the signal-to-noise ratio of the signal. In this way, the stability of the least mean square algorithm and a good signal-to-noise ratio can be taken into account.

[0006] One embodiment of the present invention provides a decision feedback equalizer, which is used to generate a decision output signal based on an input signal, and includes: a feedforward equalizer, a feedback equalizer, and a weight coefficient control unit. The feedforward equalizer includes a plurality of tapped delay lines and is controlled by a set of first weight coefficients. The feedback equalizer includes a plurality of tapped delay lines and is controlled by a set of second weight coefficients. The weight coefficient control unit is used to selectively adjust at least one of the set of first weight coefficients and determine a set of first boundary values ​​of at least one of the set of second weight coefficients. When at least one of the set of second weight coefficients does not exceed the set of first boundary values, the weight coefficient control unit increments at least one of the set of first weight coefficients.

[0007] One embodiment of the present invention provides a method for controlling a decision feedback equalizer. The method is used to generate a decision output signal based on an input signal, wherein the decision feedback equalizer has a feedforward equalizer and a feedback equalizer. The method includes: selectively adjusting at least one of a set of first weight coefficients corresponding to a plurality of tap delay lines of the feedforward equalizer; determining a set of first boundary values ​​for at least one of a set of second weight coefficients corresponding to a plurality of tap delay lines of the feedback equalizer; and when at least one of the set of second weight coefficients does not exceed the set of first boundary values, the weight coefficient control unit increments at least one of the set of first weight coefficients. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 FIG. 4 is a functional block diagram of an embodiment of a decision feedback equalizer according to the present invention.

[0009] Figure 2A and Figure 2B Schematic diagrams of the implementation architectures of the feedforward equalizer and the feedback equalizer in the decision feedback equalizer of the present invention are respectively shown.

[0010] Figure 3 Flowchart of the preliminary adjustment stage in an embodiment of the control method of the present invention.

[0011] Figure 4 Flowchart of the fine-tuning stage in an embodiment of the control method of the present invention.

[0012] Figure 5 A simplified flow chart of an embodiment of the control method of the present invention. DETAILED DESCRIPTION

[0013] In the following description, numerous specific details are described to provide a thorough understanding of the embodiments of the present invention. However, those skilled in the art will understand how to implement the present invention without one or more of these specific details, or using other methods, components, or materials. In other cases, well-known structures, materials, or operations are not shown or described in detail to avoid obscuring the core concepts of the present invention.

[0014] References in this specification to "one embodiment" mean that the specific features, structures, or characteristics described in that embodiment may be included in at least one embodiment of the present invention. Therefore, the phrase "in one embodiment" appearing throughout this specification does not necessarily refer to the same embodiment. Furthermore, the aforementioned specific features, structures, or characteristics may be combined in any suitable manner in one or more embodiments.

[0015] Please refer to Figure 1, which is a functional block diagram of an embodiment of a decision feedback equalizer according to the present invention. As shown, the decision feedback equalizer 100 includes a feedforward equalizer 110, a feedback equalizer 120, a decision unit 130, and a weight coefficient control unit 140. The decision feedback equalizer 100 is configured to receive an input signal r0 from a channel and generate a decision output signal d0. The feedforward equalizer 110 generates a feedforward output signal FF0 based on the input signal r0, thereby eliminating pre-cursor interference and post-cursor interference in ISI caused by the channel. The feedback equalizer 120 generates a feedforward output signal FB0 based on the decision output signal d0, thereby eliminating post-cursor interference in ISI. The decision unit 130 generates the decision output signal d0 based on the feedforward output signal FF0 generated by the feedforward equalizer 110 and the feedforward output signal FB0 generated by the feedback equalizer 120.

[0016] Figure 2A and Figure 2B The schematic diagrams of the structures of the feedforward equalizer 110 and the feedback equalizer 120 are shown respectively. The feedforward equalizer 110 and the feedback equalizer 120 respectively have a plurality of tapped delay lines 112_1 to 112_N and 122_1 to 122_N. The feedforward equalizer 110 and the feedback equalizer 120 are respectively controlled by a set of weight coefficients cf1 to cf2. N And a set of weight coefficients cb1~cb N The weight coefficient control unit 140 can control the weight coefficients cf1 to cf2 in the process of convergence of the decision feedback equalizer 100. N and cb1~cb N In the control mechanism of the present invention, the weight coefficient control unit 140 performs weight coefficients cf1 to cf2 on the convergence process of the decision feedback equalizer 100. N and cb1~cb N In the initial convergence stage, the weight coefficient control unit 140 of the present invention conservatively adjusts the weight coefficients cf1 to cf2 according to the least mean square algorithm and certain restrictions. N and cb1~cb N , in order to improve the system signal-to-noise ratio while ensuring the stability of convergence. When entering the convergence stability stage, the weight coefficient control unit 140 adjusts one or more weight coefficients cf in the feedforward equalizer 110 K Make adjustments and observe one or more weight coefficients cb in the feedback equalizer 120 K The corresponding changes, based on which the weight coefficient cf is gradually K And the weight coefficient cbK Approximate the least square solution obtained by the least mean square algorithm. In the early stage of convergence, the present invention will K With cb K Make certain restrictions, which makes the weight coefficient cf K The adjustable range is small. In the convergence and stability stage, the weight coefficient cb will be relaxed. K The restriction of weight coefficient cf K There is a larger adjustment range, so that the weight coefficient cf K With cb K In the convergence and stable stage, it can be closer to the least square solution to improve the signal-to-noise ratio of the system.

[0017] Please refer to Figure 3 , which is a flow chart of the preliminary adjustment phase of the control method embodiment of the present invention. First, in step 310, the decision feedback equalizer is still in the early stage of convergence, so for one or more weight coefficients cb in the feedback equalizer 110 K With cf K Limit and set each one or more weight coefficients cb K Its corresponding upper boundary value cb_max K and the lower boundary value cb_min K , requiring each one or more weight coefficients cb K Must not exceed the corresponding upper boundary value cb_max K and the lower boundary value cb_min K , and set each one or more weight coefficients cf K Its corresponding upper boundary value cf_max K and the lower boundary value cf_min K , requiring each one or more weight coefficients cf K Must not exceed the corresponding upper boundary value cf_max K and the lower boundary value cf_min K In different embodiments, different weight coefficients cb K The corresponding upper / lower boundary value cb_max K and cb_min K May be the same or different, and with different weighting coefficients cf K The corresponding upper / lower boundary value cf_max K and cf_min K They may also be the same or different. However, the upper boundary value cb_max K and the lower boundary value cb_min K The absolute value of must be smaller than the absolute value of the least square solution (MMSE solution) to avoid error propagation.

[0018] Furthermore, in step 320, the least mean square algorithm is used to gradually bring the decision feedback equalizer into a convergence state, which includes one or more weight coefficients cf in the feedforward equalizer 120. K and one or more weight coefficients cb in the feedback equalizer 110 K and ensure that one or more weight coefficients cb K Will not exceed the corresponding upper / lower boundary value cb_max K and cb_min K , one or more weight coefficients cf K Will not exceed the corresponding upper / lower boundary value cf_max K and cf_min K .

[0019] In step 330, a check is performed to determine whether a predetermined time period T_avg has elapsed and the signal-to-noise ratio (SNR) of the decision feedback equalizer 100 meets a minimum convergence stability requirement. The minimum convergence stability requirement can be determined by calculating the SNR of the decision feedback equalizer 100 and comparing it with a threshold value SNR_stable. The threshold value SNR_stable can be determined by repeatedly testing the decision feedback equalizer 100 in advance based on the SNR when error propagation occurs.

[0020] If the decision feedback equalizer 100 has been ensured to enter a stable convergence state in the above stages, the control mechanism of the present invention will enter the fine-tuning stage. At this time, the control mechanism of the present invention will relax the control of one or more weight coefficients cb in the feedback equalizer 120. K limitations in order to pursue a better signal-to-noise ratio.

[0021] Please refer to Figure 4 The flowchart shown is a flowchart of the fine-tuning phase of the control method embodiment of the present invention. First, in step 410, the control mechanism of the present invention will first check the current system signal-to-noise ratio. If the system signal-to-noise ratio is high enough, the fine-tuning phase will be terminated to avoid the weight coefficient cb of the feedback equalizer 120 K If the system SNR is too high, there may be a risk of error propagation. However, if there is still room for improvement in the system SNR, the fine-tuning phase will proceed. The system SNR will be compared with a target SNR_target. The fine-tuning process will only continue if the system SNR is less than the target SNR_target.

[0022] In step 420, the weight coefficient control unit 140 relaxes the weight coefficient cb KThe limitation is to make one or more weight coefficients cb in the feedback equalizer 120 K Compared to the initial adjustment stage, the least square solution obtained by the least mean square algorithm can be closer. The weight coefficient control unit 140 sets each one or more weight coefficients cb K The corresponding new upper boundary value cb_max_tgt K and the new lower boundary value cb_min_tgt K , and requires that each one or more weight coefficients cb K Must not exceed the corresponding upper limit value cb_max_tgt K and the lower boundary value cb_min_tgt K The boundary value cb_max_tgt set in the fine-tuning stage K and the lower boundary value cb_min_tgt K , will be better than Figure 3 The boundary value cb_max at the initial stage of convergence is shown K and the lower boundary value cb_min K By relaxing one or more weight coefficients cb K The restriction that one or more weight coefficients cb K It can be closer to the least square solution and further improve the system signal-to-noise ratio.

[0023] In step 430, the weight coefficient control unit 140 checks one or more weight coefficients cb in the feedback equalizer 120. K Does it not exceed the boundary value set in step 420? K If the boundary value set in step 420 is exceeded, the fine-tuning phase (step 480) will also end. K If the boundary value is still not exceeded, the weight coefficient control unit 140 will gradually increase one or more weight coefficients cf in the feedforward equalizer 110. K Since the weight coefficient cb K and weight coefficient cf K There is a certain dependence between them, so the weight coefficient cb K It will also increase accordingly.

[0024] In the control mechanism of the present invention, the weight coefficient control unit 140 will wait for a predetermined time T_thd before increasing the weight coefficient cf K (Step 440). When the predetermined time T_thd has passed, the process will proceed to step 450 to adjust the weight coefficient cf K If not, proceed to step 460 and maintain the weight coefficient cf K No change, and continue waiting (step 465).

[0025] In the fine-tuning phase, each time the weight coefficient cf is increased K After that, it will wait for a predetermined period of time before increasing the weight coefficient cf again. K The purpose of designing the waiting time is to ensure the stability of the system. K , which may result in transient errors. When the errors are too large, they may also undermine the convergence stability of the decision feedback equalizer 100. Therefore, slow adjustment helps to avoid such errors. On the other hand, the length of the waiting time may vary depending on the characteristics of the system. In extreme cases, it may be possible to consider ignoring the waiting time or using a very long waiting time.

[0026] In step 450, the weight coefficient control unit 140 increases one or more weight coefficients cf in the feedforward equalizer by an equal amount (eg, a step amount cf_step). K , and their corresponding upper / lower boundary values ​​cf_max K and cf_min K For example, each time step 450 is entered, the weight coefficient cf K and upper / lower boundary values ​​cf_max K and cf_min K Increase the step size cf_step by one unit. The selection of the step size cf_step is also related to the system characteristics. Afterwards, the control mechanism of the present invention will adjust the weight coefficient cf K And their corresponding upper / lower boundary values ​​cf_max K and cf_min K After adjustment, reconfirm one or more weight coefficients cb K Whether the corresponding boundary value cb_max_tgt is reached K and cb_min_tgt K (Return to step 430). If yes, the process of fine-tuning phase ends; otherwise, continue to adjust the weight coefficient cf K , and its upper / lower boundary value cf_max K and cf_min K Make the next adjustment.

[0027] Please note that in some embodiments, the control flow of the present invention may also only include Figure 4 The fine-tuning phase shown, that is, once the decision feedback equalizer 100 is confirmed to have entered the stable convergence phase, the fine-tuning phase is launched, but the previous control adjustment mechanism may be different from Figure 3 The control method of the present invention can also be simplified as follows: Figure 5 Steps shown:

[0028] Step 510: Selectively adjust at least one of a set of first weight coefficients corresponding to a plurality of tap delay lines of a feedforward equalizer;

[0029] Step 520: Determine a set of first boundary values ​​for at least one of a set of second weight coefficients corresponding to a plurality of tap delay lines of the feedback equalizer; and

[0030] Step 530: When at least one of the set of second weight coefficients does not exceed the set of first boundary values, the weight coefficient control unit increments at least one of the set of first weight coefficients.

[0031] The above steps can be modified and adjusted appropriately to obtain the same / similar processes and effects as the above embodiments.

[0032] The present invention has the following features: First, in the early stage of convergence using the least mean square algorithm, the weight coefficient cb of the feedback equalizer 120 K A strict restriction is imposed to avoid being too close to the least square solution, thereby limiting the energy of the weight of the feedback equalizer 120. In this way, the occurrence of error propagation can be avoided. Only after converging into a stable state, the weight coefficient cb K The restriction of will be relaxed, so that it can approach the least square solution, so that the decision feedback equalizer 100 can further improve the signal-to-noise ratio to an ideal target. Moreover, after entering the stable state, the weight coefficient cf is adjusted more actively. K (Regularly increasing the fixed step size) allows the system to improve its signal-to-noise ratio more quickly, while at the same time, by setting the waiting time, it avoids transient errors that occur while improving its signal-to-noise ratio. In this way, the present invention takes into account the stability of the least mean square algorithm and a good signal-to-noise ratio.

[0033] The embodiments of the present invention may be implemented using hardware, software, firmware, and combinations thereof. Software or firmware stored in a memory may be used to implement the embodiments of the present invention through an appropriate instruction execution system. As for hardware, any of the following technologies or combinations thereof may be used: separate arithmetic logic having logic gates that can perform logic functions based on data signals, an application-specific integrated circuit (ASIC) having suitable combinational logic gates, a programmable gate array (PGA), or a field programmable gate array (FPGA).

[0034] The processes and blocks in the flowcharts within this specification illustrate the architecture, functionality, and operations that can be implemented by the systems, methods, and computer software products according to various embodiments of the present invention. In this regard, each block in a flowchart or functional block diagram may represent a module, segment, or portion of program code, comprising one or more executable instructions for implementing a specified logical function. Furthermore, each block in the functional block diagram and / or flowchart, as well as combinations of blocks, may be implemented substantially by a dedicated hardware system that performs the specified function or action, or by a combination of dedicated hardware and computer program instructions. These computer program instructions may also be stored on a computer-readable medium that can cause a computer or other programmable data processing device to operate in a specific manner such that the instructions stored on the computer-readable medium implement the functions / actions specified by the blocks in the flowchart and / or functional block diagram. The foregoing description is merely a preferred embodiment of the present invention, and all equivalent variations and modifications made within the scope of the present invention are intended to be covered by the present invention.

[0035]

Explanation of symbols

[0036] 100 Decision Feedback Equalizer

[0037] 110 Feedforward Equalizer

[0038] 120 Feedback EQ

[0039] 130 Decision-making Unit

[0040] 140 Weight Control Unit

[0041] 112_1~112_N, 122_1~122_N tapped delay lines

[0042] Steps 310-340, 410-480

Claims

1. A decision feedback equalizer for generating a decision output signal based on an input signal, comprising: A feedforward equalizer comprising a plurality of tapped delay lines and controlled by a set of first weight coefficients; A feedback equalizer comprising a plurality of tapped delay lines and controlled by a set of second weight coefficients; as well as A weight coefficient control unit is used to selectively adjust at least one of the set of first weight coefficients and determine a set of first boundary values ​​for at least one of the set of second weight coefficients; when at least one of the set of second weight coefficients does not exceed the set of first boundary values, the weight coefficient control unit increases at least one of the set of first weight coefficients, and at least one of the set of second weight coefficients increases as at least one of the set of first weight coefficients increases.

2. A decision feedback equalizer according to claim 1, wherein when at least one of the set of second weight coefficients is equal to the set of first boundary values, the weight coefficient control unit no longer increments at least one of the set of first weight coefficients. 3 . The decision feedback equalizer according to claim 1 , wherein the weight coefficient control unit has a predetermined time interval between two consecutive processes of increasing at least one of the set of first weight coefficients.

4. A decision feedback equalizer according to claim 1, wherein when at least one of the set of second weight coefficients does not exceed the set of first boundary values, and the weight coefficient control unit increases the set of second boundary values ​​corresponding to at least one of the set of first weight coefficients while increasing at least one of the set of first weight coefficients.

5. The decision feedback equalizer according to claim 1, wherein before the weight coefficient control unit adjusts the set of first weight coefficients, if the signal-to-noise ratio of the decision output signal is greater than or equal to the target signal-to-noise ratio, the weight coefficient control unit will not adjust any one of the set of first weight coefficients.

6. The decision feedback equalizer according to claim 4, wherein the weight coefficient control unit further determines a set of third boundary values ​​for at least one of the set of second weight coefficients, wherein the set of third boundary values ​​is smaller than the set of first boundary values, and the set of second boundary values ​​is smaller than a least square solution obtained based on a least mean square algorithm; wherein, When the signal-to-noise ratio of the decision output signal is less than the stable signal-to-noise ratio, the weight coefficient control unit adjusts at least one of the set of first weight coefficients and at least one of the set of second weight coefficients according to the least mean square algorithm until the signal-to-noise ratio is no less than the stable signal-to-noise ratio.

7. A method for controlling a decision feedback equalizer, for generating a decision output signal according to an input signal, wherein the decision feedback equalizer comprises a feedforward equalizer and a feedback equalizer, the method comprising: selectively adjusting at least one of a set of first weight coefficients corresponding to a plurality of tap delay lines of the feedforward equalizer; determining a set of first boundary values ​​corresponding to at least one of a set of second weight coefficients for a plurality of tap delay lines of the feedback equalizer; as well as When at least one of the set of second weight coefficients does not exceed the set of first boundary values, the weight coefficient control unit increases at least one of the set of first weight coefficients, and at least one of the set of second weight coefficients increases as at least one of the set of first weight coefficients increases.

8. The method of claim 7, wherein selectively adjusting at least one of the set of first weight coefficients comprises: When at least one of the set of second weight coefficients is equal to the set of first boundary values, at least one of the set of first weight coefficients is no longer incremented.

9. The method of claim 7, wherein selectively adjusting at least one of the set of first weight coefficients comprises: A predetermined time interval may elapse between two consecutive increments of at least one of the set of first weight coefficients.

10. The method of claim 7, wherein selectively adjusting at least one of the set of first weight coefficients comprises: When at least one of the set of second weight coefficients does not exceed the first boundary value, and the weight coefficient control unit increases at least one of the set of first weight coefficients while also increasing a set of second boundary values ​​corresponding to at least one of the set of first weight coefficients.

Images