A channel estimation and adaptive equalization device

By using a fully digital channel estimation and adaptive equalization device, the problem that traditional devices cannot adapt to changes in signal modulation schemes and signal-to-noise ratio estimation is solved. It achieves efficient and stable channel estimation and adaptive equalization, adapts to multiple modulation schemes, and reduces system complexity.

CN115865578BActive Publication Date: 2024-11-05THE 54TH RESEARCH INSTITUTE OF CHINA ELECTRONICS TECHNOLOGY GROUP CORPORATION
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
CN202211533786.4
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-12-02
Publication Date
2024-11-05
Estimated Expiration
2042-12-02

AI Technical Summary

Technical Problem

Traditional channel estimation and adaptive equalization devices cannot adapt to changes in signal modulation schemes and lack signal-to-noise ratio estimation, resulting in high system complexity and low convergence speed and stability.

Method used

It employs a digital AGC, a split-delay module, a signal-to-noise ratio estimation module, a parallel convolution operation module, a convolution coefficient generation module, a data extraction module, an error extraction module, and an FFT interpolation module, combined with a modulation scheme indicator signal, to achieve fully digital channel estimation and adaptive equalization, adapting to multiple modulation schemes, and performing signal-to-noise ratio estimation and error weighting processing.

Benefits of technology

It achieves highly reliable and low-complexity channel estimation and adaptive equalization, improves system processing efficiency and stability, adapts to various modulation schemes, and reduces system complexity.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115865578B_ABST
    Figure CN115865578B_ABST
Patent Text Reader

Abstract

The present invention discloses a channel estimation and adaptive equalization device, which includes a digital AGC, a branch delay module, a parallel convolution operation module, a convolution coefficient generation module, a signal-to-noise ratio estimation module, a data extraction module, an error extraction module and an FFT interpolation module. The present invention adopts a fully digital implementation scheme, and simultaneously realizes multiple functions such as signal-to-noise ratio estimation, channel estimation and signal adaptive equalization, thereby improving system processing efficiency and reducing system complexity.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the field of communication and data transmission, and is particularly suitable for use as channel estimation and adaptive equalization in a data receiving subsystem of a satellite communication and a data ground receiving station. Background Art

[0002] Traditional channel estimation and adaptive equalization devices complete signal adaptive equalization for a specific modulation system and cannot adapt to changes in signal modulation systems. At the same time, traditional channel estimation and adaptive equalization devices output adaptive equalization coefficients as channel estimation parameters, but lack the estimated output of the received signal signal-to-noise ratio, requiring other modules in the system to implement this function, increasing the complexity of the system. In addition, traditional channel estimation and adaptive equalization devices usually adopt three methods when calculating error values:

[0003] 1) Extracting errors through hard decision. This method is simple to implement, but due to the influence of factors such as noise, the extracted errors cannot be well utilized, which reduces the convergence speed and convergence stability of the device;

[0004] 2) By transmitting pilot signals, the error can be extracted without error, but it requires additional transmission resources and reduces the system transmission utilization;

[0005] 3) By cascading with channel decoding, high-precision error extraction is achieved, but the device system is complex and has a large system delay, which leads to a smaller system feedback loop bandwidth and reduces system stability. Summary of the invention

[0006] The purpose of the present invention is to avoid the deficiencies in the above-mentioned background technology and provide a channel estimation and adaptive equalization device. The present invention has the characteristics of high reliability, high stability, low complexity and the like.

[0007] The object of the present invention is achieved in that:

[0008] A channel estimation and adaptive equalization device, comprising a digital AGC 1, a branch delay module 2, a signal-to-noise ratio estimation module 3, a parallel convolution operation module 4, a convolution coefficient generation module 5, a data extraction module 6, an error extraction module 7 and an FFT interpolation module 8;

[0009] The digital AGC1 performs signal power detection and power adjustment on the input I / Q two-way digital baseband signal, transmits the signal power detection signal to the signal-to-noise ratio estimation module 3, and transmits the power-adjusted I / Q two-way digital baseband signal to the branch delay module 2;

[0010] The branch delay module 2 branches and independently delays the received I / Q two-way digital baseband signals to generate two groups of I / Q two-way digital baseband signals, wherein the first group of I / Q two-way digital baseband signals is transmitted to the parallel convolution operation module 4, and the second group of I / Q two-way digital baseband signals is transmitted to the convolution coefficient generation module 5;

[0011] The parallel convolution operation module 4 receives the first group of I / Q two-way convolution coefficients transmitted by the convolution coefficient generation module 5, and performs convolution operation on the input I / Q two-way digital baseband signals, generates the I / Q two-way digital baseband signals after operation processing, and transmits them to the data extraction module 6;

[0012] The data extraction module 6 extracts the peak points of the input I / Q two-way digital baseband signals and outputs them respectively. The I / Q two-way digital baseband signals output by the data extraction module 6 are further divided to generate two groups of I / Q two-way digital baseband signals, wherein the first group of I / Q two-way digital baseband signals is output as the device, and the output signal is the equalized I / Q two-way digital baseband signal, and the second group of I / Q two-way digital baseband signals is transmitted to the error extraction module 7;

[0013] The error extraction module 7 processes the input I / Q two-way digital baseband signal according to the input modulation system indication signal, generates and outputs an I / Q two-way equalization error signal, and the I / Q two-way equalization error signal output by the error extraction module is further divided to generate two groups of I / Q two-way equalization error signals, the first group of I / Q two-way equalization error signals is transmitted to the signal-to-noise ratio estimation module 3, and the second group of I / Q two-way equalization error signals is transmitted to the convolution coefficient generation module 5;

[0014] The signal-to-noise ratio estimation module 3 receives the signal power detection signal transmitted by the digital AGC1 and the I / Q two-way equalization error signal output by the error extraction module 7, generates a signal-to-noise ratio estimation value through calculation and outputs it as the signal-to-noise ratio estimation signal of the device;

[0015] The convolution coefficient generating module 5 receives the second group of I / Q two-way digital baseband signals output by the branch delay module 2 and the second group of I / Q two-way equalization error signals output by the error extraction module 7, generates I / Q two-way convolution coefficient signals by calculation and outputs them;

[0016] The I / Q two-way convolution coefficient signal output by the convolution coefficient generating module 5 is further divided into two groups of I / Q two-way convolution coefficient signals, the first group of I / Q two-way convolution coefficient signals is transmitted to the parallel convolution operation module 4, and the second group of I / Q two-way convolution coefficient signals is transmitted to the FFT interpolation module 8;

[0017] The FFT interpolation module 8 performs interpolation and FFT processing on the input I / Q two-way convolution coefficient signals and outputs them as the channel estimation signal of the device.

[0018] Further, the error extraction module 7 includes a module value calculation module 9, a module value accumulation module 10, a module value comparison module 11, an error calculation benchmark selection module 12, an error weighting coefficient generation module 13 and an error calculation module 14;

[0019] The I / Q two-way digital baseband signal output by the data extraction module 6 is split to generate two groups of I / Q two-way digital baseband signals, and the first group of I / Q two-way digital baseband signals is transmitted to the module value calculation module 9;

[0020] The modulus value calculation module 9 selects the I / Q data mapping values ​​of all constellation points in the standard constellation diagram of the current modulation system according to the input modulation system indication signal, and performs complex subtraction on the I / Q data mapping values ​​of each constellation point and the input I / Q two-way digital baseband signal, and performs modulus value operation on the calculated values ​​after the complex subtraction to generate a group of modulus value data signals; the modulus value calculation module 9 outputs each group of modulus value data signals generated in real time;

[0021] The module value data signal output by the module value calculation module 9 is divided into two modules of module value data signals. The first module value data signal is transmitted to the module value accumulation module 10. The module value accumulation module 10 accumulates each group of module value data values ​​in the received module value data signal, and transmits the module value accumulation signal generated after the accumulation to the error weight coefficient generation module 13.

[0022] The second module value data signal output by the module value calculation module 9 is transmitted to the module value comparison module 11, and the module value comparison module 11 performs numerical comparison processing on each group of module value data in the received module value data signal, and transmits the minimum module value data signal obtained by the numerical comparison to the error weighting coefficient generation module 13; at the same time, the module value comparison module 11 generates a minimum module value data state indication signal, and transmits the minimum module value data state indication signal to the error calculation benchmark selection module 12;

[0023] The error weight coefficient generation module 13 performs complex subtraction on the received module value accumulation signal and the minimum module value data signal, and uses the received module value accumulation signal for normalization processing, and finally generates an error weight coefficient signal and transmits it to the error calculation module 14;

[0024] The error calculation reference selection module 12 selects the I / Q data mapping values ​​of all constellation points in the current modulation system standard constellation diagram according to the received modulation system indication signal, and selects the constellation point I / Q data mapping value corresponding to the current minimum modulus data according to the received minimum modulus data state indication signal, and transmits the constellation point I / Q data mapping value to the error calculation module 14;

[0025] The error calculation module 14 receives the I / Q two-way digital baseband signal output by the data extraction module 6, the constellation point I / Q data mapping value output by the error calculation reference selection module 12 and the error weighting coefficient signal output by the error weighting coefficient generation module 13, and performs complex subtraction operation on the received I / Q two-way digital baseband signal and the constellation point I / Q data mapping value, and performs weighted processing on the complex subtraction operation result with the received error weighting coefficient signal, and finally obtains the I / Q two-way equalized error signal, and transmits it to the signal-to-noise ratio estimation module 3 and the convolution coefficient generation module 4 respectively after branching processing.

[0026] Compared with the background technology, the present invention has the following advantages:

[0027] 1. Compared with the traditional channel estimation and adaptive equalization device, the present invention does not need to transmit pilot signals and adopts a fully digital implementation scheme, while realizing multiple functions such as signal-to-noise ratio estimation, channel estimation and signal adaptive equalization, thereby improving the system processing efficiency and reducing the complexity of the system.

[0028] 2. When performing equalization error extraction, the present invention not only extracts the current equalization error value, but also calculates the correct probability information of the current equalization error value, and uses the correct probability information of the current equalization error value to perform weighted processing on the current equalization error value before performing subsequent processing, thereby improving the convergence speed and stability of the device. In addition, since a modulation system indication signal is introduced into the error extraction module, it adapts to a variety of modulation systems, thereby improving the applicability of the device.

[0029] 3. The present invention adopts a fully digital processing flow, has a simple structure, and is highly portable. Since FPGA integrates units such as subtractors, numerical comparators, multipliers, and adders, it is easy to implement and promote. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a principle block diagram of the present invention.

[0031] Figure 2 It is a circuit schematic diagram of the error extraction module of the present invention. DETAILED DESCRIPTION

[0032] Reference Figure 1 A channel estimation and adaptive equalization device includes a digital AGC 1, a branch delay module 2, a signal-to-noise ratio estimation module 3, a parallel convolution operation module 4, a convolution coefficient generation module 5, a data extraction module 6, an error extraction module 7 and an FFT interpolation module 8.

[0033] The digital AGC1 performs signal power detection and power adjustment on the input I / Q two-way digital baseband signal, transmits the signal power detection signal to the signal-to-noise ratio estimation module 3, and transmits the power-adjusted I / Q two-way digital baseband signal to the branch delay module 2;

[0034] The branching delay module 2 generates two groups of I / Q two-way digital baseband signals after branching and independent delay processing on the received I / Q two-way digital baseband signals, wherein the first group of I / Q two-way digital baseband signals is transmitted to the parallel convolution operation module 4;

[0035] The parallel convolution operation module 4 receives the first group of I / Q two-way convolution coefficients transmitted by the convolution coefficient generation module, and performs convolution operation on the input I / Q two-way digital baseband signals, generates the I / Q two-way digital baseband signals after operation processing, and transmits them to the data extraction module 6;

[0036] The data extraction module 6 extracts the peak points of the input I / Q two-way digital baseband signals and outputs them;

[0037] The I / Q two-way digital baseband signal output by the data extraction module 6 is further divided to generate two groups of I / Q two-way digital baseband signals, wherein the first group of I / Q two-way digital baseband signals is output by the device, and the output signal is the equalized I / Q two-way digital baseband signal; the second group of I / Q two-way digital baseband signals is transmitted to the error extraction module 7;

[0038] The error extraction module 7 processes the input I / Q two-way digital baseband signal according to the input modulation system indication signal, generates and outputs an I / Q two-way equalization error signal, and the I / Q two-way equalization error signal output by the error extraction module is further divided to generate two groups of I / Q two-way equalization error signals, the first group of I / Q two-way equalization error signals is transmitted to the signal-to-noise ratio estimation module 3, and the second group of I / Q two-way equalization error signals is transmitted to the convolution coefficient generation module 5;

[0039] The signal-to-noise ratio estimation module 3 receives the signal power detection signal transmitted by the digital AGC1 and the I / Q two-way equalization error signal output by the error extraction module 7, generates a signal-to-noise ratio estimation value through calculation and outputs it as the signal-to-noise ratio estimation signal of the device;

[0040] The convolution coefficient generating module 5 receives the second group of I / Q two-way digital baseband signals output by the branch delay module 2 and the second group of I / Q two-way equalization error signals output by the error extraction module 7, generates I / Q two-way convolution coefficient signals by calculation and outputs them;

[0041] The I / Q two-way convolution coefficient signal output by the convolution coefficient generating module 5 is further divided into two groups of I / Q two-way convolution coefficient signals, the first group of I / Q two-way convolution coefficient signals is transmitted to the parallel convolution operation module 4, and the second group of I / Q two-way convolution coefficient signals is transmitted to the FFT interpolation module 8;

[0042] The FFT interpolation module 8 performs interpolation and FFT processing on the input I / Q two-way convolution coefficient signals and outputs them as the channel estimation signal of the device.

[0043] Through the above processing, the channel estimation and adaptive equalization processing of the input signal are completed.

[0044] In the embodiment, the digital AGC 1, the branch delay module 2, the signal-to-noise ratio estimation module 3, the parallel convolution operation module 4, the convolution coefficient generation module 5, the data extraction module 6, the error extraction module 7 and the FFT interpolation module 8 are all implemented using Virtex 7 FPGA produced by XILINX Corporation of the United States.

[0045] Figure 2 It is a principle block diagram of the error extraction module 7. The error extraction module 7 includes a module value calculation module 9, a module value accumulation module 10, a module value comparison module 11, an error calculation benchmark selection module 12, an error weighting coefficient generation module 13 and an error calculation module 14.

[0046] The I / Q two-way digital baseband signal output by the data extraction module 6 is split to generate two groups of I / Q two-way digital baseband signals, and the first group of I / Q two-way digital baseband signals is transmitted to the module value calculation module 9;

[0047] The modulus value calculation module 9 selects the I / Q data mapping values ​​of all constellation points in the current modulation system standard constellation diagram according to the input modulation system indication signal, and performs complex subtraction on the I / Q data mapping values ​​of each constellation point and the input I / Q two-way digital baseband signal, and performs modulus value operation on the calculated values ​​after the complex subtraction, and generates a group of modulus value data signals. The modulus value calculation module 9 outputs each group of modulus value data signals generated in real time;

[0048] The module value data signal output by the module value calculation module 9 is divided into two modules of module value data signals, and the first module value data signal is transmitted to the module value accumulation module 10; the module value accumulation module 10 accumulates each group of module value data values ​​in the received module value data signal, and transmits the module value accumulation signal generated after the accumulation to the error weight coefficient generation module 13;

[0049] The second module value data signal output by the module value calculation module 9 is transmitted to the module value comparison module 11; the module value comparison module 11 performs numerical comparison processing on each group of module value data in the received module value data signal, and transmits the minimum module value data signal obtained by the numerical comparison to the error weighting coefficient generation module 13; at the same time, the module value comparison module 11 generates a minimum module value data state indication signal, and transmits the minimum module value data state indication signal to the error calculation benchmark selection module 12;

[0050] The error weight coefficient generation module 13 performs complex subtraction on the received module value accumulation signal and the minimum module value data signal, and uses the received module value accumulation signal for normalization processing, and finally generates an error weight coefficient signal and transmits it to the error calculation module 14;

[0051] The error calculation reference selection module 12 selects the I / Q data mapping values ​​of all constellation points in the current modulation system standard constellation diagram according to the received modulation system indication signal, and selects the constellation point I / Q data mapping value corresponding to the current minimum modulus data according to the received minimum modulus data state indication signal, and transmits the constellation point I / Q data mapping value to the error calculation module 14;

[0052] The error calculation module 14 receives the I / Q two-way digital baseband signal output by the data extraction module 6, the constellation point I / Q data mapping value output by the error calculation reference selection module 12 and the error weighting coefficient signal output by the error weighting coefficient generation module 13, and performs complex subtraction operation on the received I / Q two-way digital baseband signal and the constellation point I / Q data mapping value, and performs weighted processing on the complex subtraction operation result with the received error weighting coefficient signal, and finally obtains the I / Q two-way equalized error signal, and transmits it to the signal-to-noise ratio estimation module 3 and the convolution coefficient generation module 4 respectively after branching processing.

[0053] Compared with the traditional channel estimation and adaptive equalization device, the present invention does not need to transmit pilot signals and adopts a fully digital implementation scheme, and realizes multiple functions such as signal-to-noise ratio estimation, channel estimation and signal adaptive equalization at the same time, thereby improving the system processing efficiency and reducing the complexity of the system. In addition, when performing equalization error extraction, in addition to extracting the current equalization error value, the correct probability information of the current equalization error value is also calculated, and the correct probability information of the current equalization error value is used to perform weighted processing on the current equalization error value before subsequent processing, thereby improving the convergence speed and stability of the device.

[0054] The present invention introduces a modulation system indication signal into the error extraction module, thereby adapting to a variety of modulation systems and improving the applicability of the device. The hardware logic resources required by the device are small, the implementation complexity is low, and it is easy to implement and promote.

Claims

1. A channel estimation and adaptive equalization device, It is characterized in that It includes a digital AGC (1), a branch delay module (2), a signal-to-noise ratio estimation module (3), a parallel convolution operation module (4), a convolution coefficient generation module (5), a data extraction module (6), an error extraction module (7) and an FFT interpolation module (8); The digital AGC (1) performs signal power detection and power adjustment on the input I / Q two-channel digital baseband signals, transmits the signal power detection signal to the signal-to-noise ratio estimation module (3), and transmits the power-adjusted I / Q two-channel digital baseband signals to the branch delay module (2); The branch delay module (2) branches and independently delays the received I / Q two-way digital baseband signals to generate two groups of I / Q two-way digital baseband signals, wherein the first group of I / Q two-way digital baseband signals is transmitted to the parallel convolution operation module (4), and the second group of I / Q two-way digital baseband signals is transmitted to the convolution coefficient generation module (5); The parallel convolution operation module (4) receives the first group of I / Q two-way convolution coefficients transmitted by the convolution coefficient generation module (5), and performs convolution operation on the input I / Q two-way digital baseband signals to generate the I / Q two-way digital baseband signals after operation processing and transmit them to the data extraction module (6); The data extraction module (6) extracts peak points of the input I / Q two-way digital baseband signals and outputs them respectively. The I / Q two-way digital baseband signals output by the data extraction module (6) are further divided to generate two groups of I / Q two-way digital baseband signals, wherein the first group of I / Q two-way digital baseband signals is output by the device, and the output signal is the I / Q two-way digital baseband signal after equalization. The second group of I / Q two-way digital baseband signals is transmitted to the error extraction module (7); The error extraction module (7) processes the input I / Q two-way digital baseband signal according to the input modulation system indication signal, generates and outputs an I / Q two-way equalization error signal, and the I / Q two-way equalization error signal output by the error extraction module is further divided to generate two groups of I / Q two-way equalization error signals, the first group of I / Q two-way equalization error signals is transmitted to the signal-to-noise ratio estimation module (3), and the second group of I / Q two-way equalization error signals is transmitted to the convolution coefficient generation module (5); The signal-to-noise ratio estimation module (3) receives the signal power detection signal transmitted by the digital AGC (1) and the I / Q two-way equalization error signal output by the error extraction module (7), generates a signal-to-noise ratio estimation value through calculation and outputs it as the signal-to-noise ratio estimation signal of the device; The convolution coefficient generation module (5) receives the second group of I / Q two-way digital baseband signals output by the branch delay module (2) and the second group of I / Q two-way equalization error signals output by the error extraction module (7), generates I / Q two-way convolution coefficient signals by calculation and outputs them; The I / Q two-way convolution coefficient signal output by the convolution coefficient generation module (5) is further divided into two groups of I / Q two-way convolution coefficient signals, the first group of I / Q two-way convolution coefficient signals is transmitted to the parallel convolution operation module (4), and the second group of I / Q two-way convolution coefficient signals is transmitted to the FFT interpolation module (8); The FFT interpolation module (8) performs interpolation and FFT processing on the input I / Q two-way convolution coefficient signals and outputs them as the channel estimation signal of the device.

2. A channel estimation and adaptive equalization device according to claim 1, It is characterized in that The error extraction module (7) comprises a module value calculation module (9), a module value accumulation module (10), a module value comparison module (11), an error calculation benchmark selection module (12), an error weighting coefficient generation module (13) and an error calculation module (14); The I / Q two-way digital baseband signal output by the data extraction module (6) is split to generate two groups of I / Q two-way digital baseband signals, and the first group of I / Q two-way digital baseband signals is transmitted to the module value calculation module (9); The module value calculation module (9) selects the I / Q data mapping values ​​of all constellation points in the current modulation system standard constellation diagram according to the input modulation system indication signal, and performs complex number subtraction on the I / Q data mapping values ​​of each constellation point and the input I / Q two-way digital baseband signal, and performs module value operation on the calculated values ​​after the complex number subtraction to generate a group of module value data signals; the module value calculation module (9) outputs each group of module value data signals generated in real time; The module value data signal output by the module value calculation module (9) is divided into two modules value data signals, and the first module value data signal is transmitted to the module value accumulation module (10). The module value accumulation module (10) performs accumulation processing on each group of module value data values ​​in the received module value data signal, and transmits the module value accumulation signal generated after the accumulation processing to the error weight coefficient generation module (13); The second module value data signal output by the module value calculation module (9) is transmitted to the module value comparison module (11), the module value comparison module (11) performs numerical comparison processing on each group of module value data in the received module value data signal, and transmits the minimum module value data signal obtained by the numerical comparison to the error weight coefficient generation module (13); at the same time, the module value comparison module (11) generates a minimum module value data state indication signal, and transmits the minimum module value data state indication signal to the error calculation benchmark selection module (12); The error weight coefficient generation module (13) performs complex subtraction on the received module value accumulation signal and the minimum module value data signal, and uses the received module value accumulation signal for normalization processing, and finally generates an error weight coefficient signal and transmits it to the error calculation module (14); The error calculation reference selection module (12) selects the I / Q data mapping values ​​of all constellation points in the current modulation system standard constellation diagram according to the received modulation system indication signal, selects the constellation point I / Q data mapping value corresponding to the current minimum modulus data according to the received minimum modulus data state indication signal, and transmits the constellation point I / Q data mapping value to the error calculation module (14); The error calculation module (14) receives the I / Q two-way digital baseband signals output by the data extraction module (6), the constellation point I / Q data mapping values ​​output by the error calculation reference selection module (12) and the error weighting coefficient signal output by the error weighting coefficient generation module (13), performs a complex subtraction operation on the received I / Q two-way digital baseband signals and the constellation point I / Q data mapping values, and performs weighted processing on the complex subtraction result using the received error weighting coefficient signal, finally obtaining the I / Q two-way equalization error signals, and transmitting them to the signal-to-noise ratio estimation module (3) and the convolution coefficient generation module (5) respectively after branching processing.

Citation Information

Patent Citations

  • Frequency domain channel estimation method of asynchronous multicarrier system

    CN107171984A

  • Time-domain synchronous OFDM space-frequency coding time-frequency combination channel estimating method

    CN1791084A

  • Coded orthogonal frequency division multiplexing equalizer

    KR1020000018811A

  • Transmission path equalization processing device, and transmission path equalization processing method

    WO2021100092A1