Low-complexity ofdm synchronization implementation method with low-bit non-uniform quantization

Abstract

The present application relates to a kind of low-complexity OFDM synchronous implementation methods of low-bit non-uniform quantization, belong to wireless communication technical field.It includes: in the information data of sending end, training sequence is inserted for synchronization;In receiving end, by the low-bit non-uniform quantization of set quantization threshold to received signal, then carries out delay autocorrelation, obtains coarse synchronization position;With the same quantization threshold, local training sequence is carried out low-bit non-uniform quantization, and with coarse synchronization position as starting point, the cross-correlation operation of received signal and local training sequence is obtained to final timing synchronization position.The acquisition of best non-uniform quantization threshold is carried out by the way of traversal search, and with mutual information maximum as decision criterion.The present application reduces the complexity of synchronous correlation operation by carrying out low-bit quantization, and improves the synchronization accuracy under low-bit quantization by the way of non-uniform quantization.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication technology, and mainly relates to OFDM communication system synchronization technology, specifically to a low-complexity OFDM synchronization implementation method with low-bit non-uniform quantization. Background Technology

[0002] Orthogonal Frequency Division Multiplexing (OFDM) evolved from Frequency Division Multiplexing. Its subcarriers are orthogonal to each other, offering advantages such as high bandwidth utilization, fast data transmission rate, and strong resistance to multipath interference. Synchronization is the first step in baseband processing after analog signals are converted to digital signals, and it is crucial for correct data transmission. Its performance directly determines the quality of communication; therefore, stable, reliable, and accurate synchronization is essential for the reliable operation of the entire communication system.

[0003] Compared to single-carrier communication systems, OFDM systems have much stricter synchronization requirements. Timing deviations introduce inter-symbol interference, causing significant loss in receiver performance. Therefore, symbol synchronization is a crucial prerequisite and foundation for reliable OFDM demodulation. Currently, synchronization algorithms based on training sequences are widely used due to their simplicity and good timing performance. However, they require the insertion of specially structured sequences for synchronization estimation, increasing time-frequency overhead.

[0004] Synchronization algorithms based on training sequences can be divided into two types: delayed autocorrelation algorithms and cross-correlation algorithms. Delayed autocorrelation algorithms have low hardware implementation complexity and low computational cost, but poor timing performance. Cross-correlation algorithms have high timing accuracy, but their computational cost is large and they consume too much hardware resources. Summary of the Invention

[0005] To address the problems existing in the prior art, this invention proposes a low-complexity OFDM synchronization implementation method with low-bit non-uniform quantization. This method can be used for timing synchronization in millimeter-wave OFDM communication systems, and can obtain accurate timing synchronization results while reducing hardware resource consumption.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] A low-complexity OFDM synchronization implementation method with low-bit non-uniform quantization includes the following steps:

[0008] Step 1: At the sending end, the training sequence T is framed with the information data and sent, wherein the training sequence is used as the frame header for timing synchronization at the receiving end.

[0009] Step 2: Based on the criterion of maximizing mutual information, determine the optimal quantization threshold for non-uniform quantization through a traversal search.

[0010] Step 3: At the receiving end, the received signal is converted into a digital signal y(n) by an ADC. Then, the digital received signal y(n) is subjected to N-bit non-uniform quantization using the optimal quantization threshold to obtain y(n) with 1 sign bit and N-1 amplitude bits. q (n); where 2≤N≤4;

[0011] Step 4, quantize the received signal y q (n) Perform delayed autocorrelation and threshold detection based on the autocorrelation results to obtain coarse timing synchronization results;

[0012] Step 5: Apply the same optimal quantization threshold as the sending end to the local training sequence T. L (n) Perform N-bit non-uniform quantization to obtain the local quantized training sequence T. q (n);

[0013] Step 6: Starting with the coarse timing synchronization result, convert the quantized received signal y... q (n) and the local quantization training sequence T q (n) performs cross-correlation operations and completes the final timing synchronization through threshold detection.

[0014] Furthermore, step 2 specifically includes the following steps:

[0015] (201) Determine the quantization threshold A q The search range, i.e., the amplitude range of the training sequence [A] min A max ];

[0016] (202) Given N-bit non-uniform quantization, there are 1 sign bit and N-1 amplitude bits. The quantization thresholds are (2 N-1 -1) items, that is, the quantization threshold is A. qi i = 1, 2, ... 2 N-1 -1, where 2≤N≤4; when N=2, execute steps (203)-(206); when N=3, execute steps (203)-(210); when N=4, execute steps (203)-(214);

[0017] (203) A min Set as quantization threshold A q1 The initial value;

[0018] (204) with A q1 To set the quantization threshold, the training sequence T is quantized using 2 bits, resulting in T1 = {α, β}. The specific quantization rules are as follows:

[0019] α = sgn(T),

[0020] (205) Calculate the mutual information between the 2-bit quantized training sequence T1 and the unquantized training sequence T;

[0021] (206) Using ΔA as the search step, in [A min A max Update A within the range q1 The value is A q1 +ΔA, return to step (204), until the updated A q1 Greater than A max And based on the criterion of maximizing mutual information, the non-uniform quantization threshold value A is determined. q1 The optimal value A q1,opt ;

[0022] (207) The optimal quantization threshold A in 2-bit non-uniform quantization q1,opt Based on this, A min Set as quantization threshold A q2 The initial value of A q1,opt Set as quantization threshold A q3 The initial value;

[0023] (208) with A q2 and A q3 Using a threshold, the training sequence T is quantized using 3 bits to obtain T2 = {α, β, γ}. The specific quantization rule is as follows:

[0024] α = sgn(T),

[0025] (209) Calculate the mutual information between the 3-bit quantized training sequence T1 and the unquantized training sequence T;

[0026] (210) Using ΔA as the search step, in [A min A q1,opt Update A within the range q2 The value is A q2 +ΔA, at [A q1,opt A max Update A within the range q3 The value is A q3 +ΔA, return to step (208), until A q2 and A q3 The updated value is out of range, and the non-uniform quantization threshold value A is determined based on maximizing mutual information. q2 and A q3 The optimal value A q2,opt and A q3,opt ;

[0027] (211) The optimal quantization threshold A in 3-bit non-uniform quantization q1,opt A q2,opt and Aq3,opt Based on this, A min Set as quantization threshold A q4 The initial value of A q2,opt Set as quantization threshold A q5 The initial value of A q1,opt Set as quantization threshold A q6 The initial value of A q3,opt Set as quantization threshold A q7 The initial value;

[0028] (212) A q4 A q5 A q6 and A q7 Using the threshold, the training sequence T is subjected to 4-bit non-uniform quantization to obtain T3={α,β,γ,λ}, and the specific quantization rule is as follows:

[0029]

[0030] (213) Calculate the mutual information between the 4-bit quantized training sequence T1 and the unquantized training sequence T;

[0031] (214) Using ΔA as the search step, in [A min A q2,opt Update A within the range q4 The value is A q4 +ΔA, at [A q2,opt A q1,opt Update A within the range q5 The value is A q5 +ΔA, at [A q1,opt A q3,opt Update A within the range q6 The value is A q6 +ΔA, at [A q3,opt A max Update A within the range q7 The value is A q7 +ΔA, return to step (212), until A q4 A q5 A q6 and A q7 The updated value is out of range, and the non-uniform quantization threshold value A is determined based on maximizing mutual information. q4 A q5 A q6 and A q7 The optimal value A q4,opt A q5,opt A q6,opt and A q7,opt .

[0032] Furthermore, in step 3, the digital received signal y(n) undergoes N-bit non-uniform quantization, the specific process of which is as follows:

[0033] Using the optimal quantization threshold A obtained in step 2 for non-uniform quantization qi,opt i = 1, 2, ... 2 N-1 -1 Perform N-bit non-uniform quantization on the digital received signal y(n) to obtain y with 1 sign bit and N-1 amplitude bits. q (n).

[0034] Furthermore, step 6 is performed as follows:

[0035] Using the coarse timing synchronization result as the starting point for the received signal, the quantized received signal y q (n) and the local quantization training sequence T q (n) Perform cross-correlation calculation to obtain the cross-correlation value V, as shown in the following formula:

[0036]

[0037] Where L represents the length of the training sequence;

[0038] The cross-correlation value V and the decision threshold V th Compare, if V≥V th If the synchronization is complete, then the synchronization is considered complete.

[0039] The beneficial effects of this invention are as follows:

[0040] 1. Reduced computational complexity of timing synchronization in OFDM synchronization system: This invention quantizes ADC output data into 1 sign bit and N-1 amplitude bits using low-bit quantization, and uses XOR operation of sign bit and addition operation of amplitude bit to replace complex multiplication operation of related operations to reduce computational and hardware implementation complexity.

[0041] 2. Improved timing synchronization accuracy of low-bit quantization OFDM system: This paper replaces the traditional uniform quantization method with non-uniform quantization, which improves the timing synchronization accuracy under the same quantization bit width. Based on the maximization of mutual information, the optimal quantization threshold for N-bit non-uniform quantization is obtained by traversal search. Attached Figure Description

[0042] Figure 1 This is a flowchart of the low-complexity OFDM synchronization implementation method with low-bit non-uniform quantization according to the present invention.

[0043] Figure 2 This is a diagram of the training sequence structure of the present invention.

[0044] Figure 3 This is a block diagram illustrating the implementation of the present invention. Detailed Implementation

[0045] A low-complexity OFDM synchronization implementation method with low-bit non-uniform quantization, see [link to relevant documentation]. Figure 1 and Figure 3 The specific steps are explained using 3-bit non-uniform quantization as an example:

[0046] (1) At the transmitting end, the training sequence T is framed with the information data and transmitted. The training sequence serves as the frame header for timing synchronization at the receiving end. The training sequence consists of two identical ZC sequences, and its specific structure is as follows: Figure 2 As shown.

[0047] (2) Based on maximizing mutual information, the optimal quantization threshold for 3-bit non-uniform quantization is determined through a traversal search. The specific implementation is as follows:

[0048] (201) Determine the quantization threshold A q The search range, i.e., the amplitude range of the training sequence [A] min A max ];

[0049] (202) 3-bit non-uniform quantization, with 1 sign bit and 2 amplitude bits, has 3 quantization thresholds, i.e., quantization threshold A. q1 A q2 and A q3 ;

[0050] (203) A min For quantization threshold A q1 The initial value;

[0051] (204) with A q1 To determine the quantization threshold, the training sequence T is subjected to 2-bit non-uniform quantization, resulting in T1 = {α, β}. The specific quantization rules are as follows:

[0052]

[0053] (205) Calculate the mutual information between the 2-bit quantized training sequence T1 and the unquantized training sequence T;

[0054] (206) Using ΔA as the search step, in [A min A max Update A within the range q1 The value is A q1 +ΔA, and return to step (204) until the updated A. q1 If the range is exceeded, and the non-uniform quantization threshold value A is determined based on maximizing mutual information. q1 The optimal value is A q1,opt ;

[0055] (207) The optimal quantization threshold A in 2-bit non-uniform quantization q1,opt Based on A q2 and A q3 As a threshold, set A min Set as quantization threshold A q2 The initial value of A q1,opt Set as quantization threshold A q3 The initial value;

[0056] (208) with A q2 and A q3 Using a threshold, the training sequence T is quantized using 3 bits to obtain T2 = {α, β, γ}. The specific quantization rule is as follows:

[0057]

[0058] (209) Calculate the mutual information between the 3-bit quantized training sequence T1 and the unquantized training sequence T;

[0059] (210) Using ΔA as the search step, in [A min A q1,opt Update A within the range q2 The value is A q2 +ΔA, at [A q1,opt A max Update A within the range q3 The value is A q3 +ΔA, return to step (208), until A q2 and A q3 The updated value is out of range, and the non-uniform quantization threshold value A is determined based on maximizing mutual information. q2 and A q3 The optimal value A q2,opt and A q3,opt ;

[0060] (3) At the receiving end, the received signal is converted from analog signal to digital signal y(n) by ADC, and then three optimal quantization thresholds A are used. q1,opt A q2,opt and A q3,opt Perform 3-bit non-uniform quantization on the digital received signal y(n) to obtain y(n) with 1 sign bit and 2 amplitude bits. q (n) = {α,β,γ}. The specific quantization rules are as follows:

[0061]

[0062] (4) The quantized received signal y q (n) Perform delayed autocorrelation and threshold detection based on the autocorrelation results to obtain coarse timing synchronization results.

[0063] (5) Use the same quantization threshold A as the transmitter. q1,opt A q2,opt and A q3,opt For the local training sequence T L (n) Perform 3-bit non-uniform quantization to obtain the local quantized training sequence T. q (n).

[0064] (6) Starting from the coarse timing result, quantize the received signal y q (n) and the local quantization training sequence T q (n) Perform cross-correlation operation to obtain the cross-correlation value:

[0065]

[0066] Where L represents the length of the training sequence. The cross-correlation value V is compared with the decision threshold V. th Compare, if V≥V th If the synchronization is complete, then the synchronization is considered complete.

Claims

1. A low-complexity OFDM synchronization implementation method with low-bit non-uniform quantization, characterized in that, Includes the following steps: Step 1: Send the training sequence at the sending end. The training sequence is framed with information data and sent, with the training sequence serving as the frame header for timing synchronization at the receiving end. Step 2: Based on the criterion of maximizing mutual information, determine the optimal quantization threshold for non-uniform quantization through a traversal search. Step 3: At the receiving end, the received signal is converted into a digital signal by the ADC. Then, the optimal quantization threshold is used to quantize the digital received signal. conduct Non-uniform bit quantization yields a 1-bit sign bit. Bit amplitude ;in ; Step 4: Quantize the received signal Perform delayed autocorrelation and threshold detection based on the autocorrelation results to obtain coarse timing synchronization results; Step 5: Apply the same optimal quantization threshold as the sending end to the local training sequence. Perform N-bit non-uniform quantization to obtain the local quantized training sequence. ; Step 6: Starting with the coarse timing synchronization result, quantize the received signal... With local quantization training sequence Perform cross-correlation calculations and complete the final timed synchronization through threshold detection; Step 2 specifically includes the following steps: (201) Determine the quantization threshold The search range, i.e., the amplitude range of the training sequence. ; (202) has In non-uniform bit quantization, there is a total of 1 sign bit and AND bit. Bit amplitude bits, quantization threshold has One, that is, the quantization threshold is ,in ; When the time comes, execute steps (203)-(206); When, execute steps (203)-(210). When the time comes, execute steps (203)-(214); (203) will Set as quantization threshold The initial value; (204) with For quantization threshold, the training sequence Perform 2-bit quantization to obtain The specific quantification rules are as follows: ； (205) Calculate the 2-bit quantized training sequence With the pre-quantization training sequence Mutual information; (206) with For the search step, in Update within range The value is + Return to step (204) until the updated... Greater than A max And based on the criterion of maximizing mutual information, the non-uniform quantization threshold is determined. Optimal value ; (207) Optimal quantization threshold in 2-bit non-uniform quantization Based on this, Set as quantization threshold initial value, Set as quantization threshold The initial value; (208) with and As a threshold, for the training sequence Perform 3-bit quantization to obtain The specific quantification rules are as follows: ； (209) Calculate the 3-bit quantized training sequence With the pre-quantization training sequence Mutual information; (210) with For the search step, in Update within range The value is + ,exist Update within range The value is + Return to step (208) until and The updated value is out of range, and the non-uniform quantization threshold is determined based on maximizing mutual information. and Optimal value and ; (211) Optimal quantization threshold in 3-bit non-uniform quantization , and Based on this, Set as quantization threshold initial value, Set as quantization threshold initial value, Set as quantization threshold initial value, Set as quantization threshold The initial value; (212) with , , and As a threshold, for the training sequence Perform 4-bit non-uniform quantization to obtain The specific quantification rules are as follows: ； (213) Calculate the 4-bit quantized training sequence With the pre-quantization training sequence Mutual information; (214) with For the search step, in Update within range The value is ,exist Update within range The value is ,exist Update within range The value is ,exist Update within range The value is Return to step (212) until , , and The updated value is out of range, and the non-uniform quantization threshold is determined based on maximizing mutual information. , , and Optimal value , , and .

2. The low-complexity OFDM synchronization implementation method with low-bit non-uniform quantization according to claim 1, characterized in that, Step 3 involves receiving digital signals. conduct Bit non-uniform quantization, the specific process is as follows: Using the optimal quantization threshold obtained in step 2 for non-uniform quantization Digital received signal conduct Non-uniform bit quantization yields a 1-bit sign bit. Bit amplitude .

3. The low-complexity OFDM synchronization implementation method with low-bit non-uniform quantization according to claim 1, characterized in that, The specific method for step 6 is as follows: Using the coarse timing synchronization result as the starting point for the received signal, the quantized received signal... With local quantization training sequence Perform cross-correlation calculations to obtain the cross-correlation value V, as shown in the following formula: in Indicates the length of the training sequence; The cross-correlation value V is compared with the decision threshold. If a comparison is made, If the synchronization is complete, then the synchronization is considered complete.

Images