A time-domain average-based partial matched filter signal acquisition method

Abstract

The application belongs to the technical field of communication and particularly relates to a partial matching filter signal acquisition method based on time domain averaging. The method mainly sets M partial matching filters in a receiving end, simultaneously divides the received signals into M segments, inputs the signals into the M partial matching filters for partial matching filter processing, performs a difference operation on the obtained filter outputs, obtains a correlation peak value according to a time domain average length, and finally performs signal acquisition judgment by using the correlation peak value. The method has small calculation amount, simple realization, and can obtain certain gain at a small cost.

Description

Technical Field

[0001] This invention belongs to the field of communication technology, and specifically relates to a method for capturing partially matched filtered signals based on time-domain averaging. Background Technology

[0002] In spread spectrum systems, for security reasons, signals often need to be submerged in noise. Therefore, a spreading sequence is required as a synchronization header. The receiving end must generate a synchronization header consistent with the transmitting end. Similarly, the synchronization sequence must have excellent autocorrelation; otherwise, the correlation peak will be lost, leading to synchronization failure. A simple synchronization method is to send the sequence in a pilot signal for the receiving end to acquire and analyze. However, considering security and spectral efficiency, the receiving end usually acquires the signal by directly processing the synchronization sequence. Besides noise affecting the correlation peak during synchronization, high-frequency offset also contributes to the loss of correlation peaks. High-frequency offset is often caused by the relative velocity and acceleration of the transmitting and receiving ends. When the relative acceleration at both ends changes significantly, it will lead to a large Doppler frequency offset. In low signal-to-noise ratio environments, the frequency offset affects the size of the despread correlation peak, resulting in a lower acquisition success rate. Currently, the commonly used acquisition algorithm to combat frequency offset and noise is PMF-FFT. However, this algorithm suffers from scallop loss and high implementation complexity, leading to a complex receiving end structure. Summary of the Invention

[0003] To address the aforementioned problems, this invention provides a method for capturing partially matched filtered signals based on time-domain averaging.

[0004] The technical solution of this invention is as follows:

[0005] A method for capturing partially matched filters based on time-domain averaging, characterized in that it includes:

[0006] S1. At the transmitting end, take N_syn synchronization header frames of length L_syn and send them;

[0007] S2. The receiving end performs partial matched filtering and differential operation on the received signal after down-conversion using a digitally controlled oscillator, specifically including:

[0008] At the receiving end, M = L_syn / N partially matched filters are configured, where N is the coherence length of each partially matched filter. The local synchronization sequence of length L_syn is sequentially divided into M segments of length N. The corresponding sequence is stored in each partially matched filter. The received signal is then divided into M segments and input into the M partially matched filters for partially matched filtering. The M partially matched filters are defined as PMF. i If i = 1...M, then the i-th partially matched filter performs a temporal convolution on the i-th synchronization sequence, and the convolution result is represented as COR. i ;

[0009] Perform a difference operation on the outputs of the i-th partially matched filter and the j-th partially matched filter, where j = i + 1 and j ≤ M. Summing all the differences yields the processing result COR of the received signal at time t. diff (t):

[0010]

[0011] S3, according to COR diff (t) Calculate the correlation peak COR(t):

[0012] COR(t) = COR diff (t)+COR diff (t+L_syn)+...+COR diff (t+(T-1)*L_syn)

[0013] Where T is the average length in the time domain;

[0014] S4. Determine if COR(t)>TH is true. If yes, proceed to S41. Otherwise, do nothing and continue searching until COR(t)>TH.

[0015] S41. If there exists a previous time t′ such that COR(t′)>TH holds true, determine whether the difference between time t and time t′ is equal to L_syn. If yes, increment the counter and proceed to S5; otherwise, proceed to S42. If there exists a previous time t′ such that COR(t′)>TH holds true, set the counter to 1 and record the time, then return to S4.

[0016] S42, set the counter to 1, record the moment when COR(t) > TH, and return to S4;

[0017] S5. Determine if the counter value is greater than or equal to the set value L. If so, the following conditions are met:

[0018] COR(t1)>TH,...,COR(t L )>TH

[0019] t2-t1=L_syn,t3-t2=L_syn,...,t L -t L-1 =L_syn

[0020] If a signal is detected, a synchronization flag is issued; otherwise, it is determined that no signal has been detected.

[0021] Furthermore, the method for obtaining the coherence length N and time-domain average length T of the partially matched filter is to let:

[0022]

[0023] in

[0024]

[0025] △f is the frequency offset, f s Let σ be the sampling rate. 2 For noise power, N, M, and T are all integers greater than 0, under the constraints... Given 1≤N≤L_syn and 1≤T≤N_syn, several solutions for N and T are obtained. The solution with the largest N and the smallest T is selected as the coherence length N and time-domain average length T of the required partially matched filter.

[0026] The beneficial effects of the present invention are as follows: the method of the present invention has a small amount of computation and is simple to implement. It can obtain a certain gain with relatively little cost. By increasing the averaging number T, it can not only improve the algorithm's ability to resist noise, but also improve its ability to resist frequency offset, and even outperform PMF-FFT. Attached Figure Description

[0027] Figure 1 This is a schematic block diagram of PMF (Partially Matched Filter).

[0028] Figure 2 This is a schematic block diagram of PMF (Partially Matched Filtering) based on time-domain averaging. Detailed Implementation

[0029] The specific technical solution of the present invention will be described below with reference to the accompanying drawings.

[0030] like Figure 1 The diagram illustrates the traditional PMF (Partially Matched Filtering) process. Overall, this structure achieves partial matched filtering by delaying, weighting, and accumulating the input data, enabling the detection of specific signals. The specific process is as follows:

[0031] 1) There are N Z's in the diagram. -1 Unit, Z -1 The unit of delay represents the input signal being delayed by one sampling point. Multiple delay units are cascaded in the diagram to achieve different input data delay times.

[0032] 2)C m+1 ,...,C m+N The multiplier coefficients are represented by the conjugate complex numbers of the synchronization sequences. Each delayed signal is multiplied by its coefficient. ∑ is the accumulator, which sums the signal components at different delay times after multiplication with the coefficients to obtain the i-th partially matched filter output COR. i ;

[0033] like Figure 2The diagram illustrates the improved time-domain averaging-based PMF process used in this invention. The method of this invention is for a spread spectrum system, including a transmitter and a receiver, both of which include a transmitting end (TX) and a receiving end (RX). The transmitting end includes framing and spreading modules; the receiving end includes acquisition, fine synchronization, and despreading modules. The specific process is as follows:

[0034] S1. At the transmitting end, take N_syn synchronization header frames of length L_syn and send them;

[0035] S2. The receiving end performs partial matched filtering and differential operation on the received signal after down-conversion using a numerically controlled oscillator (NCO), specifically including:

[0036] According to the formula:

[0037]

[0038] The threshold can be appropriately increased or decreased according to system requirements;

[0039] The loss is related to the frequency offset:

[0040]

[0041] In the above formula, Δf is the frequency offset, which can be set to the highest frequency offset in the system operating environment. s Let σ be the sampling rate. 2 The noise power can be set to the maximum noise power in the system operating environment, N is the coherence length of the partially matched filter, M = L_syn / N is the number of partially matched filters, and T is the time-domain averaging length. N, M, and T are all integers greater than 0, under the constraints... Under the conditions 1≤N≤L_syn and 1≤T≤N_syn, several solutions for N and T can be calculated. The solution with the largest N and the smallest T can be selected as the system design parameters, thus making the receiving end structure relatively simple.

[0042] The receiver is equipped with M = L_syn / N partially matched filters. The local synchronization sequence of length L_syn is sequentially divided into M segments of length N. The corresponding sequence is stored in each partially matched filter. The received signal is then divided into M segments and input into the M partially matched filters for partially matched filtering. The M partially matched filters are defined as PMF. i If i = 1...M, then the i-th partially matched filter performs a temporal convolution on the i-th synchronization sequence, and the convolution result is represented as COR. i ;

[0043] Perform a difference operation on the outputs of the i-th partially matched filter and the j-th partially matched filter, where j = i + 1 and j ≤ M. Summing all the differences yields the processing result COR of the received signal at time t. diff (t);

[0044]

[0045] S3, according to COR diff (t) Calculate the correlation peak COR(t):

[0046] COR(t) = COR diff (t)+COR diff (t+L_syn)+...+COR diff (t+(T-1)*L_syn)

[0047] S4. Determine if COR(t)>TH is true. If yes, proceed to S41. Otherwise, do nothing and continue searching until COR(t)>TH.

[0048] S41. If there is a previous time t′, determine whether the difference between time t and the previous time t′ that made COR(t′)>TH true is equal to L_syn. If yes, increment the counter and proceed to S5; otherwise, proceed to S42. If there is no previous time t′, set the counter to 1 and record the time, then proceed to S4.

[0049] S42, set the counter to 1, record the moment when COR(t)>TH, and proceed to S4;

[0050] S5. Determine if the counter value is greater than or equal to the set value L. If so, the following conditions are met:

[0051] COR(t1)>TH,...,COR(t L )>TH

[0052] t2-t1=L_syn,t3-t2=L_syn,...,t L -t L-1 =L_syn

[0053] If a signal is detected, a synchronization flag is issued; otherwise, it is determined that no signal has been detected.

Claims

1. A method for capturing partially matched filtered signals based on time-domain averaging, characterized in that, include: S1, at the transmitting end, the length is... Synchronization sequence Send after a synchronization header frame; S2. The receiving end performs partial matched filtering and differential operation on the received signal after down-conversion using a digitally controlled oscillator, specifically including: Set at the receiving end Partially matched filters Let the coherence length of the partially matched filter be [length]. The local synchronization sequence is divided into the following order: Section length is The sequence is divided into corresponding sequences by storing them in the matched filter at each part, thus dividing the received signal into... Segment, input to Partially matched filtering is performed in a partially matched filter, defining... The numbering of the partially matched filters is Then the first The partial matched filter for the th Performing temporal convolution on the synchronized sequence, the resulting convolution is represented as follows: ; For the The output of the partially matched filter and the first The outputs of the partially matched filters are differentially processed. and After summing all the differences, we get The processing results of the received signal at any time : ， Coherence length of partially matched filters and time domain average length The method to obtain it is to let: ， in ， For frequency offset, Sampling rate, For noise power, , , All are integers greater than 0, under the constraints. , , The result is and Several solutions, choose The largest, The smallest set of solutions serves as the coherence length of the required partially matched filter. and time domain average length ; S3, according to Calculate the correlation peak : ； S4, Judgment If the condition is met, proceed to S41; otherwise, do nothing and continue searching until... ; S41. If a previous time step exists... Make Establishment, judgment time With time Is the difference equal to If yes, the counter is incremented and the process proceeds to S5; otherwise, the process proceeds to S42. If no previous time step exists... Make If the condition is met, the counter is set to 1 and the moment is recorded, then return to S4; S42. Set the counter to 1 and record the current use. At that moment, let's go back to S4; S5. Determine if the counter value is greater than or equal to the set value. If so, then the following conditions are met: ， If a signal is detected, a synchronization flag is issued; otherwise, it is determined that no signal has been detected.

Images