Unmanned aerial vehicle co-frequency anti-interference communication method based on multi-antenna technology

Abstract

The invention belongs to the technical field of communication, and particularly relates to an unmanned aerial vehicle co-frequency anti-interference communication method based on a multi-antenna technology. According to the method, time / frequency synchronization and interference elimination are achieved at the same time through the lead code sequence, different from an existing method, the method does not need information of communication and interference channels, meanwhile, the method provided by the invention has a strict theoretical guarantee, and the method is suitable for large-scale popularization and application. The required lead code sequence length can be clearly given under the condition that the number of receiving antennas and the number of interference sources are given. According to the invention, time / frequency alignment and interference elimination can be synchronously realized under the condition that the channel state information of the transmitter / jammer is unknown. Experimental results on a software radio platform show that even if the interference signal power is 30dB stronger than the target signal power, the method provided by the invention can still achieve a lower bit error rate.

Description

Technical Field

[0001] The present invention belongs to the field of communication technology, and specifically relates to a same-frequency anti-interference communication method for unmanned aerial vehicles based on multi-antenna technology. Background Art

[0002] Unmanned Aerial Vehicle (UAV) is a flexible, low-cost and versatile aircraft that has been widely used in a variety of tasks such as aerial photography, disaster relief, material delivery, and combat assistance. UAVs are controlled by ground or space base stations, and control commands are transmitted from the base station to the remote UAV via a wireless link. Although UAVs usually enjoy good line-of-sight channel conditions, the path loss is usually high due to the long distance between the UAV and the base station, and the wireless link is susceptible to interference. In order to ensure the reliability of the UAV communication link, it is crucial to develop UAV anti-interference communication technology in an interference environment.

[0003] At present, the commonly used communication anti-interference technologies mainly include spread and frequency hopping technologies. However, the spectrum utilization rate of spread and frequency hopping technologies is low, and the communication rate that can be achieved under the same frequency resources is relatively limited. In order to achieve higher spectrum utilization, multi-antenna based co-frequency anti-interference technology has received more and more attention. Multi-antenna co-frequency anti-interference technology mainly includes adaptive beamforming, semi-blind source separation, etc. Specifically, adaptive beamforming uses the incoming direction information of the target signal to adaptively form a directional narrow beam to suppress the interference signal. However, in actual scenarios, the incoming direction information of the target signal is usually difficult to obtain. Other multi-antenna auxiliary anti-interference technologies use the information of the communication channel or the pilot sequence of the target signal to achieve interference suppression. For the former, the receiving end needs to estimate the communication channel first, and then extract the interference channel information from the covariance matrix of the received signal. However, the estimated communication channel depends on the so-called interference gap, which only exists when the interference source adopts tracking interference. When there is no interference gap or the gap is very short, the method will be affected. Semi-blind source separation technology uses a known pilot sequence to remove the target signal from the received signal to obtain the interference channel subspace information. This method requires the receiving and transmitting ends to be completely synchronized in time and frequency, which has great limitations in practical applications. Summary of the invention

[0004] In view of the above problems, the present invention proposes a multi-antenna based UAV anti-interference communication method. The present invention uses the preamble sequence to simultaneously achieve time / frequency synchronization and interference elimination. Unlike the existing methods, the method proposed by the present invention does not require information about the communication and interference channels. At the same time, the method proposed by the present invention has a relatively strict theoretical guarantee, and can clearly give the required preamble sequence length when the number of receiving antennas and the number of interference sources are given.

[0005] The technical solution of the present invention is:

[0006] The UAV co-frequency anti-interference communication method based on multi-antenna technology sets the communication scenario as downlink communication from a base station to a UAV. There are K unknown interference sources in the environment. Both the base station and the interference source have only a single antenna. The UAV is equipped with N antennas, N>K. It is characterized in that the communication method is:

[0007] The base station periodically generates a preamble sequence {s(t), t = 1, ..., T}, where T is the length of the preamble sequence. The base station sends a frame of data composed of the preamble sequence and the data sequence. At the same time, the signals sent by K interference sources are defined as

[0008] After receiving the signals from the base station and the interference source, the drone processes the signals. The specific method is as follows:

[0009] Define a space-time filter with frequency offset compensation Where ω is the frequency offset compensation coefficient, w l is the filter coefficient, L>τ is the filter order, and the optimal frequency offset compensation coefficient and filter coefficient are calculated by receiving the signal:

[0010] The received signal model is defined as:

[0011]

[0012] in, To receive the signal, represents the channel vector between the base station and the drone, represents the channel vector between the kth interference source and the UAV, represents the Gaussian white noise vector, δ f Indicates the frequency deviation between the base station and the drone, represents the effect of frequency deviation on the target signal, and τ represents the unknown delay between the transmitter and the receiver;

[0013] The maximum possible frequency deviation δ is obtained based on the maximum moving speed and carrier frequency of the UAV max , define the minimum frequency deviation as δ min , in [δ min ,δ max ] select m equidistant points, denoted as Calculate each ω i The corresponding f c (ω), select f c (ω) The minimum value of ω i As the optimal ω * :

[0014]

[0015]

[0016] Get the optimal ω * Afterwards, the optimal filter is obtained by the formula

[0017]

[0018] According to ω * and w * Perform data recovery and obtain the recovery signal:

[0019]

[0020] Among them, F i is the number of sampling points representing the i-th frame.

[0021] Furthermore, the receiver optimizes the signal processing and The calculation is transformed into calculation Specifically, the definition is:

[0022]

[0023] Among them, ε is a very small non-negative number, I is the unit matrix, for The tth row of ;

[0024] When the number of sampling points at the receiving end is an integer multiple of T, recursively calculate from t′=1 to t′=T:

[0025]

[0026] get Then we can get the optimal frequency offset compensation ω * and the optimal filter w * , {ω * ,w *} is applied to the data bit to restore the corresponding frame length range t=[τ+1:1:τ+F i ],i∈{1,2,...∞}target signal, where F i Indicates the number of sampling points in the i-th frame.

[0027] The beneficial effect of the present invention is that the present invention can synchronously realize time / frequency alignment and interference elimination when the channel state information of the transmitter / jammer is unknown. The experimental results on the software radio platform show that even when the interference signal power is 30dB stronger than the target signal power, the proposed method can still achieve a low bit error rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 It is a schematic diagram of the frame structure in the present invention;

[0029] Figure 2 It is a schematic diagram of the structure of the received signal in the time domain;

[0030] Figure 3 It is a modulation and signal processing flow chart;

[0031] Figure 4 It is a schematic diagram of the hardware experimental platform;

[0032] Figure 5 This is a schematic diagram of simulation results based on MATLAB software;

[0033] Figure 6 This is a schematic diagram of the test results based on the hardware experimental platform. DETAILED DESCRIPTION

[0035] The technical solution of the present invention is described in detail below in conjunction with the accompanying drawings and simulation examples.

[0036] The signal sending method of the present invention is:

[0037] Consider the downlink communication scenario from the base station to the UAV, where there are K unknown interference sources in the environment. Both the base station and the interference source have only a single antenna, and the UAV is equipped with N antennas, assuming N>K. Consider the downlink from the base station to the UAV, the base station periodically generates a preamble sequence {s(t), t=1,…,T}, and the target signal s(t) is transmitted from the base station to the UAV. The K interference source signals are recorded as

[0038] The signal receiving method of the present invention is:

[0039] Considering the case of single carrier narrowband, the received signal model can be written as

[0040]

[0041] in To receive the signal, represents the channel vector between the base station and the UAV, represents the channel vector between the kth interference source and the UAV, Represents the Gaussian white noise vector. In the above formula, δ f It represents the frequency offset between the base station and the UAV (hereinafter referred to as frequency offset), using represents the effect of frequency deviation on the target signal. τ represents the unknown delay between the transmitter and the receiver.

[0042] The present invention assumes that the signal channel h and the interference channel is unknown to the UAV receiver. At the same time, we also assume that the channel vector remains unchanged within a coherent time slot, and the target signal s(t) and the interference signals {i k (t)} are independent of each other.

[0043] The signal recovery method is as follows:

[0044] In the present invention, the preamble sequence is used as a reference signal. In the case where the time delay and frequency offset are unknown, a spatio-temporal filter with frequency offset compensation is designed to make the filtered signal as close as possible to the reference signal:

[0045]

[0046] where L>τ is the filter order. When there is a frequency offset, only using the coefficients independent of time cannot correct the frequency offset. Therefore, a time-varying phase e j2πω(t+l) is added in problem (2) to compensate for the frequency offset. Without loss of generality, assume

[0047] Define:

[0048]

[0049] Therefore, the optimization problem (2) can be written as

[0050]

[0051] In fact, the optimization problem (4) can be written more concisely as

[0052]

[0053] where:

[0054]

[0055] To obtain the required filter w, problem (5) must be solved. Since there is a coupling between the optimization variables ω and w, it is difficult to estimate ω and w simultaneously.

[0056] To solve this problem, first fix ω, then the least squares solution of w can be obtained:

[0057]

[0058] where denotes the pseudo-inverse of, because the length of the preamble sequence usually satisfies T<NL, so is a rank-deficient matrix.

[0059] Substitute w *Substituting into formula (5) we get

[0060]

[0061] Problem (9) only involves the optimization of ω. A one-dimensional search is used to obtain the optimal ω * According to the maximum moving speed and carrier frequency of the UAV in the actual communication scenario, the maximum possible frequency deviation δ can be calculated. max Without loss of generality, assume ω * ∈[δ min ,δ max Based on this, in [δ min ,δ max ] select m equidistant points, denoted as Calculate each ω i The corresponding f c (ω). Select f c (ω) The minimum value of ω i As the optimal ω * :

[0062]

[0063] Get the optimal ω * After that, the optimal filter can be obtained by formula (11)

[0064]

[0065] Next, we explain the sufficient conditions under which the solution to problem (5) can completely eliminate the interference signal and restore the target signal s(t).

[0066] A filter that can successfully restore the target signal s(t) must meet the following conditions:

[0067]

[0068] in Defined as

[0069]

[0070] To understand the above conditions, assume that ω = δ f is the true frequency deviation, and is a set of filters that satisfy condition (12). Applying this filter to the received signal yields

[0071]

[0072] From the above formula, we can see that the filtering result is exactly the desired target signal. Therefore, a set of {ω,w} that satisfies condition (12) can successfully filter out interference.

[0073] Regarding problem (5), it can be proved that when the length of the current guide code sequence satisfies the condition T>L(K+1), its solution must satisfy condition (12), that is, the solution to problem (5) must be a feasible filter.

[0074] In the method proposed by the present invention, it is necessary to calculate each The corresponding f c (ω). For each ω i The main amount of calculation is to calculate In order to more effectively reduce the amount of calculation, it considers the calculation Where ε is a very small non-negative number.

[0075] Notice for The tth row of It can be equivalently written as

[0076]

[0077] Based on the above structure, define

[0078]

[0079] Obviously, for

[0080]

[0081] Recursively calculating (17) from t′=1 to t′=T yields Recursively calculate It can improve time utilization - calculations can be performed whenever a set of samples is collected, and it can also reduce the overall computational complexity. Specifically, the complexity of the recursive method is O(mTN 2 L 2 ), where m represents m points in, T is the length of the preamble sequence, O(N 2 L 2 ) corresponds to the matrix-vector product in (17), that is, In contrast, directly The computational complexity of the inverse is O(mN 3 L 3 ). Since T is usually much smaller than NL, the computational complexity of the recursive method is also much smaller than the direct inversion.

[0082] The multi-antenna based UAV communication interference filtering method proposed in the present invention can be simplified as the following algorithm:

[0083] Signal transmission method:

[0084]

[0085]

[0086] The received signal structure is as follows Figure 2 As shown. Figure 2 As can be seen from the figure, the received signal does not contain the target signal within the delay t = [0:1:τ]. Whenever a set of samples is collected, steps 3 and 6 can be implemented. When the number of sampling points at the receiving end is an integer multiple of T, the process is completed. Calculate. According to steps 2 to 9, the optimal frequency offset compensation ω is obtained. * and the optimal filter w * . Then {ω * ,w *} is applied to the data bit to restore the corresponding frame length range t=[τ+1:1:τ+F i ],i∈{1,2,...∞} target signal. Among them, F i Indicates the number of sampling points in the i-th frame.

[0087] Simulation example:

[0088] In the simulation experiment, the number of receiver antennas N is set to 4, and the number of interference sources K is set to 3. The target signal adopts QPSK modulation. The specific signal processing flow is as follows: Figure 4 As shown. In signal processing, the root raised cosine finite impulse response filter is selected as the shaping filter and the matching filter. The length of the root raised cosine filter is 49 sampling points, the number of samples per symbol is 8, and the roll-off factor is set to 0.5. The frequency deviation between the transmitter and the receiver is set to 760Hz, and the unknown delay τ is randomly given in the range of [0.25us, 6.25us] (corresponding to [1, 25] sampling points at the receiving end).

[0089] The baseband communication signal is randomly generated by a 0 / 1 binary sequence. The symbol rate is set to 0.5Mb / s, the upsampling multiple is 8, and the sampling rate at the receiving end is 4MHz. The original sequence of each frame contains 164 binary numbers. The first 64 binary numbers are used to generate the preamble sequence, and the remaining binary numbers are used to generate the data sequence. Due to the symbol mapping and upsampling operations in the modulation process, there are 656 sampling points per frame. The preamble sequence contains 256 sampling points. The interference signal is a randomly generated Gaussian narrowband signal with a mean of zero, which is received at the receiving end together with the target signal. The signal channel and the interference channel are generated by Gaussian random vectors. In the present invention, we give the definition of SIR:

[0090]

[0091] from Figure 5It can be seen that the bit error rate of the method proposed in the present invention is significantly lower than that of the two-stage filtering method. Specifically, when the interference signal power is 30dB higher than the target signal power, it can still maintain good anti-interference performance. When the SIR increases from -30dB to -10dB, the bit error rate of the two-stage filtering method increases. This is because when the SIR increases, the minimum eigenvector of the covariance matrix of the received signal is no longer effective as an interference elimination filter.

[0092] In addition to simulation verification, the method proposed in the present invention is also tested based on a software radio hardware platform. Figure 3 As shown. Figure 4 In the experiment, the jammer is a single-antenna Keysight N5171B EXG signal generator; the transmitter is a single-antenna Agilent E4438C ESG vector signal transmitter; and the receiver is a dual-antenna NI USRP-2954 software radio. Signal processing operations are completed on the host computer. The QPSK modulation, symbol rate, and shaping / matching filter parameter settings are the same as the simulation experiment. The number of sampling points contained in each frame is still 656, of which the number of sampling points of the preamble sequence is 104. The frequency deviation between the transmitter and the receiver is 760Hz. The noise power received by the receiver is approximately -15dBm. Figure 6 It can be seen that the method proposed in the present invention achieves a lower bit error rate than the two-stage filtering method. When the interference signal power is greater than the target signal power by more than 30 dBm, the method proposed in the present invention can still maintain good performance.

[0093] In summary, the present invention proposes a practical anti-interference communication method based on multi-antenna UAVs. Through analysis, it can be seen that this method does not need to obtain any channel information and knowledge about the interference source, and can synchronously achieve time synchronization, frequency offset compensation and interference elimination using only the preamble sequence. The simulation results based on MATLAB and the test results based on the hardware experimental platform both show that this method is superior to the existing anti-interference methods.

Claims

1. A UAV co-frequency anti-interference communication method based on multi-antenna technology, wherein the communication scenario is set as downlink communication from a base station to a UAV, there are K unknown interference sources in the environment, the base station and the interference source have only a single antenna, and the UAV is equipped with N antennas, N>K; the characteristics are as follows: The communication method is: The base station periodically generates a preamble sequence {s(t), t = 1, ..., T}, where T is the length of the preamble sequence. The base station sends a frame of data composed of the preamble sequence and the data sequence. At the same time, the signals sent by K interference sources are defined as After receiving the signals from the base station and the interference source, the drone processes the signals. The specific method is as follows: Define a space-time filter with frequency offset compensation Where ω is the frequency offset compensation coefficient, w l is the filter coefficient, L>τ is the filter order, and the optimal frequency offset compensation coefficient and filter coefficient are calculated by receiving the signal: The received signal model is defined as: in, To receive the signal, represents the channel vector between the base station and the drone, represents the channel vector between the kth interference source and the UAV, represents the Gaussian white noise vector, δ f Indicates the frequency deviation between the base station and the drone, represents the effect of frequency deviation on the target signal, and τ represents the unknown delay between the transmitter and the receiver; The maximum possible frequency deviation δ is obtained based on the maximum moving speed and carrier frequency of the UAV max , define the minimum frequency deviation as δ min , in [δ min ,δ max ] select m equidistant points, denoted as Calculate each ω i The corresponding f c (ω), select f c (ω) The minimum value of ω i As the optimal ω * : Get the optimal ω * Afterwards, the optimal filter is obtained by the formula According to ω * and w * Perform data recovery and obtain the recovery signal: Among them, F i is the number of sampling points representing the i-th frame.

2. The method for co-frequency anti-interference communication of unmanned aerial vehicles based on multi-antenna technology according to claim 1 is characterized in that: The receiver optimizes the signal processing and The calculation is transformed into calculation Specifically, the definition is: Among them, ε is a very small non-negative number, I is the unit matrix, for The tth row of ; When the number of sampling points at the receiving end is an integer multiple of T, recursively calculate from t′=1 to t′=T: get Then we can get the optimal frequency offset compensation ω * and the optimal filter w * , {ω * ,w * } is applied to the data bit to restore the corresponding frame length range t=[τ+1:1:τ+F i ],i∈{1,2,...∞}, where Fi represents the number of sampling points in the i-th frame.

Images