Robust unmanned aerial vehicle array cooperative relay method

Abstract

The invention discloses a robust unmanned aerial vehicle array cooperative relay method, and the method comprises the following steps: S1, in a cooperative relay system, one unmanned aerial vehicle in an unmanned aerial vehicle array is a master unmanned aerial vehicle, and the other unmanned aerial vehicles are slave unmanned aerial vehicles; s2, the master unmanned aerial vehicle transmits a pilot frequency sequence c (n), and the ith slave unmanned aerial vehicle receives a pilot frequency sequence ri (n); s3, the ith slave unmanned aerial vehicle obtains time-frequency robust weight parameters wi, opt and # imgabs0 through stochastic gradient descent to minimize a mean square error; and S4, the unmanned aerial vehicle array performs collaborative relay with robustness. According to the invention, the improved cooperative relay under the condition that time and frequency deviation exists between the unmanned aerial vehicles is realized, and the problem that the cooperative relay performance of the unmanned aerial vehicle array is reduced due to misalignment of time and frequency is effectively solved.

Description

Technical Field

[0001] The present invention belongs to the field of cooperative relay, and in particular relates to a robust UAV array cooperative relay method. Background Art

[0002] Drone array collaborative relay technology refers to the cooperation of independent drones randomly distributed in space to form a virtual antenna array and perform collaborative relay, so that the transmitted signals of each drone are coherently superimposed at the destination, forming a beam pattern with a high-gain main lobe pointing to the destination, thereby enhancing the transmission distance and energy efficiency of the signal.

[0003] Collaborative relay requires time and frequency alignment between drones. However, in actual applications, due to hardware resolution limitations, random frequency drift of drones’ independent local oscillators, Doppler effects, and time-frequency algorithm synchronization errors, perfect time and frequency alignment between drones is almost impossible, resulting in a decrease in collaborative relay performance. Summary of the invention

[0004] The purpose of the present invention is to overcome the shortcomings of the prior art and provide a robust UAV array cooperative relay method, which realizes improved cooperative relay in the case of time and frequency offset between UAVs, effectively solves the problem of decreased coherent gain of UAV array under imperfect time and frequency alignment, and improves the cooperative relay performance.

[0005] The objective of the present invention is achieved through the following technical solution: A robust UAV array cooperative relay method, comprising the following steps:

[0006] S1. In the cooperative relay system, one drone in the drone array is the master drone, and the remaining drones are slave drones;

[0007] S2. The master drone transmits the pilot sequence c(n), and the pilot sequence r received by the i-th slave drone i (n);

[0008] S3. The i-th slave drone obtains the time-frequency robust weight parameter w by minimizing the mean square error through stochastic gradient descent i,opt and

[0009]

[0010] S4. UAV arrays perform robust cooperative relay.

[0011] In the step S1, a drone array consisting of N drones distributed in the xy plane is considered to form the collaborative relay system, wherein one drone is a master drone and the remaining N-1 drones are slave drones.

[0012] In step S2, the launch length of the main UAV is L c The pilot sequence c(n), n = 1,,Lc, n represents the sampling time. Since the channel h between the main UAV and the i-th (i = 2,,N) UAV i is a line-of-sight channel and the i-th slave UAV fully understands its relative position with the master UAV, so h i It can be expressed as:

[0013]

[0014] Among them, α0 is the channel power loss when the reference distance is 1m, β i is the distance between the master UAV and the i-th slave UAV, and λ is the carrier wavelength.

[0015] By multiplying After compensating for the channel effect, the pilot sequence r received by the i-th slave UAV is i (n) is:

[0016]

[0017] in, is the channel h i The conjugate of i is the time offset between the master UAV and the i-th slave UAV, is the frequency offset between the master UAV and the i-th slave UAV, f is the operating frequency of the UAV, and z i (n) is the additive white Gaussian noise from the i-th slave drone.

[0018] The step S3 includes the following sub-steps:

[0019] S301. Assume the time-frequency robust weight parameter w of the i-th slave drone i,opt and are L×1 dimensional vectors and 1×1 scalars respectively, where w i,k w i,opt The kth element of , k = 1,,L;

[0020] S302. Assume that the pilot signal received from the i-th drone forms a 1×L-dimensional vector r i (n), where r i (nj) is r i The j-th element of (n), j = -P,,P,L = 2P + 1;

[0021] S303. Assume that the error signal e in the time-frequency robust weight parameter calculation process of the i-th slave drone i (n) is:

[0022]

[0023] Among them, wi,opt(n-1) and are the time-frequency robust weight parameters w i,opt and The calculated value at the nth iteration, that is, the calculated value at the nth sampling moment;

[0024] S304. First, initialize w i,opt (0) = 0 and Then, the time-frequency robust weight parameter w is minimized by stochastic gradient descent. i,opt and The update formulas are:

[0025]

[0026] Among them, μ w and is the positive step length of stochastic gradient descent, (·)* and Im{·} represent the conjugate operation and the imaginary part operation respectively. The length of the pilot sequence is L c , through L c After iterations, the time-frequency robust weight parameter w of the i-th slave UAV is i,opt =w i,opt (L c ),

[0027] The step S4 comprises the following sub-steps:

[0028] S401. A tapped delay line structure is provided at each slave drone, including L-1 delays, L-1 adders and L multipliers, for eliminating the time offset between the slave drone and the master drone.

[0029] In the tapped delay line structure of the i-th slave drone, the input end of the first delay is the data to be transmitted by the slave drone. At the same time, the input end of the k+1th delay is connected to the input end of the k-th delay, k=1,,L-2; and the data at the input end of the k-th delay is multiplied by the k-th multiplier and the weight parameter w i,k Multiply, k = 1,,L-1, the data at the output of the L-1th delay is multiplied by the Lth multiplier and the weight parameter w i,L multiply; the first input end of the first adder is connected to the output end of the first multiplier, and the second input end is connected to the output end of the second multiplier. At the same time, the first input end of the kth adder is connected to the output end of the k-1th adder, and the second input end is connected to the output end of the k+1th multiplier, k=2,,L-1; the output of the L-1th adder is the output of the tapped delay line structure;

[0030] S402. A frequency offset compensation structure is provided at each slave drone, including a multiplier for eliminating the frequency offset between the slave drone and the master drone.

[0031] In the frequency offset compensation structure of the i-th slave drone, the data at the output of the tapped delay line structure is multiplied by a multiplier and a weight parameter multiply;

[0032] S403. Each drone in the drone array adjusts the phase of its transmitted signal by multiplying the corresponding element in the ideal steering vector to perform cooperative relay. The i-th (i=1,,N) element of the ideal steering vector for:

[0033]

[0034] Among them, (x i ,y i ) represents the position of the i-th UAV, θ and φ represent the pitch angle and azimuth angle of the UAV array, respectively, and (θ0, φ0) is the destination of the UAV array to perform cooperative relay.

[0035] The transmitting signal y1(n) of the master drone is:

[0036]

[0037] Among them, d1(n) is the data that the main UAV is ready to transmit, and its length is L d .

[0038] The transmission signal y of the i-th (i=2,,N) slave drone i (n) is:

[0039]

[0040] Among them, di(n) is the data that the i-th slave drone is ready to transmit, and its length is L d , di(n) is a 1×L-dimensional data vector, and the time-frequency robust weight parameter w i,opt and After multiplication, the time and frequency offset between it and the data d1(n) of the main drone is eliminated.

[0041] The beneficial effects of the present invention are as follows: the present invention realizes improved cooperative relaying in the case of time and frequency offset between UAVs, effectively solves the problem of decreased coherent gain of UAV arrays under imperfect time and frequency alignment, and improves the cooperative relay performance. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 Flow chart of the method of the present invention;

[0043] Figure 2 A schematic diagram of a drone array performing robust cooperative relaying in an embodiment;

[0044] Figure 3 Schematic diagram of the relationship between coherence gain and time offset in the embodiment;

[0045] Figure 4 FIG. 4 is a schematic diagram of the relationship between coherence gain and frequency offset in an embodiment. DETAILED DESCRIPTION

[0046] The technical solution of the present invention is further described in detail below in conjunction with the accompanying drawings, but the protection scope of the present invention is not limited to the following.

[0047] like Figure 1 As shown, a robust UAV array cooperative relay method comprises the following steps:

[0048] S1. In the cooperative relay system, one drone in the drone array is the master drone, and the remaining drones are slave drones;

[0049] S2. The master drone transmits the pilot sequence c(n), and the pilot sequence r received by the i-th slave drone i (n);

[0050] S3. The i-th slave drone obtains the time-frequency robust weight parameter w by minimizing the mean square error (MSE) through stochastic gradient descent i,opt and

[0051] S4. UAV arrays perform robust cooperative relay.

[0052] like Figure 2 FIG. 1 is a schematic diagram of a UAV array performing robust cooperative relay in an embodiment, as follows:

[0053] In the step S1, a drone array consisting of N drones distributed in the xy plane is considered to form the collaborative relay system, wherein one drone is a master drone and the remaining N-1 drones are slave drones.

[0054] In step S2, the launch length of the main UAV is L c The pilot sequence c(n), n = 1,,Lc, n represents the sampling time. Since the channel h between the main UAV and the i-th (i = 2,,N) UAV i is a line-of-sight channel and the i-th slave UAV fully understands its relative position with the master UAV, so h i It can be expressed as:

[0055]

[0056] Among them, α0 is the channel power loss when the reference distance is 1m, β i is the distance between the master UAV and the i-th slave UAV, and λ is the carrier wavelength.

[0057] By multiplying After compensating for the channel effect, the pilot sequence r received by the i-th slave UAV is i (n) is:

[0058]

[0059] in, is the channel h i The conjugate, D i is the time offset between the master UAV and the i-th slave UAV, is the frequency offset between the master UAV and the i-th slave UAV, f is the operating frequency of the UAV, and z i (n) is the additive white Gaussian noise from the i-th slave drone.

[0060] The step S3 includes the following sub-steps:

[0061] S301. Assume the time-frequency robust weight parameter w of the i-th slave drone i,opt and are L×1 dimensional vectors and 1×1 scalars respectively, where w i,k w i,opt The kth element of , k = 1,,L;

[0062] S302. Assume that the pilot signal received from the i-th drone forms a 1×L-dimensional vector r i (n), where r i (nj) is r i The j-th element of (n), j = -P,,P,L = 2P + 1;

[0063] S303. Assume that the error signal e in the time-frequency robust weight parameter calculation process of the i-th slave drone i (n) is:

[0064]

[0065] Among them, wi,opt(n-1) and are the time-frequency robust weight parameters w i,opt and The calculated value at the nth iteration, that is, the calculated value at the nth sampling moment;

[0066] S304. First, initialize wi,opt (0) = 0 and Then, the time-frequency robust weight parameter w is minimized by stochastic gradient descent. i,opt and The update formulas are:

[0067]

[0068]

[0069] Among them, μ w and is the positive step length of stochastic gradient descent, (·)* and Im{·} represent the conjugate operation and the imaginary part operation respectively. The length of the pilot sequence is L c , through L c After iterations, the time-frequency robust weight parameter w of the i-th slave UAV is i,opt =w i,opt (L c ),

[0070] The step S4 comprises the following sub-steps:

[0071] S401. A tapped delay line structure is provided at each slave drone, including L-1 delays, L-1 adders and L multipliers, for eliminating the time offset between the slave drone and the master drone.

[0072] In the tapped delay line structure of the i-th slave drone, the input end of the first delay is the data to be transmitted by the slave drone. At the same time, the input end of the k+1th delay is connected to the input end of the k-th delay, k=1,,L-2; and the data at the input end of the k-th delay is multiplied by the k-th multiplier and the weight parameter w i,k Multiply, k = 1,,L-1, the data at the output of the L-1th delay is multiplied by the Lth multiplier and the weight parameter w i,L multiply; the first input end of the first adder is connected to the output end of the first multiplier, and the second input end is connected to the output end of the second multiplier. At the same time, the first input end of the kth adder is connected to the output end of the k-1th adder, and the second input end is connected to the output end of the k+1th multiplier, k=2,,L-1; the output of the L-1th adder is the output of the tapped delay line structure;

[0073] S402. A frequency offset compensation structure is provided at each slave drone, including a multiplier for eliminating the frequency offset between the slave drone and the master drone.

[0074] In the frequency offset compensation structure of the i-th slave drone, the data at the output of the tapped delay line structure is multiplied by a multiplier and a weight parameter multiply;

[0075] S403. Each drone in the drone array adjusts the phase of its transmitted signal by multiplying the corresponding element in the ideal steering vector to perform cooperative relay. The i-th (i=1,,N) element of the ideal steering vector for:

[0076]

[0077] Among them, (x i ,y i ) represents the position of the i-th UAV, θ and φ represent the pitch angle and azimuth angle of the UAV array, respectively, and (θ0, φ0) is the destination of the UAV array to perform cooperative relay.

[0078] The transmission signal y1(n) of the master drone is:

[0079]

[0080] Among them, d1(n) is the data that the main UAV is ready to transmit, and its length is L d .

[0081] The transmission signal y of the i-th (i=2,,N) slave drone i (n)

[0082]

[0083] Among them, di(n) is the data that the i-th slave drone is ready to transmit, and its length is L d , di(n) is a 1×L-dimensional data vector, and the time-frequency robust weight parameter w i,opt and After multiplication, the time and frequency offset between it and the data d1(n) of the main drone is eliminated.

[0084] The proposed robust UAV array cooperative relay method is simulated, analyzed and evaluated below. The specific parameter settings are as follows:

[0085]

[0086] Figure 3 The relationship between the time offset between drones and the coherent gain of the drone array at the destination after robust cooperative relay is shown under different tap numbers of the tapped delay line structure, where the frequency offset For 10 -4When the time offset D is 10 times the sampling period, compared with the traditional cooperative relay, the proposed robust UAV array cooperative relay (L=21) increases the coherence gain at the destination from 12dB to 24dB.

[0087] Figure 4 The relationship between the frequency offset between UAVs and the coherent gain of the UAV array at the destination after robust cooperative relay is given under different tap numbers of the tapped delay line structure, where the time frequency shift D is 5 times the sampling period. For 10 -2 When the sampling frequency is times that of traditional cooperative relay, the proposed robust UAV array cooperative relay (L=21) increases the coherence gain at the destination from 12dB to 24dB compared with the traditional cooperative relay.

[0088] The present invention has been described in detail herein by way of specific embodiments, and the description of the above embodiments is provided to enable those skilled in the art to make or apply the present invention, and various modifications of these embodiments will be readily understood by those skilled in the art. The present invention is not limited to these examples, or certain aspects thereof. The scope of the present invention is described in detail by the appended claims.

[0089] The above description shows and describes a preferred embodiment of the present invention, but as mentioned above, it should be understood that the present invention is not limited to the form disclosed herein, and should not be regarded as excluding other embodiments, but can be used in various other combinations, modifications and environments, and can be modified within the scope of the invention concept described herein through the above teachings or the technology or knowledge of the relevant field. And the changes and modifications made by those skilled in the art do not depart from the spirit and scope of the present invention, and should be within the scope of protection of the claims attached to the present invention.

Claims

1. A robust UAV array cooperative relay method, characterized by: The following steps are involved: S1. In the cooperative relay system, one drone in the drone array is the master drone, and the remaining drones are slave drones; S2. The master drone transmits the pilot sequence c(n), and the pilot sequence r received by the i-th slave drone i (n); S3. The i-th slave drone obtains the time-frequency robust weight parameter w by minimizing the mean square error through stochastic gradient descent i,opt and S4. UAV arrays perform robust cooperative relay.

2. A robust UAV array cooperative relay method according to claim 1, characterized in that: In the step S1, a drone array consisting of N drones distributed in the xy plane is considered to form the collaborative relay system, wherein one drone is a master drone and the remaining N-1 drones are slave drones.

3. The robust UAV array cooperative relay method according to claim 1, characterized in that: In step S2, the launch length of the main UAV is L c The pilot sequence c(n), n=1,...,L c , n represents the sampling time, since the channel h between the main UAV and the i-th (i=2,...,N) UAV i is a line-of-sight channel and the i-th slave UAV fully understands its relative position with the master UAV, so h i It is expressed as: Among them, α0 is the channel power loss when the reference distance is 1m, β i is the distance between the master UAV and the i-th slave UAV, λ is the carrier wavelength; By multiplying After compensating for the channel effect, the pilot sequence r received by the i-th slave UAV is i (n) is: in, is the channel h i The conjugate of i is the time offset between the master UAV and the i-th slave UAV, is the frequency offset between the master UAV and the i-th slave UAV, f is the operating frequency of the UAV, and z i (n) is the additive white Gaussian noise from the i-th slave drone.

4. The method for robust UAV array cooperative relay according to claim 1, characterized in that: The step S3 comprises the following sub-steps: S301. Assume the time-frequency robust weight parameter w of the i-th slave drone i,opt and are L×1 dimensional vectors and 1×1 scalars respectively, where w i,k w i,opt The kth element of , k = 1,…,L; S302. Assume that the pilot signal received from the i-th drone forms a 1×L-dimensional vector r i (n), where r i (nj) is r i The j-th element of (n), j = -P, ..., P, L = 2P + 1; S303. Assume that the error signal e in the time-frequency robust weight parameter calculation process of the i-th slave drone i (n) is: Among them, w i,opt (n-1) and are the time-frequency robust weight parameters w i,opt and The calculated value at the nth iteration, that is, the calculated value at the nth sampling moment; S304. First, initialize w i,opt (0) = 0 and Then, the time-frequency robust weight parameter w is minimized by stochastic gradient descent. i,opt and The update formulas are: Among them, μ w and is the positive step size of stochastic gradient descent, (·) * and Im{·} represent the conjugate operation and the imaginary part operation respectively; the length of the pilot sequence is L c , through L c After iterations, the time-frequency robust weight parameter w of the i-th slave UAV is i,opt =w i,opt (L c ), 5. The robust UAV array cooperative relay method according to claim 1, characterized in that: The step S4 comprises the following sub-steps: S401. A tapped delay line structure is provided at each slave drone, comprising L-1 delays, L-1 adders and L multipliers, for eliminating the time offset between the slave drone and the master drone; In the tapped delay line structure of the i-th slave drone, the input end of the first delay is the data to be transmitted by the slave drone, and the input end of the k+1-th delay is connected to the input end of the k-th delay, k=1,...,L-2; and the data at the input end of the k-th delay is multiplied by the k-th multiplier and the weight parameter w i,k Multiply, k = 1, ..., L-1, the data at the output of the L-1th delay is multiplied by the Lth multiplier and the weight parameter w i,L multiply; the first input end of the first adder is connected to the output end of the first multiplier, and the second input end is connected to the output end of the second multiplier, and at the same time, the first input end of the kth adder is connected to the output end of the k-1th adder, and the second input end is connected to the output end of the k+1th multiplier, k=2,...,L-1; the output of the L-1th adder is the output of the tapped delay line structure; S402. A frequency offset compensation structure is provided at each slave drone, including a multiplier for eliminating the frequency offset between the slave drone and the master drone; In the frequency offset compensation structure of the i-th slave drone, the data at the output of the tapped delay line structure is multiplied by a multiplier and a weight parameter multiply; S403. Each drone in the drone array adjusts the phase of its transmitted signal by multiplying the corresponding element in the ideal steering vector to perform cooperative relay. The i-th (i=1,…,N) element of the ideal steering vector for: Among them, (x i ,y i ) represents the position of the i-th UAV, θ and φ represent the pitch angle and azimuth angle of the UAV array, respectively, and (θ0, φ0) is the destination of the UAV array to perform cooperative relay; The transmission signal y1(n) of the master drone is: Among them, d1(n) is the data that the main UAV is ready to transmit, and its length is L d ; The transmission signal y of the i-th (i=2,…,N) slave drone i (n) is: Among them, di(n) is the data that the i-th slave drone is ready to transmit, and its length is L d , d i (n) is a 1×L-dimensional data vector and a time-frequency robust weight parameter w i,opt and After multiplication, the time and frequency offset between it and the data d1(n) of the main drone is eliminated.

Images