阅读说明：本技术 无人机抗干扰通信系统及联合轨迹与波束成形优化方法 (Unmanned aerial vehicle anti-interference communication system and joint track and beam forming optimization method ) 是由 陈瑾 侯志峰 罗屹洁 方贵 谷江春 徐逸凡 于 2021-08-25 设计创作，主要内容包括：无人机抗干扰通信系统及联合轨迹与波束成形优化方法,属于无线通信抗干扰技术领域,考虑一个抗全频段强功率干扰的通信网络,该网络内基站为移动用户提供通信服务,同时一个全频段的全向干扰机以功率压制的方式对于合法通信链路进行干扰。该干扰机具备一定的隐蔽性,其信息对于合法通信双方无法准确获取。利用无人机感知环境获取状态信息。通过ε-greedy探索利用策略,依概率选择由神经网络估计的最佳回报值动作。智能反射无人机通过执行方案得到该动作的真实回报值,并将经验信息存储于滑动存储窗中。在多步经验池中的经验信息达到一定数量之后,神经网络进行更新,通过神经网络拟合环境后进行最优决策。能够很好地应用于实时抗干扰通信场景。(An unmanned aerial vehicle anti-interference communication system and a joint track and beam forming optimization method belong to the technical field of wireless communication anti-interference, a communication network resisting full-band strong power interference is considered, a base station in the network provides communication service for mobile users, and meanwhile, a full-band omnidirectional jammer interferes a legal communication link in a power suppression mode. The jammer has certain concealment, and information of the jammer cannot be accurately acquired by both legal communication parties. And acquiring state information by using the unmanned aerial vehicle to sense the environment. And selecting the action of the optimal return value estimated by the neural network according to the probability by searching and utilizing the strategy through epsilon-greedy. The intelligent reflection unmanned aerial vehicle obtains a real return value of the action through an execution scheme, and stores experience information in a sliding storage window. And after the experience information in the multi-step experience pool reaches a certain amount, updating the neural network, and fitting the environment through the neural network to perform optimal decision. The method can be well applied to a real-time anti-interference communication scene.)

1. An unmanned aerial vehicle anti-interference communication system is characterized by comprising an anti-interference communication network under the condition of full-band strong power interference, wherein a base station for carrying out communication coverage service on a mobile user is arranged in the network; the power of omnidirectional interference machine and the dual suppression characteristic of frequency channel carry on intelligent reflection surface reinforcing useful signal/offset interference signal through deploying unmanned aerial vehicle and improve the anti-interference communication performance of system, through the unmanned aerial vehicle motion trail with intelligent reflection surface passive beam forming's joint optimization, promote the anti-interference communication ability of the unknown environment of developments.

2. The unmanned aerial vehicle anti-jamming communication system of claim 1, wherein within a topological state, the signal received by mobile user D is defined as:

wherein, P_SFor base station transmitting power, P_JIs the jammer interference power, H_S、H_J、h_SDAnd h_JDThe gain of a direct link channel between a base station S and a mobile user D, the gain of a direct link channel between an interference machine J and the mobile user D, the combined channel gain of a signal of the base station S which is reflected by an intelligent reflection surface U and reaches the mobile user D, and the combined channel gain of a signal of the interference machine J which is reflected by the intelligent reflection surface U and reaches the mobile user D, x_SFor transmitting signals to the base station, x_JFor jammers to transmit signals, z is white gaussian noise received by mobile user D, with a obedient mean of 0 and a variance of δ²(ii) a gaussian distribution of;

the signal-to-interference-and-noise ratio of the mobile user D in the topological state is specifically represented as:

the achievable communication rate received by mobile user D is expressed as:

the achievable communication rate at the mobile user side is affected by the combination of the position of the intelligent reflective surface and the passive beamforming.

3. The anti-jamming communication system for unmanned aerial vehicles according to claim 1, wherein the channel of the intelligent reflective surface is strongly coupled to the beam forming direction, so that the user reception channel after reflection by the intelligent reflective surface is specifically represented as:

wherein G is_t,G_rAnd G is the antenna gain of the transmitting end, the receiving end and the reflection unit of the intelligent reflection surface, M and K are the number of rows and columns of the reflection unit contained in the intelligent reflection surface, and d_xAnd d_yIs the row and column width of the reflective element, d₁And d₂Is the distance, θ, between the intelligent reflective surface and the transmitting and receiving ends_tAnd theta_rIs the angle between the transmitting end signal and the receiving end signal and the x axis,andfor the transmission and reception of end signalsAngle, a is the adjustment of the intelligent reflective surface to the amplitude of the signal, λ is the wavelength of the signal,is the normalized power radiation function of the reflecting element,is a beamforming direction offset function, which reflects the signal amplitude attenuation caused by the beamforming direction offset, and is specifically expressed as:

wherein the content of the first and second substances,φ_k,mis the phase variation value corresponding to the first row and column of the reflective unit, when the beam forming is completely aligned with the receiving end, the phase variation value of the reflective unit can be expressed as

And according to the coupling relation between the phase of the intelligent reflection surface and the beam forming direction, the algorithm complexity of phase design is reduced through the beam forming direction.

4. The joint trajectory and beam forming optimization method of the unmanned aerial vehicle anti-interference communication system based on claim 1, wherein in the communication process of the sender and the intelligent reflection surface, state sensing is firstly performed, then an optimization algorithm based on deep reinforcement learning is executed, optimization of position deployment and beam forming is performed according to probability, and the probability is adjusted according to a feedback value after communication.

5. The joint trajectory and beam forming optimization method of claim 4, characterized in that the optimization algorithm comprises the steps of:

step 1, modeling the anti-interference communication problem into a Markov decision process by designing a state, an action, a return value and a transition probability function

And 2, initializing a system scene and a neural network, and updating parameters of the neural network according to state collection, action decision and feedback.

And 3, recording the path after the algorithm is executed and the beam forming scheme.

6. The joint trajectory and beam forming optimization method according to claim 5, characterized in that in step 1, through the design of the passive beam forming model, the motion state space of the phase change optimization algorithm of the large-scale intelligent reflection surface is reduced, which is beneficial to the enhancement of the reinforcement learning convergence speed in the path planning process, and further improves the reaction speed of the algorithm and the adaptability under the anti-interference scene.

7. The joint trajectory and beam forming optimization method of claim 5, wherein in step 1, the passive beam forming design that the intelligent reflection surface can perform after sensing the channel state is divided into two cases: enhancing the useful signal or reducing the interference signal;

the useful signal is enhanced, namely the phase between the signal which reaches the receiving end of the user after the transmitting end is reflected by the intelligent reflecting surface and the direct signal of the transmitting end and the receiving end is the same, and the strength of the useful signal after the receiving end superposes the signal is far greater than that of the interference signal, so that the aim of resisting interference is fulfilled;

reduce interfering signal, no matter how big power interfering signal interferes with, intelligence reflecting surface carries out refocusing at the receiving terminal after the signal that sends with the interference reflects, but the signal is through reverse phase processing, and interfering signal and the interfering signal of reflection of penetrating directly at the receiving terminal can offset, for useful signal, interfering signal obtains reducing, and anti-interference communication effect obtains promoting.

8. The joint trajectory and beamforming optimization method according to claim 5, wherein in step 2, the transmitting end obtains the position and channel state information of the mobile user through channel sensing and the state information communication link of the unmanned aerial vehicle, and directly corrects the beamforming direction from the transmitting end to the intelligent reflective surface end under the condition of a given beamforming direction; because the information of the jammer is unknown, the information of the jammer is explored in a deep reinforcement learning interaction mode, and an optimal intelligent reflector passive beam forming scheme and an unmanned aerial vehicle flight track are searched by utilizing the real-time control of the beam forming direction.

9. The joint trajectory and beam forming optimization method of claim 8, wherein the deep reinforcement learning algorithm comprises the steps of:

initializing, setting the iteration number to be i-1, setting a state-action-report storage space, initializing parameters of a neural network, and exploring-utilizing a probability epsilon;

step 11, the intelligent reflection surface-unmanned aerial vehicle interacts with the environment to obtain the state s at the moment i_i；

And step 12, selecting the motion of the motion trail and the passive beam forming phase design by the intelligent reflecting surface-unmanned aerial vehicle according to the probability. Exploration-utilization strategy through epsilon-greeny, which is specifically expressed as:

wherein, a_iFor the selection of the action of the intelligent reflecting surface-unmanned plane at time i, s_iFor the Intelligent reflective surface-the State of the drone at time i, Q(s)_i,a_i) For intelligent reflective surfaces-unmanned aerial vehicle at s_iIs made in a state of_iThe return value after the action is according to formula (7) at s_iUnder the state of (1), the intelligent reflecting surface has the action of selecting the probability of epsilon to ensure that the return value is maximum, and has the action of randomly selecting the probability of 1-epsilon; intelligent reflective surface by gradually increasing the value of epsilon in an iterative process-unmannedThe machine seeks balance between exploration and utilization, and an optimal scheme is obtained quickly under the condition of exploring all actions as much as possible;

step 13, perform action a_iThe next time state s is reached_i+1And obtaining the return value, and sending the state-action-return value-next-time state [ s ]_i,a_i,r_i,s_i+1]Storing in a sliding memory window;

step 14, let i equal i +1 and s_i＝s_i+1Repeating the steps 11-14 until the sliding memory window is filled with historical experience data;

step 15, after the sliding storage window is filled with the historical empirical data, calculating the multistep empirical data, wherein the multistep return value is specifically represented as:

when the gamma is 1, the influence of the unreturned value is fully considered, but the excessive oscillation of the reported value is caused, so that the convergence effect is influenced; when gamma is 0, only the influence of the reported value at the next moment is considered, the algorithm decision speed is high, but the grasp on the future trend is lost;

after obtaining the multi-step reward value, the multi-step experience information s_i,a_i:i+N,s_i+N,r_i:i+N]Stored in an experience pool for further updating of the neural network. At the same time, new s is continuously acquired_i,a_i,r_i,s_i+1]When the information is stored in the sliding memory window, the old information is discarded, and new multi-step experience information s is obtained continuously through the sliding memory window_i,a_i:i+N,s_i+N,r_i:i+N]Continuously storing the data into an experience pool;

step 16, when the multi-step experience information in the experience pool reaches a certain amount, updating the neural network, otherwise, repeating the steps 11-15;

and step 17, updating the neural network parameters. And the value of epsilon is continuously increased during the iteration process. So that the intelligent reflective surface-unmanned aerial vehicle makes action selection with a large return value in a trending manner.

Technical Field

The invention relates to the technical field of wireless communication anti-interference, in particular to an unmanned aerial vehicle anti-interference communication system and a joint track and beam forming optimization method.

Background

When the full-band interference with strong power exists in a scene, the realization of reliable and robust information transmission is an important research direction. Due to the double suppression of interference in the frequency domain and the power domain, the common anti-interference means is difficult to solve. The intelligent reflection surface has the advantages of passive reflection, flexible phase adjustment, convenient deployment and the like as a new technology, and has wide prospect in solving the problem of anti-interference communication in the scene. Through combining the unmanned aerial vehicle technique, with intelligent reflection surface deployment on unmanned aerial vehicle, through adjusting the unmanned aerial vehicle position in real time, in the dynamic environment, the interference immunity of system can obtain further promotion.

Disclosure of Invention

The invention provides an unmanned aerial vehicle anti-interference communication system and a joint track and beam forming optimization method, which are well applied to a real-time anti-interference communication network scene to realize efficient anti-interference communication under the condition of full-band strong-power interference. On the basis of a deep reinforcement learning algorithm, the passive beam forming action decision space is greatly reduced by utilizing the phase change of the intelligent reflection surface and the high coupling of beam forming, and meanwhile, multi-step experience citation is introduced into the trajectory planning, so that the accuracy of path decision is further improved through the analysis of multi-step paths.

The technical solution for realizing the purpose of the invention is as follows: an anti-interference communication model under a dynamic unknown environment is characterized in that: consider a communication network that is resistant to full-band strong power interference, in which there is a base station for performing communication coverage service for mobile users, but a full-band omni-directional jammer interferes with legitimate communications in a power-suppressing manner. The jammer has certain concealment, and information of the jammer cannot be accurately acquired by both legal communication parties.

In a certain topological state, the signal received by the mobile user D is defined as:

wherein, P_SFor base station transmitting power, P_JIs the jammer interference power, H_S、H_J、h_SDAnd h_JDThe gain of a direct link channel between a base station S and a mobile user D, the gain of a direct link channel between an interference machine J and the mobile user D, the combined channel gain of a signal of the base station S which is reflected by an intelligent reflection surface U and reaches the mobile user D, and the combined channel gain of a signal of the interference machine J which is reflected by the intelligent reflection surface U and reaches the mobile user D are obtained. x is the number of_SFor transmitting signals to the base station, x_JFor jammers to transmit signals, z is white gaussian noise received by mobile user D, obeying a mean of 0 and a variance of δ²A gaussian distribution of (a). Therefore, the signal to interference plus noise ratio of the mobile user D in a certain topological state is specifically expressed as:

thus, the achievable communication rate received by mobile user D is expressed as:

considering that the unmanned aerial vehicle motion trajectory and the intelligent reflection surface passive beam forming are optimized in a combined mode, wherein a strong coupling relation exists between an intelligent reflection surface channel and a beam forming direction, and therefore a user receiving channel after being reflected by the intelligent reflection surface is specifically represented as follows:

wherein G is_t,G_rAnd G is the antenna gain of the transmitting end, the receiving end and the intelligent reflection surface reflection unit, M and K are the number of rows and columns of the intelligent reflection surface including the reflection unit, and d_xAnd d_yIs the row and column width of the reflective element, d₁And d₂Is the distance, θ, between the intelligent reflective surface and the transmitting and receiving ends_tAnd theta_rIs the angle between the transmitting end signal and the receiving end signal and the x axis,andis the included angle between the transmitting end signal and the receiving end signal and the z axis, A is the adjustment value of the intelligent reflecting surface to the signal amplitude, lambda is the wavelength of the signal,is the normalized power radiation function of the reflecting element,is a beamforming direction offset function, which reflects the signal amplitude attenuation caused by beamforming direction offset, and is specifically expressed as:

wherein the content of the first and second substances,φ_k,mis the phase variation value corresponding to the first row and column of the reflective unit, when the beam forming direction is completely aligned with the receiving end, the phase variation value of the reflective unit can be expressed as

According to the coupling relation between the intelligent reflecting surface phase and the beam forming direction, the algorithm complexity of phase design can be reduced through the beam forming direction.

In the communication process of the sender and the intelligent reflection surface, firstly state perception is carried out, then an optimization algorithm based on deep reinforcement learning is executed, optimization of position deployment and beam forming is carried out according to probability, and the probability is adjusted according to a feedback value after communication. Wherein, the optimization algorithm comprises the following steps:

step 1, modeling the anti-interference communication problem into a Markov decision process by designing a state, an action, a return value and a transition probability function

And 2, initializing a system scene and a neural network, and updating parameters of the neural network according to state collection, action decision and feedback.

And 3, recording the path after the algorithm is executed and the beam forming scheme.

According to the invention, through the design of the passive beam forming model, the action state space of the phase change optimization algorithm of the large-scale intelligent reflection surface is reduced, the enhancement of the reinforcement learning convergence speed in the path planning process is facilitated, and the reaction speed of the algorithm and the adaptability under an anti-interference scene are further improved.

The transmitting end of the invention obtains the position and the channel state information of the mobile user through the channel perception and the state information communication link of the unmanned aerial vehicle, and can directly correct the beam forming direction from the transmitting end to the intelligent reflecting surface end under the condition of setting the beam forming direction. On the other hand, because the information of the jammer is unknown, the information of the jammer is explored in a deep reinforcement learning interaction mode, and the optimal intelligent reflector passive beam forming scheme and the flight track of the unmanned aerial vehicle are searched by utilizing the real-time control of the beam forming direction.

The passive beam forming design which can be executed by the intelligent reflecting surface after the channel state is sensed can be divided into two conditions: the useful signal is enhanced and the interference signal is reduced.

The useful signal is enhanced, namely the phase between the signal which reaches the receiving end of the user after the transmitting end is reflected by the intelligent reflection surface and the direct signal of the transmitting end and the receiving end is the same, the strength of the useful signal after the receiving end superposes the signal is far greater than that of the interference signal, and therefore the anti-interference purpose is achieved.

Reduce interfering signal, no matter how big power interfering signal interferes with, intelligence reflecting surface carries out refocusing at the receiving terminal after the signal that sends with the interference reflects, but the signal is through reverse phase processing, consequently the interfering signal of direct incidence and the interfering signal of reflection at the receiving terminal can offset, for useful signal, interfering signal obtains reducing, anti-interference communication effect obtains promoting.

The invention provides an intelligent reflection surface unmanned aerial vehicle anti-interference communication system and a joint track and beam forming optimization method based on deep reinforcement learning, and aims to provide a scheme for realizing efficient anti-interference communication under the condition of full-band strong-power interference. On the basis of a deep reinforcement learning algorithm, the passive beam forming action decision space is greatly reduced by utilizing the phase change of the intelligent reflection surface and the high coupling of beam forming, and meanwhile, multi-step experience citation is introduced into the trajectory planning, so that the accuracy of path decision is further improved through the analysis of multi-step paths.

Drawings

Fig. 1 is a network schematic diagram of an unmanned aerial vehicle anti-jamming communication system.

Fig. 2 is a schematic diagram of a trajectory and beam forming joint optimization method based on deep reinforcement learning.

Fig. 3 is a schematic topology diagram in embodiment 1 of the present invention.

FIG. 4 is a graph comparing the performance of the algorithm of the present patent with other algorithms

Fig. 5 is a graph comparing the performance of the algorithm of the present patent for different jammer locations.

Detailed Description

The technical scheme of the invention is described in detail in the following with reference to the attached drawings:

as shown in fig. 1, an anti-interference communication system for an unmanned aerial vehicle considers a communication network resistant to full-band strong power interference, where a base station exists in the communication network to perform communication coverage service for a mobile user, but a full-band omni-directional jammer interferes with legal communication in a power-suppressing manner. This jammer possesses certain disguise, and its information can't accurately acquire to legal both sides of communication, in order to reach maximum throughput, through jointly adjusting intelligent reflection surface unmanned aerial vehicle's position and the passive beam forming of intelligent reflection surface, fully explores the anti-interference efficiency of this system.

Further, in a certain topological state, the signal received by the mobile user D is defined as:

wherein, P_SFor base station transmitting power, P_JIs the jammer interference power, H_S、H_J、h_SDAnd h_JDThe gain of a direct link channel between a base station S and a mobile user D, the gain of a direct link channel between an interference machine J and the mobile user D, the combined channel gain of a signal of the base station S which is reflected by an intelligent reflection surface U and reaches the mobile user D, and the combined channel gain of a signal of the interference machine J which is reflected by the intelligent reflection surface U and reaches the mobile user D are obtained. x is the number of_SFor transmitting signals to the base station, x_JFor jammers to transmit signals, z is white gaussian noise received by mobile user D, obeying a mean of 0 and a variance of δ²A gaussian distribution of (a). Therefore, the signal to interference plus noise ratio of the mobile user D in a certain topological state is specifically expressed as:

thus, the achievable communication rate received by mobile user D is expressed as:

considering that the unmanned aerial vehicle motion trajectory and the intelligent reflection surface passive beam forming are optimized in a combined mode, wherein a strong coupling relation exists between an intelligent reflection surface channel and a beam forming direction, and therefore a user receiving channel after being reflected by the intelligent reflection surface is specifically represented as follows:

wherein G is_t,G_rAnd G is the antenna gain of the transmitting end, the receiving end and the intelligent reflection surface reflection unit, M and K are the number of rows and columns of the intelligent reflection surface including the reflection unit, and d_xAnd d_yIs the row and column width of the reflective element, d₁And d₂Is the distance, θ, between the intelligent reflective surface and the transmitting and receiving ends_tAnd theta_rIs the angle between the transmitting end signal and the receiving end signal and the x axis,andis the included angle between the transmitting end signal and the receiving end signal and the z axis, A is the adjustment value of the intelligent reflecting surface to the signal amplitude, lambda is the wavelength of the signal,is the normalized power radiation function of the reflecting element,is a beamforming direction offset function, which reflects the signal amplitude attenuation caused by beamforming direction offset, and is specifically expressed as:

wherein the content of the first and second substances,φ_k,mis a first row and column reflection unitThe corresponding phase variation value, when the beam forming direction is completely aligned with the receiving end, the phase variation value of the reflection unit can be expressed as

According to the coupling relation between the intelligent reflecting surface phase and the beam forming direction, the algorithm complexity of phase design can be reduced through the beam forming direction.

As shown in fig. 2, the specific process of the deep reinforcement learning-based joint trajectory and beamforming optimization method of the present invention is as follows: in the communication process of the sender and the intelligent reflection surface, firstly state perception is carried out, then an optimization algorithm based on deep reinforcement learning is executed, optimization of position deployment and beam forming is carried out according to probability, and the probability is adjusted according to a feedback value after communication. Wherein, the optimization algorithm comprises the following steps:

step 1, modeling the anti-interference communication problem into a Markov decision process by designing a state, an action, a return value and a transition probability function

And 2, initializing a system scene and a neural network, and updating parameters of the neural network according to state collection, action decision and feedback.

And 3, recording the path after the algorithm is executed and the beam forming scheme.

In the step 1, through the design of a passive beam forming model, the action state space of the phase change optimization algorithm of the large-scale intelligent reflection surface is reduced, the enhancement of the reinforcement learning convergence speed in the path planning process is facilitated, and the reaction speed of the algorithm and the adaptability of the algorithm in an anti-interference scene are further improved. In step 2, the transmitting end obtains the position and channel state information of the mobile user through channel sensing and a state information communication link of the unmanned aerial vehicle, and under the condition of a set beam forming direction, the beam forming direction from the transmitting end to the intelligent reflection surface end can be directly corrected. On the other hand, because the information of the jammer is unknown, the information of the jammer is explored in a deep reinforcement learning interaction mode, and the optimal intelligent reflector passive beam forming scheme and the flight track of the unmanned aerial vehicle are searched by utilizing the real-time control of the beam forming direction.

In the deep reinforcement learning algorithm, the method comprises the following steps:

initialization, setting the number of iterations to i-1, setting state-action-reporting storage space, neural network initialization parameters, and exploration-utilization probability epsilon.

Step 11, the intelligent reflection surface-unmanned aerial vehicle interacts with the environment to obtain the state s at the moment i_i。

And step 12, selecting the motion of the motion trail and the passive beam forming phase design by the intelligent reflecting surface-unmanned aerial vehicle according to the probability. Exploration-utilization strategy through epsilon-greeny, which is specifically expressed as:

wherein, a_iFor the selection of the action of the intelligent reflecting surface-unmanned plane at time i, s_iFor the Intelligent reflective surface-the State of the drone at time i, Q(s)_i,a_i) For intelligent reflective surfaces-unmanned aerial vehicle at s_iIs made in a state of_iThe return value after the action is according to formula (7) at s_iThe intelligent reflecting surface has the action with the probability of epsilon to select the action with the maximum return value, and has the probability of 1-epsilon to randomly select the action. By gradually increasing the value of epsilon in the iterative process-the drone can seek a balance between exploration and exploitation, getting the optimal solution faster, exploring all the actions as much as possible.

Step 13, perform action a_iThe next time state s is reached_i+1And obtaining the return value, and sending the state-action-return value-next-time state [ s ]_i,a_i,r_i,s_i+1]Stored in a sliding memory window.

Step 14, let i equal i +1 and s_i＝s_i+1And repeating the steps 11-14 until the sliding memory window is filled with historical experience data.

Step 15, after the sliding storage window is filled with the historical empirical data, calculating the multistep empirical data, wherein the multistep return value is specifically represented as:

where N is the length of the sliding storage window, γ is the discount rate of the reported value, which reflects the influence of future reported value on the current action, and when γ is 1, the influence of unreported value is fully considered, but the excessive oscillation of the reported value is caused, which affects the convergence effect. When γ is equal to 0, only the influence of the reported value at the next time is considered, and the algorithm has a fast decision speed, but loses the grasp of the future trend.

After obtaining the multi-step reward value, the multi-step experience information s_i,a_i:i+N,s_i+N,r_i:i+N]Stored in an experience pool for further updating of the neural network. At the same time, new s is continuously acquired_i,a_i,r_i,s_i+1]When the information is stored in the sliding memory window, the old information is discarded, and new multi-step experience information s is obtained continuously through the sliding memory window_i,a_i:i+N,s_i+N,r_i:i+N]And continuously storing the data into an experience pool.

And step 16, when the multi-step experience information in the experience pool reaches a certain amount, updating the neural network, otherwise, repeating the steps 11-15.

And step 17, updating the neural network parameters. And the value of epsilon is continuously increased during the iteration process. So that the intelligent reflective surface-unmanned aerial vehicle makes action selection with a large return value in a trending manner.

The passive beam forming design which can be executed by the intelligent reflecting surface after the channel state is sensed can be divided into two conditions: the useful signal is enhanced and the interference signal is reduced.

The useful signal is enhanced, namely the phase between the signal which reaches the receiving end of the user after the transmitting end is reflected by the intelligent reflection surface and the direct signal of the transmitting end and the receiving end is the same, the strength of the useful signal after the receiving end superposes the signal is far greater than that of the interference signal, and therefore the anti-interference purpose is achieved.

Reduce interfering signal, no matter how big power interfering signal interferes with, intelligence reflecting surface carries out refocusing at the receiving terminal after the signal that sends with the interference reflects, but the signal is through reverse phase processing, consequently the interfering signal of direct incidence and the interfering signal of reflection at the receiving terminal can offset, for useful signal, interfering signal obtains reducing, anti-interference communication effect obtains promoting.

Example 1

The first embodiment of the invention is specifically described as follows, the system simulation adopts python language, and the parameter setting does not affect the generality. Assume that the relative positions of the nodes are as shown in fig. 3. Base station position W_S＝[10,10]Mobile user initial W_D[0]＝[1000,1000]The simulation results are shown in fig. 4.

Fig. 4 shows a schematic diagram of the anti-interference communication performance of each optimization scheme under different interference powers, and it can be seen that the anti-interference performance of the system can be significantly improved in a dynamic environment compared to a single optimization method through the proposed joint trajectory and beam forming optimization. And when the interference power is higher, the obvious anti-interference communication effect can be kept, and the steady anti-interference communication under the strong-power full-frequency-band interference condition is realized.

Example 2

In the second embodiment of the invention, the system simulation adopts python software, and the parameter setting does not affect the generality. Aiming at different interference positions, the provided combined optimization method can keep higher anti-interference communication rate under the condition of unknown interference machine positions, and the anti-interference efficiency is obviously improved compared with a single optimization method. The suitability of the algorithm in unknown environments is demonstrated.

In addition, the performance of the algorithm is simulated and analyzed, and the algorithm is compared with different algorithms, and a comparison graph of the achievable communication rates is shown in fig. 5. The joint optimization algorithm proposed herein is significantly superior to the general random algorithm for different interference locations.

In summary, the patent provides a joint path and beam forming optimization scheme, which can optimize the anti-interference communication performance under the condition of a full-band high-power jammer.