Research On Multi-objective Optimization Of UAV-assisted Physical Layer Security Communication Of IoT In Forest

In forest environments,UAV-assisted communication is used in a wide range of scenarios,however,there are challenges in its secure communication and energy saving due to the complexity of the terrain,the openness of channel transmission and the limited range of UAVs.In this paper,we investigate the multi-objective optimization of UAV-assisted Internet of Things(IoT)physical layer secure communication and energy consumption reduction in forests.A virtual antenna array is formed using multiple UAVs in forest to achieve assisted IoT secure communication in the presence of eavesdroppers through collaborative beamforming techniques.However,the process of forming virtual antenna arrays consumes the limited energy of UAVs.Based on the conflicting multiple targets and for two scenarios of UAV transmission in forest environment,the specific work in this paper is as follows:(1)At low UAV flight altitude,considering vegetation interference and the signal loss caused by the ground,a forest channel model containing the tree canopy and beam loss caused by the ground is used,and the three-dimensional(3D)position information of the UAV and the array excitation current are simultaneously optimized for this lowaltitude communication scenario to establish a low-altitude multi-objective optimization problem(LAMOP)for UAVs that maximizes the safe communication transmission rate and minimizes propulsion energy consumption.An enhanced multiobjective particle swarm optimization(EMOPSO)algorithm is proposed to solve the LAMOP.The algorithm uses Circle mapping instead of the traditional random number generation method in initializing the particle swarm individual positions,which obtains the improvement of irregularity and distribution compared with the traditional random number generation method and facilitates the exploration of a wider solution space.Furthermore,the mutation method based on the optimal set is used.During each particle swarm location update process,the convergence speed of the algorithm is judged by evaluating the number of times the optimal solution is repeated,and the opportunity of particle mutation are determined to effectively avoid the algorithm falling into local optimization.The method combines the position update method based on the gray wolf algorithm as the update mechanism of EMOPSO,which jointly uses three special particles to update the particle positions,thus effectively avoiding the particles to fall into the local optimal solution to guarantee the global search ability of the algorithm.The simulation results show that the EMOPSO algorithm has better effectiveness,stability,and convergence compared to other optimization algorithms under different array element numbers.(2)The probabilistic line-of-sight(PLoS)model was used to model the wireless communication channel when the UAV was flying at a high altitude.Aiming at the application of high-altitude flight assiste IoT wireless communication,a method was proposed to optimize the 3D position information of UAV and the excitation current of array element in UAV array.A high altitude multi-objective optimization problem(HAMOP)based on the air base station of high altitude UAV with high safe transmission rate and low flight loss was established.To solve this problem,an enhanced multi-objective salp swarm algorithm(EMSSA)was proposed.The algorithm combines Chebyshev mapping and random number methods to select better performing solutions for population initialization,which enhances population diversity and improves population quality.In this algorithm,Cauchy operator is introduced to make it easier to search for a better solution,so as to avoid falling into the local optimum.EMSSA also combined the update method of wormhole in the multi-verse optimizer(MVO)in the traditional multi-objective salp swarm algorithm(MSSA)to update the position of salp chain leader.The parameter set which can control the search step size is extended to enhance the search ability of the algorithm.The experimental results show that the EMSSA has the best effect in solving the HAMOP problem in different sizes of UAVs.