Research On Coverage And Connectivity Optimization Of Wireless Sensor Networks For Ship Environment

Recently,with the development of intelligent and unmanned ships,the demand for information perception is increasing.Wireless sensor network has gradually become an important channel for ship information perception with the characteristics of simple deployment,low cost and high redundancy.The coverage and connectivity of wireless sensor networks determine the network topology and routing,which has important research value.Traditional algorithms cannot solve the multi-objective optimization problem of wireless sensor network in ship applications.In order to realize the engineering application on ships and shorten the distance from theory to practice,the problems of minimum coverage and connectivity,redundant coverage and connectivity of important targets,and energy constraints are studied.The main research work is as follows:In order to solve the problem that a fixed number of sensor nodes are used to deploy the wireless sensor networks under the background of ship application,Self-learning method is introduced to adjust the important parameters of the algorithm online,and Gaussian mutation is used to generate new harmony randomly,which increases the diversity of variables.At the same time,reverse learning competition strategy is used to select the best of the fittest vectors in harmony memory.Then ship engine room is selected as the coverage area of sensor nodes,and the influence of equipment is considered.During initialization of harmony memory and execution of the algorithm,the method of variable length harmony vector is used to minimize the number of sensor nodes during network deployment.Simulation results show that the network coverage of the proposed strategy is increased by 8%,and the number of sensor nodes is reduced by 3,which realizes the minimum deployment of the number of sensors.Aiming at the redundant coverage and connectivity of important objectives,an adaptive operator is used to accurately describe the global impact of harmony diversity and prevent the algorithm from falling into local optimal solution.Dynamic iterative optimization coefficient is introduced to update the position and generate new harmony vector,so as to get rid of the problem of parameter setting.Then local optimization strategy of cell is adopted to break through the limitation of harmony memory and interact with potential cells outside the memory,which is conducive to the realization of global optimization.The multi-objective optimization problem of coverage and connectivity is designed,and external archive is used to store the non-dominated harmony vector.At last,fast non dominated sorting and crowding distance strategy are adopted to determine the Pareto optimal front.Simulation results show that the coverage and connectivity rate of the proposed algorithm are above 92% and 95%respectively under different k and m values while and the number of sensor nodes is least.In order to further improve the robustness of wireless sensor networks,the energy constraints of sensor nodes such as energy consumption and energy balance is considered.Firstly,a routing node selection scheme considering energy consumption and energy balance is designed before the deployment of sensor nodes and the potential energy consumption is considered in advance,which will improve the energy efficiency and realizes the efficient deployment.Secondly,the initialization method of deep reinforcement learning algorithm is improved and 27 discrete actions are designed,the random selection action mode is improved to adjust the position and number of sensor nodes.Meanwhile reward and punishment mechanism are designed to improve the efficiency of the algorithm training process.At last,the algorithm experience playback mechanism is improved to shorten the algorithm training time.Simulation results show that the convergence speed of the algorithm is improved by25%,the energy consumption of the proposed strategy is lowest and the total hops required to complete the connection are least after the nodes are deployed and run for 3500 rounds.According to the above research work,the coverage and connectivity optimization of wireless sensor network in ship environment is realized.This research has practical significance to strengthen the application of wireless sensor network and promote the development of intelligent information perception on ships.