Research On Energy-saving Algorithm Based On Edge Intelligence In WSN

According to the characteristics of large scale,high density and limited energy of sensors,a hierarchical and distributed large-scale Io T architecture has been proposed.In this network architecture,a large number of passive terminal nodes are divided into homogeneous terminals and heterogeneous terminals according to different processing capabilities.terminal.Based on this network architecture,this paper uses active sink nodes with edge intelligent computing capabilities to perform sleep scheduling on passive terminal nodes,so as to meet the needs of the task itself,reduce energy consumption,and extend the life cycle of the network.Aiming at the optimal coverage problem of homogeneous terminals in a subnet,this paper introduces the concept of regional coverage contribution,transforms the problem of selecting optimal coverage subsets in a large range into the selection problem of base particles,and organically combines the The particle swarm algorithm,considering factors such as network life cycle,area-aware coverage and iteration cost,proposes a minimum coverage subset sleep scheduling algorithm based on homogeneous terminals for the hierarchical and distributed large-scale Io T architecture.In order to realize the load balancing of the whole network,the particle swarm algorithm is optimized,and two constraints are set to verify whether the base particles are legal.In the algorithm iteration process,the illegal particle matrix is corrected based on the legitimacy of the base particles.Finally,the performance of the simulation results is compared with the traditional particle swarm algorithm and greedy algorithm,which proves that this algorithm can not only meet the task coverage rate but also effectively prolong the network life cycle in large-scale scenarios,and has better algorithm efficiency.Performing cooperative tasks in a multi-hop network requires considerable processing power,which is often beyond the capabilities of a single terminal.In this paper,cooperative tasks are divided into subtasks with and without space constraints,and the dependencies between subtasks ensure the cooperativity of the entire task.Distributing different subtasks to multiple heterogeneous terminals for parallel execution is obviously a good solution.While each terminal has different processing capabilities,multi-hop wireless communication will also restrict the execution of mutually dependent subtasks.Therefore,for the hierarchical distributed large-scale Internet of Things architecture,this paper proposes a task mapping scheme based on genetic algorithm.According to the space constraints of subtasks and the shortest route between terminal nodes,this scheme only considers the most cost-effective part of the terminal node parallel processing subroutines.Task.Additionally,for system stability,network lifetime,and time constraints on task completion,a hybrid fitness function is derived and embedded into the algorithm.Finally,the simulation results are compared with the greedy algorithm and the multi-objective task allocation algorithm,and it is proved that the algorithm can comprehensively consider energy consumption balance,scheduling period,timeout rate,network life cycle,system reliability,etc.in the case of large-scale deployment.optimization goal and perform well.