Improvedant Colonyalgorithm To Solve Mobile Agent Routing Problem In Wireless Sensor Networks

Wireless sensor network is one kind of resources-restricted network withcharacteristics such as energy constraints, limited communication ability, low capabilityof calculation and memory. In order to maintain network’s longer working time,designing energy-efficient routing protocols in wireless sensor networks become one ofthe primary challenges. Adopting mobile agent technique in wireless sensor networkscan effectively reduce the redundant data transmission, save energy and prolong thenetwork existent cycle. According to the characteristics of wireless sensor network, thispaper discusses improved ant colony algorithms to solve the routing problem of mobileagent, and the main research contents in this thesis are as follows:(1) Designing wireless sensor network model based on ant system and analyzing therelationship between the performance of ant colony algorithm and relevant parameters(ants number, pheromone volatile coefficient, pheromone weighting factor, heuristicinformation weighting factor and pheromone intensity), as basis of the future research.For wireless sensor networks environment, how to choose reasonable parameters toaccelerate its convergence speed and prevent invalid paths are the key focus of theresearch.(2) Solving mobile agent routing based on an improved Ant Colony System. Usingtheory of clustering, the large scale monitoring network is divided into several clusters.Mobile agent exchanges data with cluster-head nodes and carries data to base station.This rule can reduce data transmission and save network energy to a certain extent. Inorder to improve the algorithm convergence speed, choosing partial best paths andreleasing pheromone to lead ants choose paths. Meanwhile, a mutation operation isintroduced to avoid invalid paths. The method relies on finding the common neighbornodes of invalid path end-nodes as relay nodes to reconstruct a new path. Simulationexperiments show that the improved ant colony algorithm can accelerate convergencespeed and prevent algorithm into the local optimal solution.(3) Best-Worst Ant System Based on Mutation is introduced to solve mobile agentrouting problem. In small-scale network model, mobile agent directly accessesmonitoring network data-source nodes and takes message to client after completing data collection. In this model, mobile agent doesn’t need to access dead nodes. UsingBest-Worst Ant System to solve mobile agent routing expands pheromone differencesbetween optimal path and worst path and leads ants to choose paths quickly. Also,considering the communication ability of nodes and adopting mutation operation toprevent invalid paths. The analysis of experimental data shows that the improvealgorithm suits to small-scale network, solve mobile agent routing problem effectivelyand avoid invalid paths in wireless sensor networks.