Research On Routing Algorithm Based On Ant Colony Optimization For WSN

As a new way to obtain information and disposal pattern,wireless sensor networkshave been a heated issue nowadays by scholars home and aboard. Wireless sensornetwork consists of hundreds of sensor nodes, one of which has the capabilities ofperceiving the conditions, performing simple calculation and transacting directlycommunication among other near nodes or from base station. Therefore, it has broadapplication prospects in numerous areas like military, industry, civilian use and so on.In wireless sensor networks, due to energy and other resource-constrained,survival is particularly important, therefore, the energy efficient routing is a challengefor large scale wireless sensor networks. Routing algorithm in wireless sensor networks,the use of basic ant colony optimization algorithm for the path will appear prematureconvergence, easy to fall into local optimum,it is likely to plan no path that leads toalgorithms stagnation in the large-scale wireless sensor network environment, not theoptimal path. This paper puts forward improved ant colony algorithm by means of basicant colony algorithm. Compared to the latter, improved ant colony algorithm makes thefollowing progress:(1) Improved the inspired factor of distance, the introduction of the energy factor,The higher energy level, the higher the probability of the node is selected, the wholenetwork can better achieve energy balance.(2) During the pheromone ant colony algorithm update, vulnerable to the impactof early detection of better solutions, resulting in local optimal solution，in order toovercome this phenomenon, we use the update method with a reward and punishmentmechanism. In the search process, Compared with the optimal solutions, if the betterpath is appropriate incentives, or to a certain degree of punishment.(3) Dynamic change parameters α and β in the basic ant colony algorithm, toprevent the algorithm are local optimal solution.(4) Pheromone evaporation coefficient of dynamic update, to avoid falling intolocal optimal solution.(5) In the algorithm, the ant are divided into forward ants and backward ants, ithas a life, and the information is updated using a local update, and a combination ofglobal update to prevent premature convergence of the ants, the algorithms stagnationinto local optimum. Based OMNET++3.3environment into the simulation, the improved algorithm bysimulation experiments to verify the above algorithm to improve the effectiveness andfeasibility.