Research On Computational Intelligence Based Self-organization Methods Of Wireless Sensor Networks

Wireless Sensor Networks (WSNs) are self-organized networks consisting ofresource-constrained sensor nodes in which how to build efficient energy-saving self-organizaionmethods is an important issue. Computational intelligence (CI) belongs to bionics algorithms，itspotential parallelism, self-organization and distribution can meet the requirements of theself-organization for WSNs. Memetic algorithm, self-organization feature map networks andbacterial foraging optimization algorithm of CI are able to build efficient energy-savingself-organizaion methods for WSNs.The thesis studies self-organization methods of WSNs including coverage, topology androuting and the specific work includes three aspects.Firstly, we build a multi-objective coveragemodel from the perspectives of network coverage and network lifecycle, and on the basis of thememetic algorithm, combining with dynamic local search strategy, cell encoding and the elitiststrategy, present a multi-objective optimization coverage algorithm based on dynamic memeticalgorithm, which solves the proposed coverage model and makes the WSNs get a higher coverageratio with a longer lifetime. Secondly, for the problem of the changeable topology in WSNs, atopology control algorithm based on self-organizaion feature map network is proposed with theunsupervised learning characteristics of self-organizaion feature map network, which optimizes thedistribution of cluster heads by unsupervised competitive learning among nodes, dynamicallyadjusts the transmitter power to reduce energy consumption, makes an optimized network topologyand balanced energy consumption. Thirdly, to solve the problem of uneven energy consumption inthe hierarchical routing of WSN, a routing algorithm based on bacterial foraging optimizationalgorithm is presented, which takes node residual energy, transmission energy consumption and thenumber of hops in account, adopts relay node selection mechanism to generate initialization paths,makes use of bacterial foraging optimization algorithm to search for the optimal transmission pathbetween the source node and the sink node; the proposed algorithm improves the routing efficiency,makes an effective balance of the nodes’ energy consumption and extends the life cycle of thenetwork.