Research And Improvement On Clustering Routing Algorithm For Wireless Sensor Networks

Wireless sensor network, as the core part of "the internet of things" at present, highly integrates sensor technology, communications technology, and computer technology. In recent years, it shows a good application prospects in military, environmental monitoring, healthcare, intelligent transportation and other areas. In wireless sensor network, the node energy is limited. To organize the cluster structure and use the node energy effectively, which is the important Part for routing algorithm.In this thesis, the author analyzes cluster-based routing mechanisms for wireless sensor networks at first. Then the author emphasized cluster head selection, cluster formation and data transmission in cluster-based routing protocols. From these techniques perspective systemically recent representative cluster-based routing protocols are presented, and compared with their characteristics and application areas.Second, through the analysis of the advantages and disadvantages of traditional clustering algorithm LEACH, the algorithm WBCP and EECS are also raised. In WBCP, the factor of nodes’ cluster selection is not comprehensivly considered and the weight coefficient is determined by human, lack of certain theoretical basis; In EECS, simply considering the node’s distance from cluster-head in the process of cluster formation, which can’t make nodes join into the appropriate cluster. Therefore, from the angle of two algorithms’shortage, this paper improves their cluster-head selection and cluster formation and proposes an Energy-Efficient Clustering routing Algorithm based Weight and Fitness factor for WSN (WF-EECA).In the cluster-heads selection mechanism, weight factor W is introduced as adjusting parameter to choose cluster-heads, and AHP (Analytic Hierarchy Process) is also citied to work out weight coefficients; In the attribution mechanism of non cluster-head nodes, clustering factor F is introduced, considering cluster-head’s factors of remaining energy, distance between cluster-heads and distance from nodes to base station, a new clusters mechanism function is put forward, which makes nodes choose the best clusters to join into them; And the cluster-head selection isn’t done out of periodic, it uses the dynamic cluster-head rotation mechanism based on energy threshold and does the the exchange to the local topology information, and then makes the clustering rapid and effective, reduces the number of cluster-head selection and saves the energy.Finally, we use the MATLAB to build a simulation platform and simulate the improved algorithm and others mentioned above. Simulation results proved that compared to the original algorithms, improved algorithm effectively balances network energy consumption, prolongs the network lifetime, relatively the success rate of data transmission to sink also have increased significantly. What’s more, the improved algorithm is more suited to the heterogeneous wireless sensor networks.