The Research Of The Routing Protocol Based On Self-Organizing Neural Network

With the rapid development of sensor technology, embedded computationtechniques, distributed information processing technology and communicationtechnology, wireless sensor network has a certain research value, which has been widelyused in many fields. Based on LEACH (Low Energy Adaptive clustering Hierarchy)protocol for wireless sensor networks, we propose RBCSC (Routing based on ClusteringSelf Organizing Neural Network and Chain) protocol in WSNs in this paper.LEACH is a hierarchical routing protocol which was earlier proposed in wirelesssensor networks. It is an adaptive clustering routing algorithm, which divides the entiresensor network into several clusters, and reduces conflicts in the intra-cluster andinter-clusters by TDMA/CDMA in the MAC layer mechanism. LEACH-C protocol isproposed on the basis of the LEACH, which is a typical centralized protocol, using thesimulated annealing algorithm for clustering, and the cluster heads can be more evenlydistributed in the network. LEA2C protocol improves LEACH-C protocol, and it usesself-organizing map for clustering. In the field of self-organizing neural network forclustering, scholars proposed EBC-S protocol.In this paper, the RBCSC protocol uses self-organizing map and learning vectorquantization for clustering, learning from PEGASIS (Power-Efficient GAthering inSensor Information Systems) protocol about a chain of algorithms, inter-clusters ofRBCSC chains by greedy algorithm. What is more, RBCSC selects the low-level clusterheads, and between the low-level cluster heads use the greedy algorithm into a chain.Finally, RBCSC elects only one low-level cluster head as the high-lever cluster head, thehigh-level cluster head communicates directly with the base station. The analysis andsimulation results show that, compared with LEACH, the improved algorithm increasesthe complexity, but has better performance in terms of balancing node energyconsumption and prolonging the network lifetime. The RBCSC is suitable for small-sizedand medium-sized networks. What is more, RBCSC shows a better performance in thesituation that the base station is far away from the network center. However, the initial energy is less influenced to the RBCSC.