Research And Implementation On Clustering Routing Algorithm In Wireless Sensor Networks

With the development of wireless communication technology and microelectronics in ahigh speed, the wireless sensor networks, WSNs for short, have provided a new research field forus. It is composed by lots of sensor nodes which collect information in the monitoring region bycooperating with each other. The routing algorithm is one of the pivotal technologies in WSNs. Itprovides a route scheme for the nodes which deliver data packages from the source node to thetarget node. Because of this, it is very valuable to study it for the networks. Because every nodeconsumes the limited energy unequally, the routing algorithm takes an important role to extendthe life of the network.The main contributions of this thesis are illustrated as follows.According to the characteristics and design principle of WSNs, we analyze the relationshipbetween the uniformed distribution of cluster head node and the networks load. Combined withGauss distribution, we also analyze the principle of choosing cluster head and develop the Gaussclustering routing algorithm, GCRA for short, in solution to the problem of short lifetime andvarious distributions of cluster head. In the proposed algorithm, the mean value isdetermined by the optimum distance between node heads while the variance value by therate of cluster heads. In addition to these two factors, the probability function is also determinedby the residual energy of the node. The node gets its probability by the distance to the currentcluster head and chooses the nearest cluster head as its final head.For the deficiencies of GCRA algorithm, the paper presents the modified Gauss clusteringrouting algorithm, M-GCRA, for short, by introducing the minimum probability value. In fact, itreduces the target nodes number and time complexity by dividing the residual nodes. The nodewhose probability is much lower than the minimum probability value will never turn into thecluster head. Besides, according to the Gauss distribution probability value table, it deducts therelationship of the mean value and the variance value. The simulation result shows thecorrectness of the theoretical analysis. Finally, In the MATLAB simulation environment, the simulation results show that the Gaussclustering routing algorithm can choose the cluster head which has the higher energy comparedwith traditional LEACH and CMCRP algorithms. The modified algorithm can also guarantee theuniform deployment of cluster head nodes. Besides, the nodes consume energy consistently. Thealgorithm shows reasonable results in extending network life and increasing the quantity of datapackages.