An Improved Routing Algorithm For Internet Of Things Based On LEACH Protocol

In recent years, IoT (Internet of Things) technology develops more and morequickly. WSN (Wireless Sensor Network) is able to obtain information whenever andwherever, which is considered being the foundation of IoT technology development.WSN, as an important underlying network technology of IoT, are drawing moreresearchers’ attention. More and more people are involved in research on routingtechnology, as one of the key techniques for WSN, and the work has got greatdevelopment.With deepening research, many routing protocols were proposed, but each kindof routing protocol is proposed according to different requirements, so each routingprotocol is just for improving one or several performance, whereas no protocol canimprove all performance of the networks. This thesis summarized the flat routingprotocols, hierarchical routing protocols, location routing protocols, and compared theperformance for different parameters, as well as selected the most classic routingprotocol, LEACH, as the research object at last. This thesis introduced the workingprinciples of LEACH protocol, analysed its advantages and disadvantages, andimproved the algorithm.The LEACH protocol uses the mechanism of random selection of cluster headnodes in the clustering stage, each node averages the energy consumption, goodadaptive，good fault-tolerance，applicable to a wide range of wireless sensor network.But the mechanism of random selection of cluster head nodes will lead to unevendistribution of cluster nodes, cause the question of large energy consumption andshort life cycle of wireless sensor network， and how many cluster head nodes tochoose to make the minimal energy consumption is a problem that must beconsidered.In this paper, on the basis of intensive study of LEACH algorithm, to prolong thenetwork life cycle for the principle, improved the LEACH algorithm in three aspects:1. By analyzing the energy consumption model of LEACH algorithm, derivedthe optimal number of cluster head nodes that can minimize the energy consumption; 2. Aim at the uneven distribution of cluster head nodes for clustering stage, usingthe K-means clustering algorithm to choose cluster head nodes3. Considering both of distance and energy consumption, this paper presents theEK-LEACH algorithm to cluster.Finally, this paper gave simulation on the original LEACH algorithm, K-LEACHalgorithm which chose optimal cluster head nodes and used K-means algorithm tocluster, and EK-LEACH algorithm which chose optimal cluster head nodes and usedK-means algorithm based on energy consumption. After the analysis and comparisonof simulation results conclude that the choice of the optimal number of cluster headnodes is better than other number of cluster head nodes in death nodes and networklife cycle. The EK-LEACH algorithm with the addition of K-means algorithm basedon energy consumption is better than K-LEACH algorithm and original LEACHalgorithm in the death nodes, network life cycle.