Efficient Data Collecting Algorithm In Was With Mobile Sink

Wireless sensor networks (WSN) is a hot spot in the field of frontier, which involves sensors, embedded and wireless communication technology. WSN has a broad prospect of application. The sensor node of WSN has more limited energy and communication capabilities relative to traditional network node. In wireless sensor networks, a key problem is how to effectively balance each sensor nodes' energy consumption and prolong the life cycle of the whole network. While in traditional wireless sensor networks the sink node's location is fixed, sensor node that is closer to the sink node, the sensor node consumes energy faster. The sensor node may has empty energy, which causes the whole network premature death. In recent years, researchers put forward in wireless sensor networks using mobility methods and verify it can bring the network performance improvement. Researchers from many aspects of mobility are studied:some actual deployment system, energy and routing analysis. Aiming at application problems of the wireless sensor network based mobile sink, we research efficient data collection and mobile sink network clustering.First, the paper presents an efficient data collection scheduling algorithm based on mobile sink to overcome shortcomings of MWSF algorithm. A star algorithm is used to solve the shortest path problem in the network which has obstacles, while in the MWSF algorithm the mobile sink only move in a straight line. By using a single-hop range, mobile sink not only collects data of the sensor node which it visited, but also collects data of other sensor node in single-hop range. This can improve the way of data collection. The simulation results show that the algorithm can ensure more efficient data collection.Then, because wireless sensor network sensor node has limited computing power and communication ability, there is the request to design network topology which can improve the performance of wireless sensor network. As a kind of clustering algorithm, K-Means algorithm can cluster the sensor nodes according to their geographical positions. K-Means algorithm can be applied to wireless sensor networks, but needs to be improved. There are two reasons. On the one hand, the wireless sensor network's cluster head should be sensor node, while using traditional K-means algorithm the cluster head is the center of each cluster. We can choose the sensor node which is closest to other nodes as cluster head node. On the other hand, sensor node'energy is limited. Because cluster head node has to forwarding the data of other sensor nodes, cluster head node should have more energy, and sensor node which has low energy is unfit to be cluster head, so we should consider in the surplus energy of the sensor node. In view of the above two problems, we weights the distance and the surplus energy. The improved algorithm was simulated in OMNET++. The simulation results prove that the improved algorithm can improve the network's life cycle effectively.