Mobile Sensors Dispatch Algorithm In Wireless Sensor Network

With the advancement of science and technology, wireless sensor networks (WSNs) have developed rapidly. Due to its low cost, low power, flexibility and convenience, it gets more and more attention in many areas of research and applications. Wireless sensor networks have been used in many fields, such as environment monitoring, smart agriculture, health care and intelligent transport. Since wireless sensor network has a large number of nodes, while the sensing and computing power of a single node is limited, it can’t provide a very comprehensive monitoring. It will increase the cost greatly if improving every node’s sensing and computing power. So a wireless sensor network usually has several mobile sensors which have stronger sensing and computing power to support network. It will move to the area of interest for further investigation and analysis after static nodes detecting something. We call such networks with mobile sensors hybrid wireless sensor networks.In a hybrid wireless sensor network, static sensors are responsible for detecting events while mobile sensors can move to the location for further analysis. The most energy-cost operation of mobile sensors is movement, how to reduce the moving distance of mobile sensors while allow it to complete the task is a challenging research topic. If some mobile sensors are continually dispatched to different locations, they will run out of energy soon and cause the remaining sensors load heavily. This paper presents a mobile sensor dispatch algorithm, it can balance the load of each mobile sensor, can applied for any number of mobile sensors and event locations. When there are more mobile sensors than event locations, this paper translates the problem into a maximum bipartite matching problem. When there are more event locations than mobile sensors, we first clustering the event locations, then dispatch each mobile sensor to a cluster of event location. In order to reduce the amount of message transmission, this paper further proposed a distributed algorithm. Simulation results show that the proposed algorithm can effectively extend the system lifetime.