Gravity algorithm for Wireless Sensor Networks with multiple sinks

Wireless Sensor Network (WSN) is a new form of ad-hoc network which is characterized by a very large scale of nodes. Each node has a sensing module, small memory and a limited power resource. The WSN is application specific, which means one routing algorithm cannot be adopted for all problems and applications. In finding a suitable algorithm, the main problems are the limited resource and the dynamic of the network. One of the WSN routing algorithms is called Gravity. The Gravity algorithm uses the gravity value at each node to form a virtual contour. This virtual contour will guide the message to the sink. In this paper, the modeling and implementation of the Gravity algorithm on WSN will be discussed. Different from earlier works, the emphasis here will be given to multiple sink scenario and its effects. In this study, the Gravity model is implemented using the simulator OMNeT++. The multiple sink scenarios are represented with cases with two sinks. The simulation results show that the multiple sink cases bring a huge power saving with a ratio up to around 250. In a WSN with multiple sinks, virtual node clusters are formed based on the relative distance from the nodes to the closest sink. This clustering will limit the data transmission within a certain number of nodes. Hence the node power can be saved and network transmission load is reduced. This is a very important advantage because of the power limitation on each node.