The Research Of Node Scheduling Strategies In Service-oriented Wireless Sensor Networks

Wireless sensor networks are new sensing and processing technologies, which have a great prospect and hold the promise of revolutionizing the way we interacting with the physical world. With the growing requirements for network services, supporting various services by means of simple wireless sensor networks has received great attention by scholars, and service computing will become an important trend in wireless sensor networks research.Large amounts of tiny sensor nodes are the resource carriers of wireless sensor networks as well as the services providers, which are generally cheapness and simplicity that have limited energy and distorted sensing data. Moreover, these sensor nodes are densely deployed, which has led to very large redundancy and strong correlation among the sensed data. Node scheduling is an efficient way to optimize network performance, which can improve the data sensing by selecting a group of nodes to provide services and reduce unnecessary energy consumption by switching off redundant nodes. In this context, it is important to study efficient node scheduling strategies for supporting various services in the service-oriented wireless sensor networks. We carry out the research of node scheduling strategies by following two important performance indexes, i.e. sensing quality and energy efficiency. The main works of this paper are as follows:(1) This paper has researched the node selection problem with data accuracy guarantee, which aims to deal with the challenges caused by the unreliable sensing component on sensor nodes. We analyze the spatial correlation of sensed data by joint Gaussian model, and propose an integer nonlinear programming to find the optimal solution in case that the solution space is small. We also design two heuristic algorithms, namely, separate selection algorithm that selects nodes for each service in a separate way, and combined selection algorithm that selects nodes accordingly to their contribution increment. Simulation results have shown that the proposed algorithms have good performance in terms of the number of selected nodes.(2) This paper has researched the node scheduling problem with energy efficiency. We propose an energy-aware centralized heuristic scheme, which helps to improve the energy efficiency of nodes. The proposed scheduling scheme adopts two-phase node scheduling scheme constructing strategies, and through periodic reconstruction to rotate the active nodes. Meanwhile, considering the self-organization feature of wireless sensor networks, we design a distributed scheduling scheme. Extensive simulation is performed to evaluate the proposed schemes, and the results show that the two schemes can organize the sensor nodes in efficient way to provide the required services.