Research On Pattern-based Query Strategy For Wireless Sensor Network

Wireless Sensor Network (WSN) is widely used in more and more areas; event detection is one of its applications. Pattern-based query system is an important way to realize event detection. As sensor nodes have many limitations, this paper makes research on pattern data distribution mechanism and matching mechanism on sensor nodes for pattern query system.Flooding is an important way to distribute pattern data. But in basic flooding, sensor nodes broadcast packets automatically, which makes the number of packet redundancies in the network increase quickly. In order to solve this problem, this paper proposes Level Based Flooding (LBF). In LBF, the whole network is divided into several levels according to distance (hop) between sensor nodes and sink node. When users want to distribute pattern data, the sink node sets packet’s Time to Live (TTL) according to the coverage area of the packet to avoid useless flooding. In the diffusion process, each node makes decision on how to process the packet according to neighbor nodes’ processing situation. Through this way, the number of nodes that broadcast packets is reduced, and packet copies are reduced, so the load of network is reduced.If sensor nodes run matching mechanism on the original data collected by nodes, as for large amount of data, algorithm’s running time is long, which consumes much energy. At the same time, storing original data directly wastes a lot of storage. This paper proposes Piecewise Linear Representation based on Adjacent Data (AD-PLR) to compress the original data. Through this way, the amount of data to be processed is reduced a lot, and the storage is saved, so the energy consumption for matching mechanism is reduced.This paper does simulations to evaluate LBF’s performance. Simulation results show that LBF reduces the number of nodes that broadcast packets and reduces the energy consumption of data distribution. LBF solves broadcast storm very well. This paper also does experiments on STM32W108to evaluate AD-PLR’s performance. The results show that AD-PLR’s performance improves a lot. Its running time is reduced and energy consumption is very good.