Research On Node Schedule For Wireless Sensor Networks

Ever-advancing in processor brings about the fast improvement in the capability of data processing, ever-increasing storage capacity leads to the realization of mass memory and ever-broadening communication bandwidth makes access to data very easy. Meanwhile it is noticeable that the collecting method of data lags far way behind and becomes incompatible with the above technologies. Integrating MEMS, computation and communication, Wireless Sensor Networks emerge as the times require. It consists of a large amount of micro-sensors, which are immerged in the physical environment to collect information of the targets inside surrounding area and radio them to the user. A broad outlook is presented in the fields of military and civil applications with its unique advantages of high sense accuracy, easy deployment way, and low investment budget.For the design of Wireless Sensor Networks, Quality of Sense and Lifetime are two important performance indexes. And Energy Efficiency is one critical problem since battery is the main source of power for micro-sensor. On one hand the battery for micro-sensor has a limited capacity and on the other hand there are a huge amount of nodes in Wireless Sensor Networks, which are often deployed in a remote, unguarded and abominable environment that makes it very difficult and infeasible to recharge or replace the batteries. Therefore, guaranteeing the Quality of Sense as well as prolonging the Lifetime of the entire network by optimizing energy consumption turns to be the essential target in designing Wireless Sensor Networks.Different from those energy-efficint network communication protocols, Node Schedule method is one of the effective Application-oriented way to improve the Energy Efficiency of Wireless Sensor Networks. By scheduling some nodes to work and the others to sleep within a specific time period and space region, Node Schedule optimizes data source and amount within the network and reduces energy consumption in sensing and wireless communicating. In that way Energy Efficiency is enhanced and network Lifetime is prolonged. This method is introduced against the high data redundancy of Wireless Sensor Networks, i.e. few data discrepancy between collected by part of the nodes and by all of the nodes. The other point is that it is easy to cause network jam and reduce data traffic of the network if too many nodes are activated at the same time and the consequences would be a loss of data or a delay in transfering.Application-oriented is the characteristic of Wireless Sensor Networks, which means the design of Node Schedule must aim at the application requirement and base on the application feature. In the dissertation, the design space and objective of Node Schedule are fully analyzed. And directed towards two major kinds of Wireless Sensor Networks Applications, i.e. Event Detection and Environment Monitor, we focus on research of different Node Schedule methods suitable for them based on the demand and feature of those applications. The proposed methods effectively reduce the energy consumption while meeting the demand of quality.The following are the main innovations of the dissertation.1. The design model of Node Schedule is constructed and design assumption as well as design objective are concluded. The state-of-art of Node Schedule methods are compared, which provide guide and ideas to design new Node Schedule methods.2. An effective Grid-based schedule method is proposed for event detection applications based on partial-coverage strategy aiming to reduce energy consumption and shorten event detection delay. As a cluster method based on location, grid-partition implements an even partition and balanced load among nodes, as well as a balanced energy consumption with guaranteed network connection, low overhead and good scalability. The simulation verifies the Grid-based schedule method is superior to random schedule one in the event detection quality.3. An effective TIN-based schedule method is proposed for environment monitor applications based on data spatiality aiming to reduce energy consumption and guarantee data field accuracy. The method considers spatiality as well as dynamics of field data, through which energy consumption of wireless communication is reduced as well as the amount of data need to be transfered to sink node, the dynamic of field data is matched and the accuracy of generated data field is guaranteed. It is showed by simulation that the TIN-based schedule method can achieve remarkable energy consumption than original one and provide much more scalability.4. A multi-resolution schedule model, comprehending a distributive data representation method and Grid-based hierarchy, is brought forward for environment monitor applications. An adaptive multi-resolution sense model is further presented to get variable data resolution within different regions. The result of simulation shows energy efficiency is improved with sense accuracy by supporting multi-resolution sense model to Wireless Sensor Networks.