Modeling And Analysis Of Some Fundamental Issues In Wireless Sensor Networks

Wireless sensor network (WSN) is a new information acquisition and process-ing technology, which has integrated the technologies of computer, communicationand sensor. With great application prospects, WSN has increasingly concernedby academia and industry. The connectivity,coverage and energy consumptionare fundamental issue in WSN, which a?ects the performance and lifetime of thenetwork. This dissertation further studies on some problem related connectivity,coverage and energy consumption. The dissertation includes the work as follows:Firstly, according to the characteristics of densely WSNs with uniform ran-dom deployment, a random graph model is established. On basis of randomgraph, the asymptotic properties of network connectivity is obtained. Analysisand simulation results further confirm the use of random graph model describesa topological structure of large-scale randomly deployed homogeneous WSNs ise?ective; A ?ooding algorithm of network nodes space query is given, and severalkey indicators for algorithm are given, including the average query distance, theprobability of a successful query, the cost of a successful network query.The thesis next considers a partial coverage network with node-based ran-dom scheduling mechanism. The quality of network coverage quality is concernedin any time slot, and the approximately closed expression of probability whichrandom target is sensed successful by network is given; We tackle with optimaldeployment pattern to construct a WSN with connectivity and complete coverageby the deployment of regular polygon , and obtains a good lower bound of theprobability of connectivity and coverage in random deployment by regular poly-gon subdivision; A distributed deciding algorithm with free location ofβ?partialcoverage and redundant connectivity is proposed.Based on Interfere-Protocol model, a wireless transmission model with free-interference access mechanism is given, and give an approximate estimate functionof e?ectively synchronize tra?c and a quantitative relation between spatial reuse and network parameters; Energy-based relay sub-optimal practical routing algo-rithm is proposed under Energy-based relay optimal routing algorithm, and givenits theoretical results of the analysis.Finally, according to redundant coverage of sensor nodes, a fastestredundant-remove single hop clustering algorithm is employed to construct a2-step fusion estimator for unknown parameters of the network. By analysisand comparison, it is found that 2-step fusion estimator keeps similarly accu-racy level of MSE with classical distributed estimator, but significantly reducestotal data transmission;Expanding the classical distributed detection , we pro-pose an triggered single-hop clustering detection algorithm. Simulation resultsand theoretical analysis show the proposed algorithm is more e?ective in denselyWSNs.This article studies some foundational issues of wireless sensor networks. Toemploy methods of mathematical modeling and quantitative analysis obtain somequantitative relationships between network performance and some key parame-ters, which are forward to be a frame of reference to design or to optimize aWSNs.