Wireless Mesh Based Wireless Sensor Network Simulation And Research

Wireless Sensor Network technology which integrates distributed information processing technology, embedded computing, wireless networking and communications technology, sensor technology, is widely used in various fields, and is one of the most influential technology in the twenty-first century. Different from traditional wireless network technology, Wireless Mesh Network technology is a multi-hop network, mixed Mesh networks with a flexible structure can make full use of their network advantage. Wireless Sensor Networks and Wireless Mesh Networks are the focus of current research, the introduction of Wireless Mesh Network to Wireless Sensor Network, combining the advantages of both, makes the sensor network to have the better network performance, and it will become one of the new research trends.This article is around Wireless Mesh based Wirleless Sensor Network for research. Based on the analysis of the current sensor network technology and simulation model algorithms,we combine Wireless Sensor Network technology and Wireless Mesh technology, and discusse the structure of Wireless Mesh based Wirleless Sensor Network, which can effectively reduce the energy consumption and delay the network lifetime. Then focusing on the structure, we design the corresponding node deployment strategy and the Q-leanring based GEAR improved routing algorithm. In this paper we mainly do the work in three areas:1. We introduce the Wireless Mesh based Sensor Network's general framework, function module and the node model, simulate them by OMNet + + simulation tool, and compare communication energy consumption with Multi-Sink Nodes Wireless Sensor Network, the simulation results show that the framework is better than Multi-Sink Nodes Wireless Sensor Network in communication energy consumption.2. With the node re-deployment strategy, we deployed sensor nodes by then and make use of Voronoi partition to filter redundant nodes, soon after calculate the minimum required set of nodes for complete coverage of the target area; Focusing on the Mesh node fixed deployment strategies, create new mathematical model, and apply genetic algorithms to solve the deployment issues, and come out the optimum combination of Mesh node location. The fitness function and encoding is the core of the genetic algorithm, we design fitness function that according to both connectivity and best locations, and utilize the sequence of node locations as the encoding in order to achieve the genetic algorithm. Through node deployment simulation we come to active node set and the best combination of Mesh nodes.3. Integrating GEAR algorithm with the Q-learning, we propose a GEAR improved algorithm, based on Q-value of the node select the next hop node to achieve network load balancing. For the case of network congestion, we simulate GEAR improved algorithm and compare with the GEAR algorithm, the trend of energy consumption curve GEAR algorithm's communication energy consumption curve is higher than GEAR improved algorithm, which is able to adjust Q-value dynamically to avoid network congestion zone.Finally, We make a conclusion on the full-text, that the framework can decrease node communication cost to prolong the network lifetime, node re-deployment strategy particularly Mesh node deployment strategy is practical feasibility, and Q-leanring based GEAR improved routing algorithm achieve Network Load Balancing to reduce the network communication cost.