Research On Coverage And Data Fusion Algorithm In Wireless Sensor Network

As an important part of the rising Internet of Things, wireless sensor network has profoundly changed the way people live. Wireless sensor network have been widely used in many fields of medical electronics, smart grid, smart home and so on as a new information acquisition and processing technology.Currently, research on wireless sensor network are no longer concerned about the initial stage of the hardware and software design, communication protocol design and the layered architecture, etc. gradually. Instead, research increasingly concerned about the global application-oriented solutions. The deployment and coverage, information processing and information fusion are becoming focuses and hotpots of research. However, the existing research often seek solutions from the perspective of wireless ad hoc networks, ignoring the energy-constrained limited computing and storage capacity of wireless sensor network. The problem of area coverage and multi-sensor data fusion in wireless sensor network were studied in this thesis, and corresponding algorithms were put forward. The main contributions of this thesis include the following aspects:Firstly, the concept of Node-redundancy and Cover Set-Redundancy based on Boolean coverage model were proposed. These two concepts are helpful to improve the coverage model of wireless sensor network, characterize the network with low computational complexity adjustable parameters. This model can calculate cover set performance quickly and provide reference for the design of communication protocols, routing protocol and other research for wireless sensor network.Secondly, optimization algorithm RGA-D was put forward this thesis, which pay attention to the entire lifetime of network, consider both nodes and cover set redundancy, using Genetic Algorithm to get cover set. Simulation results show that RGA-D can balance the coverage rate and active node number, solve the problem of blind area in later life of network.Thirdly, a new support function OLSF based on the least squares fit was proposed to solve the problem of complex computation of support function. It can get results which are very close to exponential function in polynomial computational complexity, reduces the amount of calculation greatly.Finally, a multi-sensor data fusion algorithm TMMDF for the numerical data:was proposed in this thesis, TMMDF eliminate the gross errors in the original data with Dixon method, and then use OLSF support function to calculate the membership degree of each data on the median and average value. Finally, the algorithm selects the appropriate Triangle Module Operator to calculate and get fusion results. TMMDF algorithm has low complexity and high fusion precision; simulation results show TMMDF algorithm is more accurate and robust.