| Summary: | 碩士 === 國立中正大學 === 通訊工程研究所 === 96 === In this paper, we consider the problem of lifetime-distortion tradeoff in wireless sensor network. From the resource limitation of sensor node, we categorize the problem into two subproblems: (OBJ 1) As transmission rate R is fixed, finding a minimal distortion D* for a given transmission rate; (OBJ 2) As maximal allowed distortion is fixed, finding a minimal transmission rate R* for a given distortion. We use vector quantization technique to solve these two objective functions. The goal is to minimize the average transmission rate for spatial correlated sensor data. The static and dynamic traffic of sensor data are considered in this paper. When the sensor data is static traffic we formulate mathematic models (distortion rate model and rate distortion model) to determine the optimal solutions (D* and R*), while the sensor data is dynamic traffic the algorithms (distortion rate algorithm and rate distortion algorithm) are used instead, which reduces the computational complexity. Although the algorithm is guaranteed to converge to a local minimum, it is possible to arrive at the global minimum by repeated applications of the algorithm with different initializations. The performance analysis is conducted under two difference data patterns: uniform distribution U(a, b) and normal distribution N(mu ,sigma^2).
|
|---|
