採用權重式的感測網路定位演算法

Abstract

本篇論文針對無線感測網路提出一個兩階段的定位演算法，在第一階段每個感測節點以DV-hop的方式得到初始的位置;在第二階段感測節點利用鄰居的位置與距離，反覆地更新與交換資訊，並根據資訊的可靠程度給予不同權重來進行多角定位。根據模擬結果顯示，當感測節點鄰居少的時候，位置平均誤差小於無線電波範圍半徑的20%，而可被定位的節點數目大於總節點數目的70%，最後也驗證了所提定位演算法的效能優於其他定位系統。

In this thesis, we present two-phase location algorithm for wireless sensor networks (WSNs). In the first phase, each sensor node obtains its initial position by DV-hop method; in the second phase, each sensor node utilizes neighbors’ positions and distances to neighbors to update and exchange location information periodically, and then operates multilateration with different weight values. According to the simulation results, when the average number of neighbors is low, the average position error is less than 20% of the radio range, and the number of sensor nodes which can be located is larger than 70% of total nodes. Finally we prove the performance of our proposed algorithm is better than other location systems.