以類電磁演化計算為基礎解決無線感測網路拓璞最佳化問題

本研究提出一種演算法，針對無線感測網路的感測器擺放問題，我們在合格解的前提，也就是連結性和探測機率的滿足下，從感測器的總數和整個系統的電力消耗來優化。在電力消耗的角度上參照了相關研究的方向，從感測器間距離、封包流量、封包流量標準差以及整體拓璞深度去分析。另外最大的貢獻點在於，我們修改了傳統類電磁演算法的機制，第一個將其使用在此題目上。值得一提的是在庫倫靜電力求值時，因為類電磁機制對本題目的天生適應性，我們將之相當程度的簡化，並提出電量線性衰退機制來避免求解過程的震盪現象。實驗結果也確實證明了我們演算法的核心精神，也就是類電磁機制的使用上，相較於對照組結果是能求出更佳的結果，讓整個拓璞能以更少的感測器，在一定完整度上延長系統的運作時間。

關鍵字

無線感測網路；拓璞最佳化；類電磁機制演算法；多目標最佳化

並列摘要

In this research, we propose an algorithm to solve Wireless Sensor Network (WSN) deployment problem. In this problem, topology connectivity and requested event detection probability must be ensured as a qualified solution. Then with five directions to optimize the topology according to our related work, the employed sensor count and lifetime are expected to be improved. The main contribution of our algorithm is that we simplify the original Electromagnetism-like Mechanism (EM), and being the first research uses it to solve this WSN deployment problem, because we think Coulomb's electrostatic force is natural fit the problem. In addition, an electric charge linear aging method is applied to avoid oscillation through computation process. In the end, our simulation result shows the positive analysis when comparing our work with related literature. The solution generated by our algorithm can be more stable and balance in energy consumption, furthermore with fewer sensors