適用於無線微型感測網路之具容錯性與攻擊抵抗性之多重路徑資訊傳送機制

Abstract

對於大多數無線微型感測網路的應用而言，資訊傳輸的可信賴程度與其安全性為兩項主要的考量因素。由於感測節點通常佈建於無人照料或敵人充斥的環境中，並且感測到的資訊必須經由節點間本質即不穩定的無線連結傳輸至後端基地台，如何正確、安全且即時地傳送感測到的資訊實為一大挑戰。為了解決此項議題，我們提出了一個基於封包切割概念的多重路徑資訊傳送機制，FAMIDS。在此機制之下，資料來源節點將原始封包利用增加冗位元的方式切割成數個小封包，並且利用多重路徑安全地傳送該封包。即使部分的小封包接收錯誤，後端基地台仍能重建原始封包。 除此之外，我們建立了分析FAMIDS所能達到的傳輸可信賴程度之分析模型。針對不同的網路條件，我們研究該分析模型之表現情形，並且提出能找出最佳封包切割方式的有效率演算法，以達到最佳的傳輸可信賴程度。相較於傳統之暴力法，我們提出之演算法能有效降低每次封包傳輸之計算複雜度。

Transmission reliability and information confidentiality are two of the major concerns for most applications of wireless sensor networks. Since sensor nodes are usually deployed in unattended or adversarial environments and sensed information is transmitted through inherent instable wireless links among sensor nodes to the base station, how to deliver sensed information correctly, securely and instantaneously becomes a challenging problem. To deal with this issue, we propose an enhanced multipath information delivery scheme FAMIDS based on the concept of packet splitting. With the proposed scheme, the source node splits the original data packet into several sub-packets by adding redundancy, and securely transmits the sub-packets through multiple paths. The base station can reconstruct the original data packet even if part of sub-packets is incorrectly received.

In addition, we develop an analytical model for evaluating communication reliability achieved by our scheme and study the behavior of the analytical form under different network conditions. For each network model, we propose an algorithm for finding the optimal way of packet splitting with highest communication reliability. Compared with the traditional brute-force approach, the proposed algorithms efficiently reduce the computational complexity for each packet transmission.