無線感應器網路上可調適及節能之資料儲存與查詢機制

Abstract

在無線感應器網路上，節能一定是各種機制設計的主要考量議題。因此，直覺性的 資料儲存與查詢機制如外部儲存或是本地儲存，都不適用在這樣的環境下。資料中 心儲存可以減少很多網路上的傳輸負載，並且將負載平均分配到整個網路的所有節 點上。然而，在無線感應器網路上，事件發生頻率和查詢的發送頻率是時時變動且 經常彼此不相稱的。若採用上述的靜態資料儲存與查詢機制，將會因為無法隨即時 狀況調整而產生一些不必要的傳輸成本。 在這篇論文中，我們提出了Seesaw。它是一個在無線感應器網路上可動態調適之機 制，可以達到節能的資料儲存與查詢。Seesaw的基本概念來自於GHT (Geographic Hash Table)裡面的結構複製 (Structured Replication)。藉由選擇最佳的複製參數，Seesaw可以動態調整儲存和查詢的策略 以達到節能的目標。此外，可動態調適之複製的概念也可以增進系統的負載平衡以 及可靠性。模擬實驗的結果證實了Seesaw不但能在低能源消耗的狀況下，完成資料 儲存與查詢的任務，同時也能夠達到負載平衡，高擴充性，以及高可靠度的目標。

Energy efficiency is a major concern of mechanism design in wireless sensor networks. Therefore, heuristic data storage and query approaches like external storage or local storage are not suitable in this kind of environment. Data-centric storage is a useful method which not only reduces total message load, but distributes load throughout the whole network. However, event happening and query dissemination frequencies are time-varying and usually mismatched. The above static data storage and query mechanisms can not adapt to the realtime situation and hence incur unnecessary transmission costs. In this thesis we propose Seesaw, an adaptive scheme for supporting efficient data storage and query in sensor networks. Seesaw is based on the concept of structured replication in GHT (Geographic Hash Table). By choosing the optimal replication parameters, Seesaw can dynamically adjust its storage and query strategies to achieve energy efficiency. Besides, the concept of adaptive replication also improves load balance and reliability. The simulation results demonstrate that Seesaw can not only successfully complete storage and query tasks in lower energy consumption, but achieve load balance, scalability, and robustness simultaneously.