三維水下無線感測網路上以深度為基礎的多路徑路由協定

Abstract

陸上無線感測網路 (terrestrial wireless sensor network) 發展已有很長的時間，也已經廣泛地運用在生活上，而最近幾年水下無線感測網路（underwater sensor network）開始受到重視，相對於陸上無線感測網路，水下無線感測網路不管在技術與應用上仍有很大的進步空間，主要是水下無線感測網路使用聲波 (acoustic waves) 作為傳輸的媒介，不同於陸上無線感測網路使用無線電波(radio waves)，也由於聲波與無線電波兩者特性截然不同，因此傳統陸地上的方法無法運用在無線水底感測網路，必須重新發展針對無線水底感測網路的新方法與技術。 本論文我們提出N level binary tree的路由協定，N level binary tree利用感測器彼此的深度差來決定路由的方向，再使用簡單的區塊定位，透過多點跳躍 (muti-hop)的方式，使感測資料能送達海面上的基地站。N level binary tree採用多路徑 (mutipath)來傳送感測資料，N level binary tree會限制傳送次數，因此不會產生過多的傳送次數而且封包到達率不會太低。 聲波在水中通訊容易受到破壞導致發生封包遺失 (packet loss)，陸地上常用的解決方法是採用自動重傳請求 (automatic repeat request, ARQ) 機制，在我們的模擬實驗中，N level binary tree不使用ARQ機制，我們將與單一路徑且使用ARQ的Greedy short sink方法來比較，由實驗結果得知，在不需要極高的封包到達率 ( Packet Delivery Ratio)，N level binary tree能有較低的傳送次數，因此N level binary tree能節省感測器的電力，增加整個網路生命週期 (network lifetime )。In the last few years there has been are intensive study in routing protocols for terrestrial wireless sensor networks. Terrestrial wireless sensor networks have been used extensively in many land-based applications. Underwater Sensor Networks (UWSNs) have attracted a lot of attention recently. Since data packets in Underwater Sensor Networks are transmitted by acoustic signals, the characteristics of a Underwater Sensor Networks are different from those of a terrestrial wireless sensor networks. However, due to the different nature of the underwater environment and applications, there are drawbacks with respect to the suitability of the existing solutions for underwater acoustic networks. In this paper, we propose a N level binary tree protocol. Sinks are usually situated at the water surface. The based on the depth information of each sensor, N level binary tree forwards data packet towards the water surface. We make use of multipath data delivery. N level binary tree will limit the number of transmissions, it does not produce excessive transmission times and packet arrival rate is not too low. Acoustic communications vulnerable to damage in the water result in packet loss, land used solution is to use automatic repeat request (ARQ) techniques. In our simulations, N level binary tree does not use ARQ, we will work with a single path and use the ARQ Greedy short sink method to compare. The results indicate that, without the need for high packet arrival rate, N level binary tree can have a lower number of transmissions, the N level binary tree can save electricity sensor, increasing the network lifetime.