無線感測網路中認知式擇路協定的跨層設計

Abstract

感知無線（cognitive radio，CR）的概念在於強調頻段的有效利用。在無線感測網路的範疇下，本論文提出了一個“以網路配置向量來協助的認知式擇路協定”（CNAR），根據跨層的概念達到更有效率地使用有限的資源。其中，網路配置向量（NAV）被用來估測頻道的狀態，結果再應用於封包路徑的選擇。此外，“認知式路徑轉換”及“降低連結點的區域性路徑修補”亦被提出來，進一步地提升CNAR的效能。最後，一個理論上的研討及數個模擬結果將用做性能的評估與比較。其結果顯示出，在不大量耗費控制封包的情況下，所提出的CNAR比現今許多類似的擇路協定表現更佳，特別是在頻道擁塞的情況下，更能突顯出CNAR的優勢。

Cognitive Radio (CR) is a paradigm which aims at improving the spectrum efficiency by means of the prior information and the ability to adapt. In this thesis, a Cognitive NAV-Assisted Routing (CNAR) algorithm is proposed to effectively make use of the limited channel resources with a cross-layer approach. The existing NAV vectors obtained from the RTS/CTS packets are utilized for the estimation of the channel status around a sensor node. The routing path is selected based on the channel-free probability along the route for the transmission of data packets. Moreover, the cognitive path-switching scheme and the Reduced Hops Local Repair (RHLR) mechanism further enhance the routing performance of the proposed CNAR algorithm. The effectiveness of the proposed CNAR protocol is evaluated via both the analytical study and the simulation results. Without consuming excessive power for each sensor node, the CNAR algorithm can achieve better performance comparing with other existing schemes, especially under the scenarios with the channel congestion.