非信標多通道IEEE 802.15.4e無線網路系統效能探討

Abstract

IEEE 802.15.4是一種適合被採用於近距離無線感測器網路(WSN)通訊的規範。它定義了實體層(PHY)和媒介層(MAC)的協定內容。近來隨著物聯網的興起，感測網路功能越來越多元，除了感測功能還能進行主動控制，例如無線感測致動網路(WSAN)。網路的結構也變得複雜，更多的節點、更大的通訊流量需求因應而生。IEEE 802.15.4e就是為了支持日趨繁複的WSAN的發展而制定的一項延伸性協定標準。它包含了許多擴增功能，其中最主要的就是把多通道通訊協定(Multi-channel communication protocol)明確的定義在其公開的標準裏面，該標準依據不同的媒介存取模式(MAC behavior)將多通道通訊分為：信標模式(beacon-enable mode)，代表性協定如Deterministic & Synchronous Multi-Channel Extension(DSME)及非信標模式(non-beacon mode)，代表性協定如Asynchronous Multi-Channel Adaption(AMCA)，其中AMCA的出現讓非信標模式通訊增加實際應用的亮點。非信標模式傳統上有MAC層單位傳輸資料長度可變的彈性，但由於在單通道環境常造成不如信標模式中各結點可獲取較平均接取網路資源的較少傳輸等待時間而喪失其應用機會。如配合多通道方法，則各通道可以進行不同MAC層單位資料長度的通訊，且由於非信標模式對網路接取的立即性更高，彈性及更有效率的網路效能將有更多的應用可能性。我們在既有802.15.4e AMCA的基礎上提出了增進版AMCA模式，把AMCA運作過程區分為四個步驟並提出其功耗分析模型，事實上也等同傳輸時間模型，分析了我們所提出的方法比較既有AMCA 模式效能的增進程度。而這些跟微控制器應用有關的方法也可提供真實部署無線感測網路時的實作方法。本篇論文最後採用OPNET網路效能模擬軟體把既有Native AMCA和我們所提的Enhanced AMCA在較少結點數目與較多結點數目搭配Cluster Tree和Mesh混合網路拓撲做出許多效能模擬實驗結果，經整理分析其結果證明了我們提出的E-AMCA確實比既有N-AMCA效能增強，尤其在結點數多時其表現出增進的作用更為明顯。

IEEE 802.15.4 is a suitable standard to be adopted for wireless sensor networks. It defines the communication physical layer (PHY) and media access control layer (MAC) for wireless personal area network (WPAN). Recently, with the multidisciplinary development of Internet of Things (IoT) the WSN becomes more powerful for its functions, e.g. the wireless sensor and actuator network (WSAN). IEEE 802.15.4e standard is designed to support the complicated development of WSN/WSAN. It officially defines multi-channel communication protocols as part of the standard. In the standard, multi-channel communication can be categorized into beacon-enabled as example of Deterministic & Synchronous Multi-Channel Extension (DSME) and non-beacon enabled as example of Asynchronous Multi-Channel Adaption (AMCA). Traditionally it has flexibility on variable unit data length for MAC layer on non-beacon enabled WSN but less performance balanced due to unpredictable waiting time for network access to every node in the network. With the introduction of the AMCA, it can run with different MAC unit data lengths in different channels in parallel so it’s getting more use cases. We propose enhancements based on native AMCA of the IEEE 802.15.4e and build up an energy analysis model for comparing of native AMCA and enhanced AMCA modes. There are main four approaches to our proposition and the algorithms are easy to be implemented on a micro controller system as a practical method for real deployment of WSN. We finally use OPNET network simulator software to form networks with different node sizes under mesh and cluster-tree hybrid topologies as simulation scenarios. The results present our E-AMCA approaches have better performance than N-AMCA especially in the networks with high volume nodes.