在蜂巢式機器通訊網路下之高能量效率分群演算法與效能分析

Abstract

隨著越來越多的機器具備通訊功能， 近年來IEEE 802.16 和3GPP也開始關注機器對機器（M2M）通訊，並積極的在訂定相關的標準，然而，目前也只有對可能的系統架構、特色、議題做討論，還沒有一個明確的規格，因此有賴我們去定義問題並解決。現在大部分的機器都是電池供電，加上機器的數量遠超過人口數，低上行功耗和大量存取的管理對機器對機器傳輸更顯的重要。在此論文中，我們提出分群和協調者選擇的演算法來解決上述的兩個問題；此外，為了考慮協調者和基地台之間的通道狀況，更引入了加權因子(weighting factor）；最後，由於協調者的功耗遠大於其他非協調者的功耗，我們也對此設計了功率配置的方法，使協調者可以更有效率的分配能源。

As more and more machines are networked, machine-to-machine (M2M) communication have captured attention of IEEE 802.16 and 3GPP which, however, have only described possible system architecture, features, and issues, but no specification of M2M communication so far. Therefore, it is important for us to define the problems and provide solutions. Since most of the machine devices (MDs) are battery equipped, and the number of MD is expected to be significantly more than human population, there are two critical requirements in M2M communications, low uplink energy consumption (EC) and massive access management. In this thesis, we propose MD access control algorithms to ensure low uplink EC of M2M communication under the constraint of maximum number of access from MD to BS by implementing grouping and coordinator selection. Moreover, weighting factor which takes channel condition between the coordinator and base station (BS) into consideration is introduced to further improve EC performance. Finally, since EC of coordinator is dominant compared with EC of MDs, Newton Method is applied to implement power allocation on the coordinators to allocate the power resource better.