無線網路系統之節能無線電資源配置

Abstract

隨著無線網路技術的快速發展與行動裝置爆炸性的成長，移動使用者已開始偏好多媒體服務，而移動使用者對於多媒體服務的使用行為將使行動裝置消耗大量的能源在接收多媒體資料。有鑑於以上的觀查，本論文研究無線網路系統之節能無線電資源配置，主要的目標是最小化行動裝置在接收資料的能耗。在本篇論文，我們首先著重節能的時槽分配問題於視訊群播，且同時考量分層式視訊編碼和可調式調變編碼技術。我們證明此問題無論是採用scalable video coding (SVC) 或multiple description coding (MDC) 都為一個NP-hard 之問題，除非P = NP。針對此問題在考量SVC 之下，我們提出一個2 倍之近似演算法，當MDC 技術應用於此問題，我們證明不存在優於2 倍的近似演算法且提出一個假多項式時間(pseudo-polynomial time) 之2 倍近似演算法。接著，我們發現調整基地台傳輸的功率，有助於減少行動裝置在接收資料時的能耗。為了達到在行動裝置端的節能，我們因此進一步研究功率和時糟分配問題於下行資料傳輸。我們證明此問題為一個NP-hard 且不存在優於4/3倍之近似演算法，除非P = NP。為此，我們提出一個2倍之近似演算法。更進一步的，我們延伸功率和時糟分配問題，在考量行動裝置更為普遍計算能耗的尺制之下，提出一個演算法，且證明演算法的效能和最佳解之間的差距。最後，透過一連串的實驗來評估所提方法的性能，以提供更多有助於無線網路系統節能資源分配設計之觀點。

With rapid advance of wireless networking technologies and explosive growth of mobile devices, mobile users have become addicted to multimedia applications with extremely downlink-intensive transmissions. Such the usage behavior leads to significant energy consumption in mobile devices to receive downlink data. The observation motivates us to address energy-efficient radio resource allocation in wireless systems. The objective is to minimize total energy consumption of mobile devices for data reception. In this dissertation, we first target the energy-efficient slot allocation problem for video multicast with the consideration of layer-based video coding and adaptive modulation-coding scheme. We prove that the problem is NP-hard when scalable video coding (SVC) or multiple description coding (MDC) is applied. We propose a 2-approximation algorithm for the problem with SVC. When MDC is applied, we prove the problem cannot be approximated within a ratio better than 2 and propose a pseudo-polynomial time 2-approximation algorithm to solve the problem. Then, we observe that the adjustment of the transmit power of a base station is helpful to reduce the energy consumption of mobile devices. We further study the power allocation as well as the slot allocation for downlink data transmission to achieve energy efficiency in mobile devices. We prove the problem is NP-hard and cannot be approximated in polynomial time with a ratio better than 4/3, unless P = NP. Then we propose a 2-approximation algorithm to solve the problem. Finally, we extend the power and slot allocation problem to consider a more general metric for energy consumption of mobile devices. The solution is proposed and proved that its performance gap to an optimal solution is tightly bounded. The capabilities of the proposed algorithms mentioned above are evaluated by a series of simulation experiments, which provides insightful and encouraging results in energy-efficient radio resource allocation design for wireless systems.