正交分頻多工無線蜂巢式系統下上行高能源使用效率通訊方法之研究

Abstract

本論文旨在研究在正交分頻多工無線蜂巢式系統下利用上行資源配置方法增進手機使用者端能源使用效率問題。在此，能源使用效率意即無線手持裝置每傳送一單位位元所消耗的焦耳數。本論文分別探討了各種不同的使用情境及系統環境下的能源使用效率問題，透過發展一系列上行資源配置演算法，來達到最大的系統總能源使用效率。本論文針對單細胞蜂巢式系統提出低複雜度上行資源配置演算法以最大化平均能源使用效率，並同時考慮了資源配置公平性的條件。再者，在單細胞蜂巢式系統下本論文亦探討了考量即時能源使用效率的最大化問題，針對單一以及多個使用者情境並同時考量使用者的服務品質提出了不同的上行配置演算法。除一般正交分頻多工系統之外，本論文亦針對正交分頻多工協同式中繼系統，提出了上行資源配置方法以最大化平均能源使用效率。針對多細胞蜂巢式系統，本論文針對非合作式與協同式基地台，分別提出了上行資源演算法以大幅改善平均系統總能源使用效率。再者，本論文針對機器對機器通訊於非頻率選擇性衰減之正交分頻多工蜂巢式系統下的能源消耗問題，並同時考慮資源配置量以及平均延遲的限制，提出了不同的分群通訊方法以求達到最小的總能量消耗。更進一步地，在頻率選擇性衰減的環境下，經由機器裝置之發射功率配置，以及透過發展針對機器裝置系統裡之協調者(coordinator)的上行資源配置，結合前述之分群通訊方法，可更加減低總能源消耗量。同時，所提出之機器對機器演算法亦可有效率地決定出可達到最小能量消耗之最佳分群組數。

In this dissertation, we investigate the problems of uplink resource allocation (RA) in OFDMA-based cellular networks for optimizing mobile users’ efficiency of energy utilization defined as number of bits transmitted per joule consumed, namely bits-per-joule (bpj) in short. Effective and low-complexity energy-efficient resource allocation algorithms are proposed under several wireless transmission scenarios and assumptions, including the single and multi-user scenario, single- and multi-cell OFDMA network, different QoS constraints, practical transmission format (modulation adaptation), partial CSI feedback, and the cooperative relay network. Both the time-averaged and instantaneous energy efficiency (bpj), expressed as the average (instantaneous) transmission rate over the average (instantaneous) power consumption, are investigated. We propose a low-complexity 2-dimensional uplink RA scheme with users’ minimum assigned resource unit (RU) requirements for maximizing the sum of MSs’ time-averaged bpj in an OFDMA cellular network. Then, the average energy efficiency (EE) maximization problem is re-formulated in the practical OFDMA uplink system where the discrete transmission rate is achieved with different modulation scheme. We defined so-called energy-efficiency gap (EEG) to capture the performance loss for the practical transmission compared to the ideal transmission (Shannon capacity), and proposed a sub-optimal centralized two-stage RA scheme. The average EE maximization problem is also addressed in the OFDMA-based cooperative DF relay network. We developed a low-complexity subcarrier-power allocation and relay selection scheme in which the maximum ratio combining (MRC) is exploited. Also, instead of considering average EE, we address the RA problem of maximizing instantaneous bpj with the consideration of users’ QoS requirement, expressed as maximum power budget and minimum rate constraint. Then, we extend the scenario of the problem to the uplink multi-cell OFDMA network where each base station (BS) has no information about users outside its serving cell, namely that BSs can only work in a non-cooperative manner. Both the ideal (Shannon Capacity) and practical transmission format are considered. The non-cooperative distributed power allocation scheme and centralized RA and modulation adaptation (MA) scheme are proposed. Furthermore, in the multi-cell BS-coordinated OFDMA network, we develop a centralized joint modulation adaptation (MA), SC assignment (SA), and power allocation (PA) scheme for maximizing network energy efficiency, in which subcarrier-based coordination games are formed. Additionally, an effective partial CSI feedback scheme is devised and incorporated. Tradeoffs between energy efficiency and spectral efficiency and the advantages of the proposed schemes are shown. Moreover, we also study the energy minimization problem for heterogenous machine-to-machine (M2M) communications in an OFDMA cellular network embedded with a M2M system where machines communicate with each other via a different wireless protocol. We propose several group-based (clustered) M2M schemes which apply the 2-hop hierarchy transmission architecture to perform massive access control (MAC) with the resource amount and average delay constraint. On top of the group-based M2M, we develop joint MAC and RA schemes, which cluster MNs, perform RA for coordinators, and determine the proper number of groups, to minimize total energy consumption in both flat- and frequency-selective fading channel. Numerical results show the advantages of the proposed schemes. Tradeoff between EC, assigned resource amount, number of coordinators, and system delay are investigated.