下世代網路無線電資源效率性管理與節能之研究

Abstract

隨著新興無線通訊網路技術及廣電傳輸技術的持續進步，下世代無線網路之系統容量將能夠得到顯著的提升。然而，由於無線通道的基本物理特性，資料往往不能被成功地傳送到目的地。因此，尤其對於延遲敏感的交通流，如何於下世代無線網路中充分利用無線資源來設計快速且高效率的資料重傳機制是非常重要的。此外，以目標為導向的無線資源分配與管理及節能的設計可以有效地幫助提升整體系統各方面的效能。在此博士論文中，我們將探討下世代無線網路中之5個重要的效能改善研究議題，並做不同觀點探討其效率提升。

首先，我們設計搭配下世代無線網路多通道同時傳輸環境並限制重傳次數的機制，使延遲敏感流能做較有效用的傳輸。我們提出數種有效率的編解碼機制，分別結合此多通道環境來做重傳機制的設計。以最大化應用層流量為目標，根據通道品質好壞及封包大小，設計多種對應的最佳重傳策略。經由數學分析與模擬結果顯示，提出的重傳機制在有效流量及延遲效能均可明顯優於現存的重傳機制，且複雜度不高，不但省電也可節省硬軟體資源。

接著，我們於下世代無線網路多使用者合作性傳輸環境下針對延遲敏感交通流提出快速重傳的全新設計。使用此提出的最佳重傳策略，各方面系統效能均明顯優於未合作性傳輸模型，藉由合作性同時互相幫助傳輸，數學分析與模擬結果顯示在同一時間內可傳輸明顯較多的有效資訊量，非常省電。 再者，由於封包排程機制是處理資源分配問題的有效率手段之一，因此我們針對下世代多載波網路環境下分別提出兩種不同目標觀點的封包排程演算法設計，模擬結果顯示提出的機制能顯著提升整體系統的各種效能(如封包延遲、系統流量及公平性等)，並將頻寬有效率利用且維持演算法低複雜度，因此不但省頻譜資源也非常省電。 另一方面，目前商業化的3.5G網卡大多數為多模網卡，而選擇使用哪一種網路系統上網的行為模式，往往第一優先選擇為網路瞬間速度最快者，反而忽略了整體網路的使用狀況與使用效率。因此，我們提出一套基於額度的動態網路切換機制來將連網用戶之連線進行有效的資源分配，來改善目前3.5G網路面臨的資源限制。模擬結果顯示此機制能顯著提升整體系統滿意度及頻譜的使用效率並且降低使用者被拒絕連線使用的機率。 最後，由於各種無線網路技術的同步蓬勃發展與多重無線網路的大量建置，已在電力消耗上造成許多國家電信業者的沉重負擔，尤其是來自於基地台的耗能。因此，針對基地台網路架構下，我們分別針對滿意度與公平性之需求限制下提出兩種耗能最佳化的數學模型，接著分別提出對應的低複雜度優化節能演算法，數值結果顯示提出的傳輸演算法能顯著的節能，並滿足個別使用者的服務品質需求。

For next generation wireless networks equipped with promising technologies such as 4th generation (4G) cellular network, their system capacities are supposed to be significantly improved. Nevertheless, because of fundamental physical characteristics of wireless channels, data packets often cannot be delivered to the destination successfully. As a result, the design focusing on the efficient and fast retransmission scheme especially for delay-sensitive flows to sufficiently utilize the radio resource in such a next generation wireless network plays a highly crucial role. Additionally, good objective-oriented radio resource allocation and management as well as energy-saving designs can help to efficiently improve in the aspects such like the system performance, the utility, the power consumption, and etc, respectively. In this dissertation, we will focus on five potential issues for improving the efficiency in various points of view in the next generation wireless network. In the first topic, we present several multi-transceiver multi-channel fast packet retransmission schemes intended for transporting delay-sensitive flows in a multi-channel network environment. The proposed schemes are designed to allow the retransmission(s) for only one time using one or multiple channels simultaneously unlike other ones where the retransmission(s) of a link packet can only be in one channel and continue until it is successfully received. By using the application throughput as the objective, the optimal retransmission policy can be determined based on the estimated channel quality and the application-layer packet size. The optimized retransmission schemes are shown able to achieve better effective throughput (goodput) than those of other ones in various fading environments. The presented multi- transceiver multi-channel system model and fast retransmission schemes can be applied to the long term evolution-advanced network environment, in which the aggregation of multiple component carriers (CCs) or spectrum is considered an alternative to achieve significant throughput improvement. The second topic addresses the design problem of a fast packet retransmission scheme intended for transporting delay-sensitive flows in a cooperative diversity (CD) network environment. This cooperative fast retransmission scheme exploits the advantages of the CD network environment, while allowing retransmission just one time via a cooperating user (i.e., partner) or via both the sender and the partner simultaneously. Complementary link packets are used for the retransmission whose policy can be adjusted on the basis of the qualities of channels among the sender, the partner and the receiver, as well as the application-layer packet size, using the application throughput as the objective. The CD-based optimized fast retransmission scheme is shown able to achieve better effective throughput than other CD-based or non-CD-based retransmission schemes in various fading environments. As a result, the proposed scheme should be an excellent fast retransmission mechanism for real-time multimedia transport in many CD environments. In the study of the third topic, we aim to explore efficient packet scheduling schemes for users in multi-CC network environments. Two packet scheduling schemes are presented on the basis of the quantized water-filling criterion and the proportional fair criterion, respectively. The quantized water-filling packet scheduling scheme is designed to intentionally minimize the mean packet delay, where a close upper bound of the mean packet delay is derived, while maintaining the delay fairness among users, when the traffic load is unsaturated. The proportional fair based packet scheduling scheme is designed to improve the overall system performance, while maintaining fairness among all users. These two schemes are shown have much better performance than those of a network where CCs are not aggregated but used independently. The forth topic presents a dynamic network selection scheme via a quota-based admission control design for accommodating access requests in a multiple heterogeneous and orthogonal network environment, which may become common for today’s 3G operators. The design philosophy of this scheme is to let the system utilization in the fastest networks be statistically balanced and sufficiently utilized. Performance metrics in terms of the overall system utility (i.e., the satisfaction index), the blocking probability, and the system utilization are investigated. From simulation results, it is shown that the proposed scheme has significant improvement in the aspect of the overall system performance than that of existing approaches. Last, we study the topic of a power optimization model developed with respect to the radio resource allocation and the activation in a multi-CC network environment. We first formulate and solve the power-minimization problem of the base station (BS) transceivers for multiple-CC network environment, while maintaining the overall system and respective users’ utilities above minimum levels. The optimized power consumption based on this model can be viewed as a lower bound of that of other algorithms employed in practice. A suboptimal scheme with low computation complexity is proposed. Numerical results show that the power consumption of our scheme is much better than that of the conventional one in which all CCs are always active, if both schemes maintain the same required utilities. Next, we formulate a power-minimization problem of the BS transceivers for multiple-CC networks, while maintaining respective user types’ fairness indexes and respective users’ data rates. An efficient scheme is subsequently proposed to solve it. Numerical results demonstrate that the total power consumption of our proposed scheme is significantly much better than that of the case where all CCs are always active when the traffic load is relatively light.