Title: | WiMAX網路之資源與功率配置設計 Resource and Power Allocation in WiMAX Networks |
Authors: | 梁家銘 Liang, Jia-Ming 曾煜棋 林寶樹 Tseng, Yu-Chee Lin, Bao-Shuh Paul 資訊科學與工程研究所 |
Keywords: | IEEE 802.16標準;全球無線微波存取;點對多點網路;中繼網路;網狀網路;跨層設計;公平排程;區塊配置;資源管理;功率配置;時槽排程;傳遞延遲;IEEE 802.16;WiMAX;PMP network;relay network;mesh network;cross-layer design;fair scheduling;burst allocation;resource management;power allocation;slot scheduling;transmission latency |
Issue Date: | 2010 |
Abstract: | IEEE 802.16 (WiMAX)是新一代遠距無線網路標準,提供低廉的最後一哩網路存取,支援更高速的傳輸及更多樣的媒體服務。根據網路承載量及覆蓋範圍,WiMAX提供三種型態的網路架構,分別為: 1)點對多點網路架構、2)中繼網路架構及3)網狀網路架構。由於WiMAX網路下的行動用戶端較多、分佈較廣,因此WiMAX網路的資源由基地站統一負責管理,這使得WiMAX資源配置問題成為該網路上最重要的議題之一。本篇論文主要包含三個研究主題,分別探討WiMAX三種網路架構下的資源配置問題。第一個主題針對WiMAX點對多點網路,探討如何確保行動用戶端的服務品質問題。第二個主題考量在WiMAX中繼網路上,探討如何利用中繼站來協助行動用戶端傳輸,改善用戶端能源耗損的問題。最後,第三個主題針對WiMAX網狀網路,探討如何妥善配置資源以減少multi-hop傳遞造成的延遲問題。
在第一個研究主題中,我們觀察到WiMAX點對多點網路架構上的特殊二維訊框結構,該訊框結構下配置資源時會產生額外的資源浪費,因此容易造成網路效能降低。目前文獻中,主要的方法是利用單一的鏈結層排程單元或單一的實體層burst配置單元來減少資源的浪費,但兩者獨立運作下,不但運作複雜度無法降低外,實際可減少的資源浪費卻是有限的。因此,這個研究提出利用鏈結層的排程單元及實體層burst配置單元的協同設計,來有效降低運作的複雜度外,更減少訊框配置時造成的資源浪費。我們提出的跨層資源配置方法,包含一個二階權重的排程單元於鏈結層及廂型配置方法於實體層。此排程方法會根據用戶端的頻道品質、佇列資料、要求速率及容許延遲時間決定資料的優先權,再根據廂型配置法則將資料配置於二維訊框中。透過跨層的協同設計,可確保資料流的即時性及速率滿足度外,更可以公平分配行動用戶端的資源以及提升網路吞吐量。
在第二個研究主題中,我們指出在WiMAX中繼網路下,大部分研究採取提高行動用戶端的傳輸速率或增大平行傳送的個數以提升用戶端在中繼網路的吞吐量。但是,一旦用戶端使用更高的傳送速率或更多的平行傳送則會過度消耗他們的能源,這種設計對電池供電的用戶端格外造成傷害。因此,我們進而探討如何配置用戶端的頻寬資源及功率,滿足他們在上鏈傳輸的需求外,同時使他們耗用的能源最少。在這個研究中,我們首先說明這個問題是NP-complete,進而提出具節能的資源配置方法。首先,此方法會儘可能利用中繼站來協助用戶端傳輸,使更多的資源能有效的被利用。當有剩餘資源時,此方法會利用這些資源來調整用□端的傳送速率、路徑以降低用戶端的能耗。根據我們的了解,這是第一個在WiMAX中繼網路上,考量能源與頻寬配置的研究。
在第三個研究主題中,我們考量在WiMAX網狀網路上的微型時槽排程設計。我們認為,一個有效的時槽排程設計需要考慮到實體層的傳遞耗費、排程的運作複雜度及排程結果的訊息發佈耗費。我們特別針對長鏈狀及格狀網狀網路等擁有許多應用的拓蹼上作設計。和其他方法相比,我們的排程方法擁有較低的運作複雜度外,更能減少排程訊息發佈的耗費。其主要的貢獻在於利用一個簡易且具規律性的時槽配置規則,透過在傳送耗費及管線效應下取得平衡,使multi-hop傳遞產生的延遲得以減少,這個方法易於實作且效能媲美最佳結果。模擬結果說明了我們的方法能大幅的改善排程的運作時間,同時也維持較低的傳遞延遲。 The IEEE 802.16 (WiMAX) is developed to provide broadband wireless access. The standard provides three network architectures to support the last mile wireless access: 1) the point-to-multipoint (PMP) architecture, 2) the relay architecture, and 3) the mesh architecture. However, the resource allocation of these architectures is left open to the implementation. Thus, we propose the resource and power allocation for WiMAX networks, which includes three works. The first work considers the cross-layer resource allocation in WiMAX PMP networks to guarantee the quality of service (QoS) requirements of mobile subscriber stations (MSSs). The second work discusses the energy conservation issue in WiMAX relay networks. The third work focuses on the reduction of multi-hop transmission latency in WiMAX mesh networks. In the first work, we observe that the WiMAX PMP downlink subframes have a special 2D channel-time structure. Allocation resources from such a 2D structure incurs extra control overheads that hurt network performance. Existing solutions try to improve network performance by designing solely either the scheduler in the MAC layer or the burst allocator in the physical layer, but the efficiency of overhead reduction is limited. In this work, we point out the necessity of `co-designing' both the scheduler and the burst allocator to efficiently reduce overheads and improve network performance. We propose a cross-layer framework which contains a two-tier, priority-based scheduler in the MAC layer and a bucket-based burst allocator in the physical layer. The scheduler assigns priorities to MSSs' traffics in a two-tier manner and allocates resources to these traffics based on their priorities. The burst allocator divides the free space of each downlink subframe into a special structure which consists of several `buckets' and then arranges bursts in a bucket-by-bucket manner. Both the scheduler and the burst allocator are tightly coupled together and thus can significantly increase network throughput, maintain long-term fairness, alleviate real-time traffic delays, and improve frame utilization. In the second work, we point out that under WiMAX relay networks, existing studies only target at improving network throughput by increasing the transmission rates of MSSs or maximizing concurrent transmissions. However, using a higher transmission rate or allowing more concurrent transmissions could harm MSSs in terms of their energy consumption, especially when they are battery-powered. Therefore, we consider the energy-conserved resource allocation problem in the uplink direction of a WiMAX relay network. This problem asks how to arrange the frame usage with satisfying MSSs' demands as the constraint and minimizing their total energy consumption as the objective. We prove this problem to be NP-complete and develop an energy-efficient heuristic. The heuristic first exploits relay stations to allow more concurrent uplink communications to improve the transmission efficiency. When there are remaining resources, the heuristic lowers some MSSs' transmission power by adjusting their transmission rates and paths to save their energy. To the best of our knowledge, this is the first work considering both energy and bandwidth allocation in a WiMAX relay network. In the third work, we consider the mini-slot scheduling problem in WiMAX mesh networks. An efficient mini-slot scheduling needs to take into account the transmission overhead, the scheduling complexity, and the signaling overhead to broadcast the scheduling results. We are interested in chain and grid-based WiMAX mesh networks, which are the basic topologies of many applications. We propose scheduling schemes that are featured by low complexity and low signaling overhead. Compared to existing works, this work contributes in developing low-cost schemes to find periodical and regular schedules that achieve near-optimal transmission latencies by balancing between transmission overhead and pipeline efficiency that are more practical and easier to implement. In this work, we show that our schemes can significantly improve over existing works in computational complexity while maintain similar or better transmission latencies. |
URI: | http://140.113.39.130/cdrfb3/record/nctu/#GT079555832 http://hdl.handle.net/11536/41427 |
Appears in Collections: | Thesis |
Files in This Item:
If it is a zip file, please download the file and unzip it, then open index.html in a browser to view the full text content.