以鏈結層模型解法建立一個高效能無線網狀網路---參數分析、演算法設計、及雛型系統實作

Abstract

由於單點跳躍無線區域網路的固有覆蓋範圍限制及高部署費用(來自於有線主幹下的每一個有效的存取點的硬體需求)，在過去十年的期間，大家開始注意無線多點跳躍協力(wireless cooperative multi-hop)的無線隨意網路(ad hoc networking)，並且認為是一個可行的架構。定義上，一個無線隨意網路可視作使用非固定的寬頻網路架構(no fixed infrastructure)或中央控管機制(centralized)的一群無線主機來合力建立通訊。而近興起的無線網狀網路(wireless mesh network)則被視為無線隨意網路在應用層面所提出的無線網際網路最後一哩(last-mile)可行的解決方案，無線網狀網路的確打開了新的研究方向及可觀的數量在研究協定設計上，不過許多有關效能的問題仍尚待解決。過度的封包遺失(packet loss)、不可預測的通道行為(unpredictable)導致無法找出一個穩定且高流量的路徑(high-throughput paths)、內部資料流量(intra-flow)和資料間的互相干擾(inter-flow interference)導致網路流量的下降，都屬於尚未被完全解決的問題。 在這原規劃三年執行期間的計畫中，我們已獲得國科會一年期的補助，並且有一些初步研究成果，也因此打算繼續提出後兩年的補助申請，目標是針對鏈結層潛藏的效能問題做一系列的深入分析，並且提出可能問題解法之協定及效能改進的演算法，朝著優化無線鏈結的定義/特性以及本身上層協定設定最佳化的方向來提昇系統效能。我們希望可以(i)瞭解無線鏈結如何影響實體層(PHY)/媒體存取層(MAC)屬性和影響範圍以及其他環境因素的影響，(ii)實現一個嚴謹的分析和理論來描述無線鏈結的行為，和(iii)明確定義實體層/媒體存取層相關效能控制因素，使得網路容量能夠最佳化。具體來說，我們規劃為期3年的計畫其相關研究設計要點的步驟如下：

In light of the inherent coverage limitation and high deployment cost (incurred from the requirement of the available wired backbone behind each access point) of single-hop WLAN scenarios, during the past decade, the wireless cooperative multi-hop ad hoc networking has been noticed and considered as a promising operational architecture. By definition, a wireless ad hoc network is viewed as a collection of wireless hosts, which cooperatively establish communications, using no fixed infrastructure or centralized administration. The emerging wireless mesh network (WMN) is a variant of wireless ad hoc network proposed for last-mile wireless Internet access solution. Interests in WMNs have indeed opened up new research venues and led to a fair amount of research activities in protocol design, implementation, and deployment. However, a number of performance related problems have also been identified. Excessive packet losses, unpredictable channel behaviors, inability to find stable and high-throughput paths, and throughput degradation due to intra-flow and inter-flow interference are among those most cited while not yet fully exploited. We originally planned a 3-year execution period to realize this project, and have already obtained a 1-year sponsorship from NSC. Since we have preliminary results that support the promising aspects of our approaches, we therefore seek sponsorship to complete the technical contents of the rest two years we proposed. The main idea of this project aims to take a bottom-up link-layer approach. With the 3-year uninterrupted investigation model, we plan to tackle issues that work toward a better definition/characterization of wireless links and its implication for higher layer protocol design and optimization. We would like to (i) understand how, and to what extent, wireless links are affected by PHY/MAC attributes and other environmental factors, (ii) characterize the behavior of wireless links in such a way that they become amenable to rigorous analysis and reasoning, and (iii) identify control knobs in the MAC/PHY layers with which the network capacity can be optimized. Specifically, we plan to conduct our research along the following closely-related R&D thrusts: