根據 Federal Communications Commission(FCC)的研究報告指出, 任一時刻超過70% 的已授權頻譜(licensed spectrum)沒有在使用。然而未授權的頻譜(unlicensed spectrum) 資源已被新興的無線服務和應用完全耗盡。感知無線電(cognitive radio)技術為這個問題提供了可能的答案。感知無線電是一種能感測閒置頻道的無線設備。非授權頻段的使用者(secondary user)可利用感知無線電得知已授權頻段是否閒置，當其閒置時，非授權頻段的使用者便可進行傳輸而不會干擾到原本授權頻段的使用者(primary user)。 擁有感測頻道能力的網路稱為感知網路(cognitive network). 大部分在感知網路上的研究集中在實體層(physical layer)與媒體存取層(medium access layer)。而傳輸層上仍有許多未開發的研究議題. 本子計畫之研究目的是為感知無線隨意行動網路(cognitive ad-hoc networks)設計出良好的傳輸協定(transport protocol)，並於三年的執行時間中逐年增加研究要素。第一年計畫著重於增加感知無線隨意行動網路之傳輸吞吐量(throughput)。主要的作法是利用從底層獲得之閒置頻寬資訊(spectrum-aware)，改進End to End的傳輸效能。第二年則加入合作(cooperative)的概念，探討與設計合作式的傳輸協定，讓資源(頻道)的分配與網路中每個節點的工作負擔更為平均，同時也讓傳輸的品質更為提昇。除此之外，我們將導入博弈理論(game theory)的概念，期以達到更有效率的設計，第三年的計畫將會以省電與降低延遲為主題，以重視延遲時間(delay-sensitive)及限制延遲時間(delay-constrained)兩種面相來討論。我們打算結合第二年設計合作式傳輸協定的經驗，以減少頻道感測時間、預測其他節點傳輸行為、降低節點干擾與減少失敗傳輸等方向進行。除此之外，本子計畫將會依照各年之研究議題，實作、設計與開發實驗平台作為驗證各年之研究成果的實際效能。According to studies sponsored by Federal Communications Commission (FCC), over 70% of the allocated spectrum is not in use at any time even in a crowded area where the spectrum usage is intensive. The remaining portion of the unlicensed spectrum is being exhausted by emerging wireless services and applications, leading to the so-called spectrum scarcity problem. Cognitive radio (CR) techniques provide a solution to this problem. A CR is a self-managing device that is capable of sensing radio spectrum and opportunistically utilizing unoccupied frequency bands. A CR can access licensed frequency bands without interfering with its primary users (PUs) by alternating between a spectrum sensing state and an available spectrum accessing state. This cognitive ability of a CR offers an excellent opportunity to secondary users (SUs) for accessing frequency bands not licensed to them. A network with cognitive capabilities is called a cognitive network. Vast majority of past and current research and developments for cognitive networks are in the lower two layers. While the research and development for the physical layer and the medium access layer have been advancing in tandem with the CR research, the research for development of the transport layer protocols for cognitive networks is still an unexplored area. This project is intended to identify requirements of protocols for the transport layer of cognitive networks, and design, implement, and evaluate transport protocols. The first-year project is intended to propose a spectrum-aware transport protocol to maximize the throughput in competitive cognitive networks, where multiple SUs compete with each other for resources (transmit power, rate, channel). The second-year project is intended to propose a transport protocol for solving spectrum allocation problem in cooperative cognitive networks, where SUs cooperate for achieving the same goal (i.e., fairness for spectrum allocation). The third-year project is intended to solve the power-saving problem in cooperative cognitive networks. We aim to design power-saving protocols for delay-sensitive applications and delay-constrained applications, respectively. Besides, we try to develop an experimental platform for cognitive networks. By the aid of the experimental platform, we are able to evaluate the performance of our proposed protocols. 研究期間：10008 ~ 10107
