周俊廷臺灣大學:電信工程學研究所王有聖Wang, Yeo-ShengYeo-ShengWang2010-07-012018-07-052010-07-012018-07-052009U0001-1808200914424200http://ntur.lib.ntu.edu.tw//handle/246246/188348在許多現今的無線網路中，裝置之間共同協調來完成某些任務，對網路的運作和個別裝置的效能而言是非常關鍵的（例如：在節省電力的網路中，裝置同時進入非睡眠模式以進行溝通、裝置與裝置在多頻帶網路能找到對方）。為了達成這些協調，裝置通常必須藉著互相交換訊息以調整個別的行為。然而，在許多特殊的環境裡，由於硬體上的限制、網路拓樸或是管理法規上的限制，裝置間要交換訊息並不是一件容易的事。本論文中，我們設計一種新穎的演算法。透過此演算法，裝置間可以在不交換任何訊息、不需要時間上同步或是第三者幫助的情況下，達成某種型態的協調。此方法最基本的概念是各裝置透過一種特殊設計的序列來建立自己的“任務排程”，而這些序列是由各裝置獨立產生的。有了這些序列，裝置之間可以在完全分散式的架構下共同來完成特定的任務。我們的演算法最直接的應用便是在以動態頻譜存取為基礎的通訊技術中，用來偵測授權裝置的方法。這種以動態頻譜存取為基礎的通訊技術需要次級裝置共同偵測和保護授權裝置，使其避免受到干擾。我們的演算法可以保證這些次級裝置，即使在互相聽不到對方的條件下，仍然能有效率的協調偵測以保護授權裝置。們建立一套數學模型來分析所提出的演算法，並透過數值分析來說明它的效能與保證。為了呈現我們方法的可行性，我們研究兩個例子：（1）電視空白頻譜通訊中授權裝置的偵測（2）WiMax與UWB的共存問題。根據這兩個例子，我們利用Opnet模擬所提出的演算法，並證明我們的方法能夠在多變的網路環境中，100%的保護這些授權裝置。In modern wireless networks, coordination among devices is crucial for network management andndividual devices'' performance (e.g., effective power saving, or neighbor discovery in aulti-channel network). In order to establish coordination, devices usually have to exchangenformation with each other and adjust individual behaviors accordingly. However, in manyractical environments, coordination via exchange of information is infeasible due toardware limitation, network topology or regulatory constraints.n this thesis, we design new algorithms for devices to have certain form of coordination withoutny information exchange, temporal synchronization, or third-party assistance. The basic ideaf the proposed scheme is that each device establishes its own "task schedule" via a specially designedequence. Each sequence is locally and independently determined by individual devices. With suchequences, devices can jointly complete the task in a fully distributed manner. An immediate applicationf our algorithms is the detection of licensed radio users in dynamic spectrum access (DSA)-basedommunication. DSA-based communication requires secondary devices to jointly detect and protect licensed usersrom interference in a timely manner. Our algorithms guarantee that secondary devices, even whenidden to each other, achieve coordinated detection to protect the licensed users effectively.e develop a mathematical model to analyze our algorithms and conduct numerical analysis to show theerformance guarantee. To demonstrate the feasibility of our solution, we study two cases: (1) TV-bandhite-space communication and (2) coexistence between WiMax and UWB networks. We simulatehe proposed algorithm in these two cases using the Opnet Modeler, and show that our algorithmsrovide 100\% protection to the licensed users in various network environments.ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iiIST OF TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viiIST OF FIGURES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viiiHAPTER 1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . 1.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 The main objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.1 Interactive tasks . . . . . . . . . . . . . . . . . . . . . . . . . 2.2.2 Blind tasks . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.3 Related work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6.3.1 Asynchronous wake-up protocol . . . . . . . . . . . . . . . . 7.3.2 Quorum-based protocol . . . . . . . . . . . . . . . . . . . . . 7.3.3 Slotted seeded channel hopping (SSCH) protocol . . . . . . . 7.3.4 Asynchronous neighbor discovery and rendezvous . . . . . . . 8.4 Thesis organization . . . . . . . . . . . . . . . . . . . . . . . . . . . 9HAPTER 2 CASE 1: DETECTION IN TV-BANDWHITE SPACEOMMUNICATION . . . . . . . . . . . . . . . . . . . . . . . . . . . 10.1 FCC’s rules . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.2 TVBDs objectives in detection . . . . . . . . . . . . . . . . . . . . . 11.3 Detection problems . . . . . . . . . . . . . . . . . . . . . . . . . . . 12.3.1 Alien secondary devices . . . . . . . . . . . . . . . . . . . . . 12.3.2 Mismatched coverage . . . . . . . . . . . . . . . . . . . . . . 14.3.3 Hidden-node interference . . . . . . . . . . . . . . . . . . . . 14.4 Potential solutions? . . . . . . . . . . . . . . . . . . . . . . . . . . . 16.5 PHY-layer consideration . . . . . . . . . . . . . . . . . . . . . . . . 17HAPTER 3 ASYNCHRONOUS COORDINATION ALGORITHMSED IN DETECTION FOR TV-BAND WHITE SPACE COMMUNICATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19.1 Asynchronous coordination algorithm . . . . . . . . . . . . . . . . . 19.2 Proof of performance guarantee . . . . . . . . . . . . . . . . . . . . 23.3 Mathematical analysis . . . . . . . . . . . . . . . . . . . . . . . . . . 28.4 Real-valued offset . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33.5 Simulation and performance evaluation . . . . . . . . . . . . . . . . 34.5.1 Scenario I: the influence of clock drifting . . . . . . . . . . . 34.5.2 Scenario II: varying of detection time with different parameters 35.5.3 Scenario III: the length of overlapping silent periods with differentlot length . . . . . . . . . . . . . . . . . . . . . . . . . 38.6 Extension to frequency-domain detection . . . . . . . . . . . . . . . 39HAPTER 4 CASE 2: COEXISTENCE PROBLEM BETWEENIMAX AND UWB DEVICES . . . . . . . . . . . . . . . . . . . . 42.1 Detection of WiMax activity by UWB devices . . . . . . . . . . . . 44.2 Initial entry of WiMax CPEs . . . . . . . . . . . . . . . . . . . . . . 45.2.1 Step 1: scanning and synchronization . . . . . . . . . . . . . 45.2.2 Step 2: obtain UL parameters . . . . . . . . . . . . . . . . . 47.2.3 Step 3: initial ranging . . . . . . . . . . . . . . . . . . . . . . 47.3 Time limitation of initial entry . . . . . . . . . . . . . . . . . . . . . 48.4 Some potential problems . . . . . . . . . . . . . . . . . . . . . . . . 50HAPTER 5 ASYNCHRONOUS COORDINATION ALGORITHMSED IN COEXISTENCE PROBLEMBETWEENWIMAX ANDWB DEVICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54.1 Approach I: randomization . . . . . . . . . . . . . . . . . . . . . . . 54.2 Approach II: exhaustive repetition . . . . . . . . . . . . . . . . . . . 55.3 Proof for the existence of moving-left slot for Approach II . . . . . . 65.4 Simulation and performance evaluation . . . . . . . . . . . . . . . . 68.4.1 Consideration of Ls and N . . . . . . . . . . . . . . . . . . . 70.4.2 Slot length vs. detection time . . . . . . . . . . . . . . . . . 73.4.3 Duty cycle vs. average detection time . . . . . . . . . . . . . 74.4.4 Duty cycle vs. maximum detection time . . . . . . . . . . . . 75HAPTER 6 CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . 77949880 bytesapplication/pdfen-US非同步協調動態頻譜存取分散式授權裝置偵測asynchronouscoordinationDSAdistributedunlicenseduserdetectionthesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/188348/1/ntu-98-R96942108-1.pdf