認知網絡頻譜接入關鍵技術分析.pdf

fPJ ~ ~~ ~ {* W # rl:iJ 00 % ff :￥:if!) i l ~ tJ L 11J J! ~ H~ X 8<] ~ f!11~= ~o Fg r rW., ftititXflZ~I~~D1~1~ o /js:J\t~t Z.l~~Jffij( ￥IJJ ~~:Q-~~1ilitY:.8<] 萬方數據上海交通大學 學位論文原創性聲明 本人鄭重聲明：所呈交的學位論文 《認知網絡頻譜接入關鍵技術研究》，是本人在導師的指導下，獨立 進行研究工作所取得的成果。除文中已經注明引用的內容外，本論文 不包含任何其他個人或集體已經發表或撰寫過的作品成果。對本文的 研究做出重要貢獻的個人和集體，均已在文中以明確方式標明。本人 完全意識到本聲明的法律結果由本人承擔。 學位論文作者簽名： 日期： 年 月 日 萬方數據上海交通大學 學位論文版權使用授權書 本學位論文作者完全了解學校有關保 留、使用學位論文的規定，同意學校保留并向國家有關部門或機構 送交論文的復印件和電子版，允許論文被查閱和借閱。本人授權上海 交通大學可以將本學位論文的全部或部分內容編入有關數據庫進行檢 索，可以采用影印、縮印或掃描等復制手段保存和匯編本學位論文。 保密 □，在 年解密后適用本授權書。 本學位論文屬于 不保密 □。 （請在以上方框內打“ √ ” ） 學位論文作者簽名： 指導教師簽名： 日期： 年 月 日 日期： 年 月 日 萬方數據上海交通大學碩士學位論文 I 認知網絡頻譜接入關鍵技術研究 摘 要 動態頻譜接入技術是認知無線 網絡研究的重要內容之一，其實現方式包括動態授權和動態頻譜共享等。 動態頻譜共享技術將頻譜資源同時授予多個網絡，使頻譜資源得到共 享。本文的主要研究方向就是動態頻譜接入技術。 本文將主要分析， 當多個頻譜資源由多個網絡 （用戶）共享時，用戶應該采用怎樣的頻譜接入策略。 本文首先分析了考慮信道切換 損耗的單用戶多信道網絡。在用戶無法獲知信道質量先驗信息的情況下， 通過假設信道在不同時刻的信道質量具有參數化的獨立同分布特性，本課題提出了一種改進的Lai-Robbins 接入策略。該策略將用戶多信道接入問題建模為 MAB（ Multi-Armed Bandit）問題，以實現用戶收益的最大化。通過理論分析證明，該算法的收益損失隨著時間增 長的速度能夠達到理論上最優的對數階數。考慮到上述算法在實現過 程中存在頻繁的信道切換，因而存在較大的增長系數， 為了減少收益損失， 本文提出了 RSAP（ Reduce Switch with Advanced Play）策略，并給出了 RSAP 策略的收益損失的理論分析。在 RSAP 策略中，收益損失隨著時間依然以對數階數增長，但是由于較低的信道切換次數而具有較低的增長系數。 接著本文對多用戶異構網絡的 動態頻譜接入技術進行了分析。在異構網絡中，用戶在接入不同類型的接 入點時具有不同形式的收益函數，且同一信道對不同接入用戶具有不同信道 質量。在本課題分析中，允許多個用戶同時接入同一個信道。當 多個用戶接入同一個信道時，所有接入該信道的用戶的收益均會下降 。本文將該問題建模為收益函數與用戶相關的非對稱單元素擁塞博弈 問題，并證明了該博弈問題存萬方數據上海交通大學碩士學位論文 II 在純策略納什均衡點。由于收益函數未 知，這是一個不完全信息博弈問題，用戶需要在博弈的動態過程中學 習外界環境，逐漸找到對自己最有利的策略。為了解決該問題，本文 提出了較優響應分布式學習算法，并且證明在該算法下，用戶的策 略最終將收斂到納什均衡點。 關鍵詞： 認知無線電，動態頻譜接入策略， MAB 問題，非對稱單元素擁塞博弈，納什均衡 萬方數據上海交通大學碩士學位論文 III RESEARCH ON DYNAMIC SPECTRUM ACCESS IN COGNITIVE RADIO ABSTRACT Since spectrum resources are becoming more significant and rarer in wireless communication, researchers proposed dynamic spectrum access to improve efficiency of spectrum usage. In dynamic spectrum access, users are allowed to share spectrum in various s, which is called spectrum access policy. It’s very common that different network uses different spectrum access policy. In wireless communication system, getting an appropriate spectrum access policy is very crucial, which affects not only the throughput of network but also the QoS. In this paper, we focus on a scenario in which users are not ined about the channel quality in network. So the users must learn about the network in order to maximize rewards. Our research consists of two parts. In the first part, we consider a wireless network with only one user. The novel point is that we take switch cost into consideration. We suppose the operation of switch will reduce the user’s rewards due to the time delay and extra energy consumption. We model this problem as the multi-armed bandit problem. Firstly, a modified Lai-Robbins policy is proposed. We find the regret under modified Lai-Robbins policy grows with large leading constant due to the frequent channel switch. Then we proposed the RSAP (Reduce Switch with Advanced Play) policy, under which the regret is proved to grow with much smaller leading constant. Since heterogeneous network has become the trend of future wireless communication system, in the second part, we propose a distributed learning algorithm for heterogeneous wireless network. In heterogeneous network, the situation is very complicated because of the following reasons. First, it’s a distributed system, so there is no control center. Second, the 萬方數據上海交通大學碩士學位論文 IV payoff functions are various depending on different types of base stations. Third, users do not know the payoff functions exactly, since channel quality and actions of others are not known. We model the dynamic spectrum access problem in heterogeneous network as the ASPS (Asymmetric Singleton congestion game with Player-Specific payoff functions). It’s proved that the ASPS has Nash equilibrium. Then we propose the better-reply distributed learning algorithm. We can prove that under the better-reply distributed learning algorithm, the ASPS problem will converge to Nash equilibrium. KEY WORDS: cognitive radio, dynamic spectrum access, multi-armed bandit problem, congestion game, Nash equilibrium 萬方數據上海交通大學碩士學位論文 V 目 錄 第一章 緒論 ················································································································· 1 1.1 本文研究的背景與意義 ····················································································· 1 1.1.1 認知無線電技術的來源與意義 ·································································· 1 1.1.2 認知無線電的概念 ······················································································ 2 1.1.3 認知無線電架構 ·························································································· 3 1.1.4 認知無線電標準化進程 ·············································································· 6 1.2 本文研究內容 ····································································································· 7 1.2.1 動態頻譜接入技術 ······················································································ 7 1.2.2 本文研究目的 ······························································································ 9 1.3 本文篇章結構 ··································································································· 10 第二章 MAB 模型和博弈論基礎 ·············································································· 11 2.1 引言 ·················································································································· 11 2.2 MAB 問題 ········································································································· 11 2.2.1 MAB 模型 ·································································································· 11 2.2.2 Lai-Robbins 策略及其性質 ········································································ 12 2.3 博弈論基礎 ······································································································· 15 2.3.1 博弈問題的定義 ························································································ 15 2.3.2 混合策略與純策略 ···················································································· 17 2.3.3 納什均衡 ···································································································· 17 2.4 本章小結 ·········································································································· 17 第三章 單用戶頻譜接入策略 ···················································································· 19 3.1 引言 ·················································································································· 19 3.2 單用戶系統的 MAB 模型 ················································································ 19 3.2.1 系統模型介紹 ···························································································· 19 3.2.2 頻譜接入策略性能指標 ············································································ 20 3.3 改進的 Lai-Robbins 策略 ················································································· 21 3.3.1 改進的 Lai-Robbins 策略流程 ·································································· 21 3.3.2 改進的 Lai-Robbins 策略的性能分析 ······················································· 24 萬方數據上海交通大學碩士學位論文 VI 3.4 RSAP 策略 ········································································································ 28 3.4.1 RSAP 策略流程 ·························································································· 29 3.4.2 RSAP 策略的性能分析 ·············································································· 32 3.5 實驗仿真結果 ··································································································· 35 3.6 本章小結 ·········································································································· 39 第四章 多用戶異構網絡頻譜接入策略 ···································································· 40 4.1 引言 ·················································································································· 40 4.2 系統模型 ·········································································································· 41 4.2.1 網絡模型 ···································································································· 41 4.2.2 博弈模型 ···································································································· 42 4.3 純策略納什均衡點的存在性 ··········································································· 43 4.4 較優響應分布式學習算法 ··············································································· 46 4.5 仿真結果及分析 ······························································································· 49 4.5.1 仿真的參數設置 ························································································ 49 4.5.2 仿真的收斂性分析 ···················································································· 51 4.6 本章小結 ·········································································································· 53 第五章 結束語 ··········································································································· 54 5.1 主要工作與創新點 ··························································································· 54 5.2 后續研究工作 ··································································································· 54 參 考 文 獻 ··············································································································· 55 附錄 1 英文縮寫語表 ································································································· 58 附錄 2 概率密度函數所滿足條件 ············································································· 59 致 謝 ························································································································· 60 攻讀碩士學位期間已發表或錄用的論文 ·································································· 62 萬方數據上海交通大學碩士學位論文 VII 圖 錄 圖 1-1 頻譜感知 CR 的認知環 ···················································································· 3 圖 1-2 集中分布架構示意圖 ························································································ 6 圖 1-3 動態頻譜接入技術分類圖 ················································································· 8 圖 2-1 Lai-Robbins 策略 ···························································································· 14 圖 3-1 N 個信道的單用戶模型 ··················································································· 20 圖 3-2 時間 T 與虛擬時間 s 的關系示意圖 (次級用戶在 T=3,5 切換信道 ) ············· 22 圖 3-3 考慮信道切換損耗條件下的 Lai-Robbins 策略 ············································· 23 圖 3-4 RSAP 策略 ······································································································· 30 圖 3-5 N=2， K=2， Ct=1 時的 RSAP 策略實例的示意圖 ········································· 31 圖 3-6 存儲接入策略 ································································································· 33 圖 3-7 不同策略下信道 2 被選擇的次數 ·································································· 37 圖 3-8 不同策略下信道 3 被選擇的次數 ·································································· 37 圖 3-9 不同策略下信道切換的次數隨著時間的增長曲線 ······································ 38 圖 3-10 不同策略下收益損失的增長系數 ································································ 39 圖 4-1 多用戶異構網絡示意圖 ··················································································· 42 圖 4-2 學習算法中的用戶幀結構 ··············································································· 48 圖 4-3 較優響應分布式學習算法 ·············································································· 49 圖 4-4 基站與用戶分布示意圖 ·················································································· 50 圖 4-5 較優響應分布式學習算法下策略的分布 ······················································ 52 萬方數據