在感知無線電網路下基於拍賣架構之頻譜管理

Abstract

感知無線電科技被視為可以有效解決頻帶主要無線電網路對頻譜使用率普遍偏低的問題。然而為了實現感知無線電的精神，感知無線電用戶在使用頻帶主要網路所擁有之頻譜帶時，勢必提供服務提供者及頻帶主要無線電網路額外的獎勵或利益，這相對提升了頻譜管理的難度。在本論文，我們考慮感知無線電基地台、服務提供者以及感知無線電用戶所組成的架構。我們提出一個基於拍賣模型來特徵化以及分析一些感知無線電使用者在感知無線電網路下之動態頻譜管理既有的問題特色(例如:感知無線電使用者之間的競爭關係以及對於無線通訊環境的不確定性)。基於這個模型，我們分析服務提供者如何經由感知無線電基地台利用”預先拍賣“來避免本身效能之下降。假設感知無線電使用者是自私的，我們表現這場拍賣是一個非合作式賽局，每一個感知無線電使用者理性的做出行為來最大化本身的利益，而整個系統的網路容量也達到最大化，同時服務提供者更可以動態調整得自感知無線電網路之額外利益收入。

The cognitive radio(CR) is proposed to utilize the unused bandwidth of primary radio network(PRN) where the authorized users(called primary users, PU) have the priority to transmit. Once when the cognitive users are allowed to transmit their traffic, it is inevitable to interfere the PU due to the collision. Though many channel detection algorithms have been proposed, however, the phenomena can not be eliminated. The incentive policies are usually to deploy so as to compensate the PU but to increase the difficulty of resource management. In order to realize the CR concept, the Cognitive Radio Base Stations(CR-BS), classified as the PU, have usually been assigned to manage the channels which allow CR users access the network. In this thesis we study the access control policy with variety performance indicators such as capacity and revenue under the QoS constraint. The auction algorithms are adopted to analyze the performance. In our study, the system is composed by regular PU, CR-BS, and CRs. The once when CR-BS detect that there are free channels, it inform the CR about the opportunity of access into the free channel. This auction-based control policy can be used to determine the permission of rejection of access requests from CR by control policy, capacity or revenue. The channel characteristics such as noise, interference, and threshold of receiving power intensity are also taken into consideration. Since the auction mechanism follows the concepts described in non-cooperative game theory, the primary assumptions for user behaviors such as the selfish and seeking the minimum cost and maximum profit or capacity are derived in the analysis. Variety performance indicators are used to evaluate those control algorithms under these assumptions. The main contributions for our study are to study those performance indicators including channel capacity and revenue with different access control algorithms by the using of auction theory. The optimization between the system capacity and revenue is also addressed in this paper.