认知无线电动态频谱接入的竞争与干扰规避策略研究

发布时间：2018-05-18 07:32

本文选题：认知无线电 + 动态频谱接入　；参考：《哈尔滨工业大学》2014年博士论文

【摘要】：认知无线电概念是现代通信发展历程上的一个革命性观点，在过去十年中一直受到广泛的关注。通过软件定义无线电，认知无线电可以察觉信道状态、能源效率、带宽可用性、地理位置、频谱使用、网络连接性等等。并且，通过与无线环境和其他用户的交互作用，认知无线电可以自适应地改变其操作参数从而达到最佳的频谱使用率。此外，认知无线电概念还被引入到下一代无线和移动计算系统中用来提高频谱的使用效率。动态频谱接入是认知无线电的关键技术，采用该技术认知无线电用户可以伺机接入无线电频谱。认知用户在接入无线电频谱过程对其他用户产生的干扰是一个值得研究的问题。该问题可以从三个方面进行研究。一、授权用户可以制定规则来指导认知用户的频谱接入，减少来自认知用户的干扰；二、认知用户通过频谱感知避免对授权用户或其他认知用户造成干扰；三、认知用户之间通过频谱竞争或合作来共享频谱资源。本文从上述三个方面研究了认知用户频谱竞争与干扰避免的问题，并根据博弈理论和优化算法，提出了解决的方案。首先，从授权用户角度出发，针对认知用户与授权用户频谱共享中的干扰问题，提出基于干扰补偿定价的认知周期优化算法和基于定价机制的混合频谱接入管理算法。针对认知用户在授权用户随机出现情况下无法及时退出所导致的干扰问题，传统的认知周期优化算法以最大化认知用户的平均吞吐量为目标，忽略授权用户的体验。基于干扰补偿定价的认知周期优化算法使授权用户参与制定认知周期，根据授权用户的频谱利用率调整认知周期时长，并且平衡认知用户的干扰概率和信道空闲率。基于定价机制的混合频谱接入管理算法针对认知用户混合频谱接入时频谱接入方式对授权用户造成的不同影响。该算法可以根据授权用户频谱利用率调整认知用户选择频谱Overlay模式和选择频谱Underlay模式的频谱接入价格，通过频谱接入价格影响认知用户对这两种频谱接入方式的选择。其次，从认知用户角度出发，针对认知用户对授权用户的干扰以及认知用户之间的干扰问题，提出了时间加权频谱感知优化算法，建立频谱感知博弈模型并以此为基础提出非对称碰撞规避策略。时间加权频谱感知优化算法针对授权用户在认知用户频谱感知环节突然回归当前信道的现象，通过对采样信号分组并按照时间顺序进行线性加权来提高对授权用户的检测概率。算法分析对比了数据融合准则和决策融合准则下的时间加权优化算法。结果表明，采用决策融合时间加权优化算法，在相同分组数量的情况下的检测性能较好。同时针对授权用户随机出现的场景，时间加权优化算法会明显的提高检测概率。频谱感知博弈模型模拟了认知用户吞吐量竞争的自私行为以及干扰。基于频谱感知博弈模型，本文提出非对称碰撞规避策略。该策略在保证自私用户的minmax收益基础上，最大化非自私用户的频谱接入概率，使得非自私用户与自私用户可以共享信道。最后，针对独立的认知用户之间频谱共享时的干扰问题，提出了竞争退出策略和非对称主动合作策略。在独立的认知用户频谱共享模型中，两个认知用户频谱竞争的均衡结果是认知用户将各自的功率扩展到整个频谱。针对均衡时认知用户之间的相互干扰问题，本文提出了基于频谱移动功能的竞争退出策略。该策略以最大化认知用户的长期收益为目标，为不同的感知结果指定了一个竞争退出时间。通过与“干扰-立即退出”方案相对比，基于竞争退出策略均衡时的方案，认知用户明显的减少了信道切换次数，有利于提高网络稳定性。此外，，本文提出了基于频谱接入功能的非对称主动合作策略。认知用户以最大化长期收益问目标，通过认知用户可以单方面采取合作行为，另一个认知用户响应合作行为的方式来建立合作。非对称主动合作是一种隐式的共谋策略，为彼此独立的认知用户提供了一种优化频谱共享策略。
[Abstract]:Cognitive radio (cognitive radio), a revolutionary concept in the development of modern communications, has attracted wide attention over the past ten years. Through software defined radio, cognitive radio can detect channel state, energy efficiency, bandwidth availability, geographic location, spectrum use, network connectivity and so on. And, through wireless environment and wireless environment, The interaction of other users, cognitive radio can adaptively change its operating parameters to achieve the best frequency utilization. In addition, the concept of cognitive radio is also introduced to the next generation wireless and mobile computing system to improve the efficiency of spectrum.
Dynamic spectrum access is the key technology of cognitive radio. This technology is used to recognize radio users to access radio spectrum. Cognitive users' interference with other users in the radio spectrum process is a problem worthy of study. This problem can be studied from three sides. First, authorized users can make rules. To guide the spectrum access of cognitive users and reduce interference from cognitive users; two, cognitive users avoid interference with authorized users or other cognitive users through spectrum sensing; three, cognitive users share spectrum resources through spectrum competition or cooperation. This paper studies cognitive user spectrum competition from the above three aspects. And the problem of interference avoidance. Based on game theory and optimization algorithm, a solution is proposed.
First, from the point of view of authorized users, this paper proposes a cognitive cycle optimization algorithm based on interference compensation pricing and a hybrid spectrum access management algorithm based on pricing mechanism, aiming at the interference problem in the spectrum sharing between the cognitive users and the authorized users. The traditional cognitive cycle optimization algorithm aims at maximizing the average throughput of the cognitive users and neglects the experience of the authorized users. The cognitive cycle optimization algorithm based on the interference compensation pricing makes the authorized users participate in the cognitive cycle, adjusts the cognitive cycle according to the spectrum utilization rate of the authorized users and balances the cognitive users. The interference probability and the channel idle rate. The hybrid spectrum access management algorithm based on the pricing mechanism has different effects on the authorized users when the cognitive users are mixed with the spectrum access. This algorithm can adjust the frequency spectrum Overlay mode and the spectrum Underlay mode of the cognitive users according to the spectrum utilization of the authorized user. The price of spectrum access affects users' choice of these two spectrum access modes through the spectrum access price.
Secondly, from the perspective of cognitive users, this paper proposes a time weighted spectrum sensing optimization algorithm, based on the interference of cognitive users to authorized users and the interference between cognitive users, and establishes a spectrum sensing game model and proposes an asymmetric collision avoidance strategy based on it. Time weighted spectrum sensing optimization algorithm is aimed at authorized users. The phenomenon of the sudden return of the spectrum perception link to the current channel, the detection probability of the authorized users is improved by grouping the sampled signals and weighted to the time sequence. The algorithm analysis compares the data fusion criterion and the time weighted optimization algorithm under the decision fusion criterion. The results show that the decision fusion is adopted. The inter weighted optimization algorithm has better detection performance in the case of the same number of groups. At the same time, the time weighted optimization algorithm will obviously improve the detection probability for the random scene of the authorized users. The spectrum sensing game model simulates the selfish behavior of the cognitive user throughput competition and the interference. In this paper, an asymmetric collision avoidance strategy is proposed, which maximizes the spectrum access probability of unselfish users on the basis of guaranteeing the minmax revenue of selfish users, making the unselfish users share the channel with the selfish users.
Finally, a competitive exit strategy and an asymmetric active cooperative strategy are proposed for the interference problem of spectrum sharing among independent cognitive users. In an independent cognitive user spectrum sharing model, the equilibrium results of the two cognitive users' spectrum competition are that the cognitive users extend their respective power to the entire spectrum. In this paper, a competitive exit strategy based on spectrum mobility is proposed in this paper. This strategy aims at maximizing the long-term income of cognitive users, and specifies a competitive exit time for different perceived results. By comparing with the "interference immediate exit" scheme, the strategy is based on the equilibrium of the competitive exit strategy. In the case, cognitive users obviously reduce the number of channel switching and improve the network stability. In addition, this paper proposes an asymmetric active cooperative strategy based on the spectrum access function. The cognitive users can maximize the long-term returns to the goal, and the cognitive users can take cooperative behavior unilaterally, and another cognitive user responds to the cooperative behavior. In order to establish cooperation, asymmetric active cooperation is an implicit collusion strategy, which provides an optimized spectrum sharing strategy for independent cognitive users.
【学位授予单位】：哈尔滨工业大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TN925

【参考文献】