【作者】 张立；

【导师】 郑国莘；

【作者基本信息】 上海大学 ， 通信与信息系统， 2010， 博士

【摘要】 随着无线通信业务需求的快速增长,宝贵而有限的无线电频谱资源变得越来越稀缺。而认知无线电技术主张利用动态频谱的思想,主动感知无线通信环境,根据一定的学习和调度算法,实时自适应地改变系统的工作参数,动态地检测和有效地利用空闲频谱,允许在时间上、频率上以及空间上进行多维的频谱复用,这将大大降低频谱和带宽限制对无线通信技术发展的束缚,从而提高频谱利用率。为了在认知无线电网络多个用户之间实现动态频谱共享,媒体访问控制(Media Access Control, MAC)层关键技术面临诸多挑战,例如认知用户的信道接入机制及控制信令的设计、多个请求接入网络的应用调度、信道间高效的切换技术以及认知无线电网络的安全机制等。本文主要围绕认知无线电网络MAC层的关键问题进行研究,分别从认知用户的接入调度、MAC层控制信令设计及多信道切换的角度出发,对认知用户如何自适应地利用频谱资源同时又满足一定的服务质量(Quality of Service,QoS)参数的问题进行了研究,针对动态ad hoc应用环境提出了相应的解决方案。另外,对认知无线电网络中频谱感知结果的安全认证也进行了理论探讨。本文的主要工作和贡献如下:1.为了充分利用信道机会并高效地进行数据发送,将认知无线电网络中认知用户的可用信道机会进行建模,提出了一个QoS感知的自适应机会信道接入(Adaptive Opportunistic Channel Access,AOCA)调度算法。基于所感知到的信道状态信息,将单信道和多信道情景下认知用户的可用信道机会建模,根据信道接入需求动态地调度物理层的信道感知操作,从而使得用户自适应地最大化信道利用率,同时不增加认知用户的额外开销。所提出的AOCA调度算法的优越性在于它能够自适应调度信道接入并提供QoS支持,所采用的QoS参数包括空闲缓冲区的大小、队列长度以及冲突比。仿真结果表明,该算法具有很好的自适应性和灵活性,并能够取得较好的效率。2.为了在认知无线电网络中灵活高效地利用动态的信道机会,提出了一种基于控制信道预约的认知无线电网络MAC协议。针对现有认知无线电网络MAC协议控制信道利用率低和限制网络规模扩展的缺点,该协议引入了竞争控制信道和预留控制信道。其中竞争控制信道为不同类型的业务预留相应的接入机会。认知用户在接入过程中,首先在竞争控制信道上预约预留控制信道子帧,然后再通过预留控制信道,基于请求带宽预约数据信道,同时用户通过维护信道状态表来分配和回收信道资源。仿真结果表明,该协议可以有效地调度多用户多信道的接入,具有很好的可扩展性。与其它基于固定控制信道的MAC协议相比,大大降低了发生冲突和拥塞的概率,进而降低了端到端时延,并提供了良好的端到端吞吐性能。3.在异类认知无线电网络中,为了保证认知用户在多信道切换过程中满足一定的QoS需求,同时又尽可能地避免隐藏信道和暴露信道对性能的影响,提出了一个灵活的自适应垂直信道切换算法。该算法基于频谱共享的概念,假设存在一个频谱池,认知用户据此动态自适应地调度信道感知,维护一定的备用空闲频谱信息,以便认知用户在必要时进行切换调度。该算法采用信道带宽作为QoS参数来刻画网络的异类性。仿真主要考虑了切换阻塞概率、切换开销和冲突概率这三个评价指标。结果表明在异步认知无线电网络和同步认知无线电网络中,所提出的切换算法都明显地改善了切换期间的服务性能,并提供了较稳定的通信吞吐。4.针对基于无线电环境图(Radio Environment Map,REM)的认知无线电网络的认证技术进行了研究,提出了一种基于混沌序列和信道编码技术的认证协议。在分布式的感知数据汇集过程和决策结果的分发过程中,认证信息被当作秘密数据,并以随机噪声的形式嵌入到信道数据(即经过信道编码的感知数据)中,而由秘密数据的嵌入所造成的错误能够通过纠错编码技术纠正。接收用户能够提取该秘密信息并判断所接收到的数据是否来自合法用户。通过安全分析和讨论证明,所提出的认证方案具有很好的可靠性和安全性。更多还原

【Abstract】 The rapid growth of the ubiquitous wireless services has imposed great pressure on the fixed and limited radio spectrum. Given this fact, the research society has adopted the new concept of cognitive radio (CR) to change the traditional static spectrum allocating and to use spectrum dynamically. Cognitive radio has been proposed as a way to improve spectrum efficiency by detecting the unoccupied spectrum, changing the system operation parameters (e.g., symbol rate, power, bandwidth, latency) in real-time to use radio resources according to a certain learn and schedule algorithm. Therefore, in theory, it can achieve the multi-dimension spectrum utilization including frequency, time and space, which greatly reduces the limitation of spectrum and bandwidth to wireless communication technology. Thus, spectrum utilization is enhanced. However, there are lots of new challenges in order to realize dynamic spectrum sharing among cognitive users in cognitive radio networks. First of all, media access control (MAC) layer faces many difficulties that are not presented in the conventional wireless networks. Especially, how to design a cognitive users accessing mechanism and control signaling, how to schedule several channel access request and how to implement channel handoff in dynamical environment are all open research issues. Additionally, the security mechanism for cognitive radio also has lots of threats.This thesis mainly focuses on the MAC layer key technology of CR networks and performs the theory analysis and experiment research. During the process of access schedule, MAC layer control signaling and multi-channel handoff, how CR user adaptively utilizing spectrum and satisfying some quality of service (QoS) parameters are studied in detail. Several solutions are presented for CR ad hoc application environment. In addition, the authentication security for the spectrum sensing result and membership maintaining of CR networks are also explored. The main contents and contributions are as following:1. In order to fully utilize channel opportunity and efficiently transmit data in cognitive radio networks, this chapter models the available channel opportunity and proposes an adaptive opportunistic channel access (AOCA) schedule algorithm. Based on the channel state information sensed, available channel opportunity is modeled in single channel and multichannel scenario, respectively. The schedule algorithm can adaptively schedule channel sensing leveraging the MAC layer decisions and maximize channel utilization without imposing any additional burden on cognitive users. Different from other CR algorithms, AOCA algorithm provides cognitive users with QoS support. The QoS is identified as terminals’buffer space and queue length along with the collision ratio with other users. The simulation results demonstrate that the proposed channel schedule scheme can utilize the channel adaptively and flexibly in CR networks. Therefore it can achieve good efficiency.2. In order to utilize the dynamic channel opportunity flexibly and efficiently in cognitive radio networks, a multi-channel adaptive MAC protocol is presented. The proposed protocol employs the concept of competing control channel (CCC) and reserving control channel (RCC) to improve the low utilization and lack of expansibility in related works. In order to access the channel, cognitive users reserve the sub-frame of RCC through CCC, which obligates the access opportunity for different traffic. Then they apply the data channel based on the required bandwidth using RCC. The channel is distributed and recycled by updating a channel state table. Simulations show that the proposed protocol can effectively schedule the accessing of users and exhibit good scalability. Compared with other fixed control channel based protocol, it significantly reduces the probability of collision and congestion, thereby improves the end-to-end delay and network throughput.3. For the sake of executing vertical handoff for cognitive user without interfering primary user while offering the desired QoS and eliminating the influence caused by hidden channel and exposed channel simultaneously, this paper proposes a flexible vertical channel handoff algorithm for heterogeneous cognitive radio networks. This algorithm is very flexible and effective. It can adaptively schedule channel sensing by utilizing the concept of spectrum pooling. The key QoS parameter adopted in this algorithm is channel bandwidth which denotes the heterogeneity of the networks. The handoff process is analyzed using Markov model. We evaluate the major performance metrics such as handoff blocking probability, handoff cost and collision probability by simulation. The results indicate the proposed algorithm can significantly improve the performance and provide stable communication in the synchronous and asynchronous CR networks.4. With the intent of efficiently occupying under-utilized spectrum, radio environment map (REM) based CR networking is proposed to facilitate the distributed spectrum sensing. In this chapter, we identify the security threats and propose a chaos sequence and channel encoding based adaptive authentication protocol for REM-based CR networks. In the duration of distributed data aggregating and decision results distributing, authentication information is secret data and sensing information is useful data. Employing channel encoding techniques and chaos sequence, authentication data is embedded into channel data as random noise. The error of host data caused by secret data can be corrected via error correction encoding technique. The receiver can then use the secret information extracted from channel data to authenticate if the received data is sent by the user claimed. To the best of our knowledge, that is the first attempt to provide authentication mechanism for REM-enabled CR network.更多还原