OFDM及认知OFDM系统关键技术研究

发布时间：2018-06-13 17:18

本文选题：正交频分复用 + 认知无线网　；参考：《西安电子科技大学》2014年博士论文

【摘要】：随着新的高速无线应用的不断涌现,现有无线业务的不断发展,带宽的需求量在迅速增加,频谱资源匮乏问题日益严重,然而,美国联邦通信委员会(FCC)实测数据表明,频谱资源匮乏的真实原因来源于频谱利用率低下。为了解决频谱资源短缺,提高频谱利用率,认知无线电技术应运而生。在认知无线网络中,认知用户在频谱感知的基础上,机会式地接入授权用户的空闲频带进行通信,在避免对授权用户产生有害干扰的同时,应用动态资源分配技术,大大提高了频谱利用率。作为一种多载波调制技术,正交频分复用(OFDM)因其频谱利用率高、抗多径能力强等特点而被广泛应用于宽带无线通信。OFDM灵活的频谱成形技术使得它不需要连续的频带即可进行数据传输,有助于高效利用授权用户的空闲频谱,这种频谱不连续的OFDM调制技术称之为非连续OFDM(NC-OFDM),成为了认知无线网络物理层技术的首选。然而,自OFDM被提出至今,其固有的峰均比高的问题仍未得到完美的解决,同时,在非认知OFDM系统中,考虑用户公平性的资源分配算法的研究仍显得不足。在频谱资源匮乏的今天,认知OFDM无线网络的提出,具有重要的意义,然而,要实现认知系统和授权系统的真正共存,必须有效控制认知OFDM系统对授权系统的带外干扰,同时,要实现频谱利用率的有效提高,必须研究行之有效的认知OFDM系统的动态资源分配算法。针对非认知OFDM系统及认知OFDM系统存在的问题,本文将从非认知OFDM系统的峰均比、认知OFDM系统的带外抑制,以及非认知OFDM系统及认知OFDM系统的资源分配算法等几方面进行研究,具体的研究工作和创新性成果如下:1.针对OFDM信号高峰均比(PAPR)问题,提出了一种基于门限的分段压扩算法,以较小的误比特率(BER)损失,获得了有效的峰均比抑制。该算法根据OFDM信号幅度的统计分布,将OFDM信号分成了三类:小(幅度)信号、平均(幅度)信号和大(幅度)信号,通过选择合适的门限,对大信号进行压缩以减小峰值,同时放大小信号,起到提升平均功率的作用,从而达到峰均比的降低。在接收端,通过迭代检测算法估计并消除压扩噪声,有效减小了分段压扩造成的带内失真,减小了系统的BER性能损失。另外,NC-OFDM信号频谱的不连续性使得其峰均比的统计特性有别于传统OFDM信号,文中还研究了NC-OFDM信号峰均比的统计分布,为进一步研究NC-OFDM信号峰均比抑制算法提供了帮助。2.针对多用户非认知OFDM系统下行链路的动态资源分配问题,提出了一种考虑用户优先级的自适应子载波分配算法。首先,以两个用户的子载波分配为例分析了用户选择子载波的先后顺序对系统容量的影响,在此基础上,提出了一种新的冲突解决算法。在该算法中,每个用户独立地在所有子载波中搜寻最佳信道增益的子载波,当出现冲突,即多个用户同时选择一个子载波时,由“平均信道增益”的大小决定用户选择冲突子载波的优先级。文中定义了两种不同的优先级准则,一种定义平均信道增益较高的用户优先级较高,另一种则定义平均信道增益较低的用户优先级较高,相应地,给出了两种不同的冲突解决方案。仿真结果表明,由平均信道增益较低的用户优先选择冲突子载波,系统容量可以近似达到理论上限值。3.针对OFDM信号谐波功率较大的问题,在认知OFDM无线网络中,提出了一种基于星座扩展的带外抑制(谐波抑制)算法。在传统的星座扩展算法中,由于原始数据符号独立地进行星座扩展,很难实现对相邻扩展星座点的控制,一旦相邻扩展星座点出现实部或者虚部符号相同的情况,将对谐波抑制非常不利。与传统星座扩展算法不同的是,该算法的星座扩展基于符号对,而不是单个符号,在该算法中,首先将相邻的原始数据符号两两组成一组,然后以组为单位来进行星座扩展,控制扩展星座点反相或者接近反相来实现带外辐射的抑制。在接收端,充分利用了发射端的星座扩展表来纠正判决过程中可能出现的错误,减小了由于星座扩展带来的BER性能损失,理论和仿真结果均证明了该算法的有效性。4.针对认知OFDM无线网络的资源分配问题,首先,分析了认知用户因带外辐射和非理想的感知信息对授权用户造成的干扰,并在资源分配算法的干扰约束中进行了综合考虑;其次,考虑到授权用户的信道占用在评价资源分配算法的重要地位,介绍了确定性和随机性两种授权用户的信道占用模型,在此基础上,对认知OFDM无线网络的资源分配问题进行建模。在本文所研究的认知无线网络中,认知用户分别支持实时(RT)业务(延迟敏感业务),和非实时(NRT)业务(延迟容忍业务),相应的认知用户分别为迟延敏感用户(DS-CUs)和迟延容忍用户(DT-CUs)。由于延迟容忍用户关心吞吐量的大小,而延迟敏感用户则更关心切换时延的大小,将所研究的认知无线网络资源分配问题建模为一个双层规划问题,在该双层规划中,上层优化目标是在满足延迟敏感用户最小速率需求的基础上最小化其切换时延,下层优化目标是延迟容忍用户的吞吐量。由于双层规划问题求解的复杂度较高,我们将该最优化问题分解为子信道分配和功率增强两个问题的求解,仿真结果证明了该算法在所研究场景中的有效性。
[Abstract]:With the continuous emergence of new high-speed wireless applications and the continuous development of the existing wireless services, the demand for bandwidth is increasing rapidly, and the shortage of spectrum resources is becoming more and more serious. However, the actual data from the Federal Communications Commission (FCC) of the United States shows that the real source of the lack of spectrum resources is derived from the low utilization of spectrum. In the cognitive wireless network, cognitive users have access to the free frequency band of authorized users on the basis of spectrum sensing in cognitive wireless networks. In order to avoid harmful interference to authorized users, the cognitive users should use dynamic resource allocation technology to improve the spectrum utilization greatly. As a multi carrier modulation technology, orthogonal frequency division multiplexing (OFDM) is widely used in broadband wireless communication.OFDM flexible spectrum forming technology because of its high spectrum utilization and strong anti multipath ability. It makes it not need continuous frequency band to carry out data transmission, and is helpful to efficiently use the free spectrum of authorized users, this frequency The discontinuous OFDM modulation technology is called discontinuous OFDM (NC-OFDM), which has become the first choice in the physical layer technology of cognitive wireless network. However, since OFDM has been proposed, its inherent peak to average high problem has not been solved perfectly. At the same time, in the non cognitive OFDM system, the research of resource allocation algorithm considering the fairness of the user still appears. In the lack of spectrum resources today, it is of great significance to recognize the OFDM wireless network. However, in order to realize the real coexistence of the cognitive system and the authorization system, it is necessary to effectively control the external interference of the cognitive OFDM system to the authorization system. At the same time, to realize the effective enhancement of the frequency utilization rate of the spectrum, the effective cognitive OFD must be studied. The dynamic resource allocation algorithm of M system, in view of the problems of non cognitive OFDM system and cognitive OFDM system, this paper will study the peak to average ratio of non cognitive OFDM system, out of band suppression of cognitive OFDM system, and the resource allocation algorithm of non cognitive OFDM system and cognitive OFDM system, and the specific research work and innovation will be made. The results are as follows: 1. in view of the peak to average ratio (PAPR) problem of OFDM signal, a threshold based segmented compression expansion algorithm is proposed. The effective peak to mean ratio suppression is obtained with smaller bit error rate (BER) loss. The algorithm divides the OFDM signal into three categories: small (amplitude) signal, average (amplitude) signal and large (amplitude) (amplitude) signal based on the statistical distribution of the amplitude of OFDM signal. ) the signal, by selecting the appropriate threshold to compress the large signal to reduce the peak value, and at the same time, play the role of increasing the average power, and thus achieve the decrease of the peak to average ratio. At the receiving end, the iterative detection algorithm is used to estimate and eliminate the pressure expansion noise, effectively reducing the intra band distortion caused by subsection compression and reducing the BER of the system. In addition, the discontinuity of the NC-OFDM signal spectrum makes the statistical characteristics of the peak to average ratio different from the traditional OFDM signal. The statistical distribution of the peak to average ratio of the NC-OFDM signal is also studied in this paper, which provides the dynamic capital for the further study of the NC-OFDM signal peak mean ratio suppression algorithm for the multiuser and non cognitive OFDM system downlink. An adaptive subcarrier allocation algorithm considering user priority is proposed. First, the effect of the sequence of user selection subcarriers on the system capacity is analyzed with two users' subcarrier allocation. On this basis, a new conflict resolution algorithm is proposed. In this algorithm, each user is independently in the algorithm. All subcarriers search for the best channel gain subcarrier. When there is a conflict, when multiple users select a subcarrier at the same time, the size of the "average channel gain" determines the priority of the user's choice of the conflict subcarrier. Two different priority criteria are defined in this paper, and a user priority with higher average channel gain is defined. The higher, the other is the higher priority of the user with lower average channel gain. Accordingly, two different conflict solutions are given. The simulation results show that the user with lower average channel gain preferred the conflict subcarrier, the system capacity can approximate to the theory upper limit value.3. for the higher harmonic power of the OFDM signal. In the cognitive OFDM wireless network, an extra band suppression (harmonic suppression) algorithm based on the constellation extension is proposed. In the traditional constellation extension algorithm, it is difficult to control the adjacent extension constellations because of the independent constellation expansion of the original data symbols. Once the adjacent extension constellations appear in the real or imaginary parts of the constellation, the same sign is the same. Different from the traditional constellation extension algorithm, the constellation extension of the algorithm is based on the symbol pair rather than the single symbol. In this algorithm, the adjacent original data symbol 22 is first formed, then the group is used to expand the constellation, control the inverse of the extension constellation or close to the inverse. At the receiving end, we make full use of the constellation extension table of the transmitter to correct the possible errors in the decision process and reduce the BER performance loss due to the constellation extension. Both the theoretical and simulation results prove the effectiveness of the algorithm for the allocation of the resources of the cognitive OFDM wireless network, first of all,.4.. First, it analyzes the interference caused by the cognitive users to the authorized users because of the out of band radiation and the non ideal perceptual information, and takes the comprehensive consideration in the interference constraints of the resource allocation algorithm. Secondly, considering the important status of the authorized user's channel occupancy in the evaluation of resource allocation algorithms, two kinds of authorized users are introduced. On this basis, we model the resource allocation problem of cognitive OFDM wireless networks. In this cognitive wireless network, cognitive users support real-time (RT) services (delayed sensitive services), and non real-time (NRT) services (delayed tolerant services), and corresponding cognitive users are delayed sensitive users (DS-CUs), respectively. And delay tolerant users (DT-CUs). Because the delay tolerant users are concerned about the size of the throughput, the delay sensitive users are more concerned with the size of the handover delay. The study of the cognitive wireless network resource allocation problem is modeled as a double layer programming problem. In the two-layer programming, the upper level optimization goal is to satisfy the minimum speed of the delay sensitive user. On the basis of the rate requirement, the switching delay is minimized. The lower layer optimization target is the throughput of the delay tolerant user. Because of the high complexity of the two-layer programming problem, we decompose the optimization problem into two problems of subchannel allocation and power enhancement. The simulation results prove the effectiveness of the algorithm in the research scene.
【学位授予单位】：西安电子科技大学
【学位级别】：博士
【学位授予年份】：2014
【分类号】：TN925

【相似文献】