【作者】 张然然；

【导师】 刘元安；

【作者基本信息】 北京邮电大学 ， 电磁场与微波技术， 2009， 博士

【摘要】 为了提高无线通信系统的有效性和可靠性,自适应传输技术被广泛引入下一代无线通信系统的研究和设计中。一方面,随着人们对未来无线通信系统的可靠性、可支持移动速度和单位比特性价比等方面不断提出更高要求,研究者不仅要关注自适应调制技术的高效性,更对算法的可靠性和实现复杂度有了更高的追求。另一方面,认知无线电(CR,Cognitive Radio)技术的出现完全打破了传统的频谱授权机制对无线通信系统频谱使用的桎梏,与频谱共享、动态频谱接入(DSA,Dynamic Spectrum Access)等概念紧密地结合在一起。本文在对自适应传输领域的经典理论和算法进行认真的讨论和分析的基础上,对高多普勒频率条件下的MIMO-OFDM系统的自适应调制技术、信道状态信息(CSI,Channel Station Information)不精确条件下的MIMO自适应调制技术以及CR系统中的OFDMA自适应资源分配技术进行了深入研究,从系统的有效性、可靠性和计算复杂度等多个角度对其中的自适应传输算法进行分析,对所涉及的关键问题进行深入讨论,使自适应传输技术能够更好地满足未来无线通信系统的各种应用需求,更适合在实际系统和环境中的应用。论文的主要工作和创新之处在于:首先,对高多普勒频率条件下的MIMO-OFDM系统的自适应调制技术进行了研究。通过分析MIMO-OFDM系统等效信道的特点和规律,提出了基于马尔科夫(Markov)模型的自适应调制算法,并对算法的性能、实现的复杂度做了数值和仿真分析。针对传统自适应调制算法可靠性方面的不足,所提算法在考虑不理想CSI影响的情况下改善MIMO-OFDM系统的性能。一方面,可以显著提高系统有效信息速率和吞吐量,同时在信道多普勒频率较高造成CSI质量恶化的情况下,所提算法仍能够满足系统误比特率(BER,Bit Error Rate)的要求;另一方面,所提算法的实现复杂度明显低于传统的利用信道预测器的自适应调制算法。其次,对不精确CSI条件下的鲁棒MIMO自适应调制技术进行了研究,提出了三种全新的鲁棒自适应调制算法。首先提出了鲁棒贪婪算法。相比传统的贪婪算法,所提出的鲁棒贪婪算法可以在CSI不精确的条件下为MIMO系统提供稳定的BER性能,即便CSI误差较大,所提算法仍然可以满足系统BER要求。然后,针对鲁棒贪婪算法实现复杂度较高的问题,提出了鲁棒贪婪算法的简化算法,极大地降低了算法复杂度。接下来,提出了一种最大化信息传输速率的鲁棒算法,使系统的传输可靠性和频谱效率都显著提高。再次,本文研究了CR系统中的OFDMA自适应资源分配技术。提出了三种新的自适应资源分配算法来优化CR系统下行链路的性能。第一种算法为CR系统下行链路的最优自适应资源分配算法,分析了CR用户对授权用户的干扰并给出了设置干扰限制的方法,在此基础上将功率和干扰受限的最大化系统容量的自适应资源分配问题转化为线性不等式约束的凸优化问题,并借助最优化理论给出了解决方案。第二种算法在总功率和BER受限的条件下实现下行CR系统中最优的子载波、比特和功率分配,可以理解为第一种算法在实际系统中的具体实现方法。第三种算法为最小化发射功率的下行CR系统资源分配算法,在用户信息传输速率和干扰受限的条件下最小化系统发射功率。更多还原

【Abstract】 To improve the transmission efficiency and reliability, adaptive transmission technology has been widely researched and used in the next generation wireless communication systems. On the one hand, with the increasing demands of high reliability, high mobility, and low per-bit cost, researchers has paid more and more attentions to improve the performance of adaptive modulation, especially its reliability and complexity. On the other hand, the emergence of cognitive radio (CR) technology has totally broken the restriction of wireless communications by conventional spectrum authorization mechanism. Henceforth, adaptive transmission has been closely integrated with the new concepts of spectrum sharing, dynamic spectrum access (DSA) and so on.On the basis of international current research work on adaptive transmission technology, adaptive modulation for multiple-input multiple-output/orthogonal frequency division multiplexing (MIMO-OFDM) systems in high Doppler frequency environment, adaptive modulation for MIMO systems with imperfect channel state information (CSI), and orthogonal frequency division multiple access (OFDMA) resource allocation for cognitive radio (CR) systems have been investigated. The adaptive transmission algorithms have been studied on the points of efficiency, reliability, computational complexity. And some key problems have been investigated to satisfy the requirements of future wireless communications and make the adaptive transmission more practical. Main innovations of this thesis are included in the following parts.Firstly, adaptive modulation for MIMO-OFDM systems in high Doppler frequency environment has been investigated. Based on the peculiarity of equivalent channels in MIMO-OFDM systems, an adaptive modulation algorithm with Markov channel model has been proposed. Its performance and computational complexity has been analyzed in theory, and also validated by simulation. To solve the reliability problem of conventional adaptive modulation algorithms, the proposed algorithm have taken CSI imperfection into account. On the one hand, the proposed algorithm has good performance. It has high valid data rate and throughput. When the Doppler frequency is relatively high, it can satisfy the bit error rate (BER) requirement. On the other hand, its computational complexity is much lower than that of the conventional adaptive modulation algorithms using channel predictor.Secondly, robust adaptive modulation for MIMO systems with imperfect CSI has been investigated. Three novel robust adaptive modulation algorithms have been proposed. The first algorithm is robust greedy. Comparing with conventional greedy algorithm, robust greedy algorithm is more reliable. It can satisfy the BER requirement when the channel estimation error is remarkable. Then a novel algorithm with lower complexity has been proposed to make the robust adaptive modulation more practical. Subsequently, a robust adaptive modulation to maximize transmission rate has been proposed, by which the reliability and spectrial efficiency have been improved significantly.Thirdly, OFDMA resource allocation technology for CR has been investigated. Three adaptive resource allocation algorithms for downlink CR systems have been proposed. The first one is optimal resource allocation algorithm. Based on interference analysis, the problem to maximum the downlink capacity under transmission power and inference constraint is constructed as a convex optimization with linear inequality constraint and solved by optimization theory. The second one is a subcarrier, bit and power allocation algorithm under transmit power and BER constraints for downlink CR systems. It can be considered as a way to approach the capacity derived by the first resource allocation in practical systems. The third one is used to minimize the total transmit power of downlink CR systems under the transmission rate and interference constraints.更多还原