《多用户MIMO-OFDM系统中的无线资源管理关键技术研究》由会员分享，可在线阅读，更多相关《多用户MIMO-OFDM系统中的无线资源管理关键技术研究（84页珍藏版）》请在金锄头文库上搜索。
1、上海交通大学硕士学位论文多用户MIMO-OFDM系统中的无线资源管理关键技术研究姓名：杜学松申请学位级别：硕士专业：电子与通信工程指导教师：罗汉文20071201V多用户多用户 MIMO-OFDM 系统中的无线资源管理关键技术研究系统中的无线资源管理关键技术研究摘摘 要要多输入多输出（MIMO，Multi-Input and Multi-Output)技术，正交频分复用（OFDM，Orthogonal Frequency Division Multiplexing) 技术和自适应传输技术的结合，被认为将会是未来无线通信系统发展的主要趋势和基础。一般说来，自适应无线资源管理算法可以更加有效利用系
2、统内有限的资源，从而使得系统性能大大提高。因此，对于 MIMO-OFDM 系统自适应无线资源管理优化算法的研究，就成为了目前国内外学术研究领域内的热点问题之一。本文首先提出了一种 OFDM 系统中自适应子载波和功率分配优化算法。动态子载波分配算法指的是在多用户 MIMO-OFDM 系统中，根据分散在小区中不同位置的各个用户瞬时的信道状况，动态地调整系统内所有子载波在不同用户之间的分配。理论研究表明，这种分配方式相比于传统的静态子载波分配方式，可以最大程度地利用由于用户在系统内随机分散而带来的多用户分集效应，使得系统容量得到很大提高。本文提出的算法还考虑了不同用户数据传输速率比例性要求，可以事先
3、设定不同用户的数据传输速率，更加符合实际系统设计要求。由于最优分配子载波，功率问题是一个非常复杂的多参数、非线性优化算法问题，如果要精确求解其最优解，则将会是一个计算复杂度非常高的问题，很难应用在实时的自适应技术中。因此，本文中研究的各种算法都是降低了计算复杂度之后的次优算法，虽然是次优算法，但是系统性能相比于传统的静态分配算法依然得到了很大的提高。本文接着研究了基于遗传算法的自适应资源管理算法。由于遗传算法特别适合应用于多参数非线性的优化问题，相比于传统的贪婪算法，基于遗传算法的子载波分配方案可以进一步提高系统性能。根据 OFDM 系统无线资源管理优化算法问题的特殊性，在对遗传算法中的编码、
4、交叉、变异研究基础上，本文设计了一种特殊的遗传算法，性能仿真结果表明，遗传算法可以进一步地提高系统容量，这是因为，相比于传统的贪婪算法得到局部最优解，遗传算法得到的是系统优化问题的全局最优解。本文最后研究了 MIMO-OFDM 系统的无线资源管理问题。联合考虑最优化分配子载波和比特加载自适应算法，以及基于注水算法的功率分配来共同提高系统性能。此外，在多天线系统中，又增加了一个空间自由度。因此，本文根VI据 Frobenius 范数理论通过最大化信道矩阵的奇异值来提高系统容量，提出了基于奇异值分解（SVD, Singular Value Decomposition）方法的子载波功率次优分配算法，
5、性能仿真表明这种算法在高信噪比条件下相比于传统的波束成型技术大大提高了系统性能，而且可以充分利用多天线带来的空间分集效应和多用户带来的多用户分集效应。由于研究时间和作者个人能力的限制，本文对于无线资源管理算法的研究还是比较浅显的。本文最后归纳了作者的研究成果，提出了将来可以进一步深入研究的几个方向。关键词：关键词：正交频分复用，多输入多输出，自适应，无线资源管理，遗传算法VIIRESEARCH ON WIRELESS RESOURCEMANAGEMENT IN MULTIUSER MIMO-OFDMSYSTEMABSTRACTMIMO, Multi-Input and Multi-Output
6、, and OFDM, the Orthogonal FrequencyDivision Multiplexing, combined with adaptive transmission method have beenwidely considered as the major trend and basis of the development of the nextgeneration mobile communication system. Generally speaking, the adaptive wirelessresource allocation can effecti
7、vely use the limited resources to enhance systemcapacity. Therefore, the research on adaptive wireless resource allocation algorithmhas been one of the hot topics during recent academic researches all over the world.Firstly, this paper proposed one combined adaptive subcarrier and powerallocation al
8、gorithm for OFDM system. Adaptive subcarrier allocation algorithm inmultiuser MIMO-OFDM system means that we can rely on the instantaneous channelconditions of different users located in different positions within the cell to adjust thesubcarrier allocation among them. Theoretical researches have sh
9、own that comparingto the traditional static allocation method, adaptive allocation algorithm can enhancethe system capacity significantly by exploiting multiuser diversity resulting from therandom locations of different users. The algorithm here also has a proportionalrequirement of different users.
10、 To meet the practical system design requirement, wecan set a reasonable data transmission speed for each user in advance. However, theoptimization problem for adaptive subcarrier and power allocation is a very complex,multiple parameters and non-linear problem. It requires large computation effort
11、tofind the exact optimized solution. Therefore, all the algorithms in this paper are thesuboptimal algorithms based on certain simplifications. But the system throughputcan be improved significantly compared to those traditional static allocationalgorithms.Secondly, this paper studies the adaptive r
12、esource allocation based on geneticalgorithm. Genetic Algorithm is a very effective search technique of finding thesolution for the complex, multiple parameters and non-linear optimization problem.VIIIThe adaptive subcarrier allocation algorithm based on genetic algorithm improvessystem capacity, co
13、mpared to the greedy algorithm. Investigation to the differentstages of genetic algorithm, such as coding, crossover and mutation, has also beendone in this paper to design a unique genetic algorithm for the wireless resourceallocation in OFDM system. Simulation results show that the performance can
14、 beimproved. Genetic algorithm can find the global optimum solution while the greedyalgorithm can just find the local optimum solution to the problem.The third part of this paper focuses on the adaptive wireless resource allocationin MIMO-OFDM system. The adaptive subcarrier allocation, bit loading
15、algorithmand power allocation using water filling algorithm are considered together to improvesystem throughput. As we know, multiple antenna systems add another degree offreedom, spatial freedom. Therefore, according to the Frobenius Norm theory, thesystem performance can be improved by maximizing
16、the singular values of thechannel matrix. Then we propose a new dynamic subcarrier and power allocationscheme based on SVD (Singular Value Decomposition) method. From the simulationresult, we can see that the proposed scheme outperforms traditional beamformingschemes in the range of high SNR (Signal Noise Ratio). The proposed schemeexploits comprehensively both the spatial diversity resulting from multiple antennasand also the multiuser diversity resulting from the multiuser system.Due to the li
