基于Triple Play业务的OFDM系统信道划分  

Optimization Algorithm of Channel Partition in OFDM System Based on Triple Play

作　　者：岳欣[1] 李一兵[2] 杨莘元[2] 

机构地区：[1]北京邮电大学大系统实验室,北京100876 [2]哈尔滨工程大学信息与通信工程学院,黑龙江哈尔滨150001

出　　处：《微电子学与计算机》2007年第3期117-119,共3页Microelectronics & Computer

基　　金：第四代移动通信关键技术研究(GZ04A127)

摘　　要：研究了基于Triple play业务的OFDM系统信道划分问题。首先证明了在信道状态和传输总量已知的情况下,基于传输时差最小的OFDM系统信道划分的优化问题属于NP完备问题,因此必须使用随机算法代替穷举法来求解。引入了遗传算法和蚂蚁算法,并在原有方法的基础上进行了改进,在大量仿真实验的基础上,给出了仿真结果,并证明尽管随机算法不能保证解的最优性,但搜索时间较短;同时在相同条件下,蚂蚁算法可以显示出比遗传算法更强的搜索能力,因此对于OFDM这类实时性要求较高的系统,具有更大的实用性。Triple play has been one of the focuses of broadband operation. This paper paid more attention to research optimization algorithm of channel partition in OFDM system based on triple play, It was proved that this problem belonged to NP complete problem, and it needed to be solved by random method. Genetic algorithm and ant algorithm （AA） optimization were shown in this paper. In order to solve the problem of channel partition, a modified method was provided in this paper. Through simulation we can find AA is better than GA when they are finding the best result under the same situation. So AA could be widely used in the future.

关 键 词：OFDM系统 遗传算法 蚂蚁 最优化理论 三重播放业务 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]