项目名称: 联合数据区分编码理论的有记忆物理层网络编码方法
项目编号: No.61201148
项目类型: 青年科学基金项目
立项/批准年度: 2013
项目学科: 电子学与信息系统
项目作者: 于启月
作者单位: 哈尔滨工业大学
项目金额: 25万元
中文摘要: 物理层网络编码由于对同时接入的用户进行信号处理,极大提高单位时间系统吞吐量并降低系统时延,因此得到了越来越广泛的关注。而物理层网络编码的核心思想,则是如何将同时接入的多用户数据加以区分并实现物理层数据映射。目前常用的数据区分编码,具有各自的局限性:比如基于扩频码的CDMA系统或采用空时编码的高维传输系统,其符号携带平均信息量较低;比如采用"预编码"的多用户MIMO技术,对用户的信道信息和信道特性要求较高、运算复杂度较大。据此,本课题根据代数几何、组合数学和近世代数的理论,在伽罗华域针对不同用户的数据流通过构造可分辨的欧氏距离,提出用户数据流区分的编码和译码原理,以期在较低的复杂度下得以有效提高多址接收端每符号携带的平均信息量;并将该编码理论应用在物理层载波调制的携带信息位,首次提出有记忆物理层网络编码,在频谱带宽等资源受限的情况下,有效提高系统容量、频谱利用效率。
中文关键词: 物理层网络编码;唯一可译码;和速率;频率选择性衰落信道;功率分配
英文摘要: Recently, Physical-layer Network Coding (PNC) is very attractive, since it is able to increase the throughput and improve the time-delay significantly. The basic thought of PNC is to coordinate transmissions among users, in contrast to "straightforward" network coding wich performs coding arithmetic on digital bit streams after they have been received, PNC makes use of the additive nature of simultaneously arriving electromagnetic (EM) waves for equivalent coding operation. Thereby, how to separate the hybrid multiple-sources information and then do coding is the core topic of PNC. The most popular methods to separate the hybrid multiple-sources information are code division multiple access (CDMA), space-time block coding (STBC), and multi-user multiple-input multiple-output (MIMO). However, the mentioned methods have some problems. For example, spreading signals always oppupy a wider frequency-domain bandwidth, while 4 or more antennas STBC will decrease the information rate. The channel state information (CSI) is necessary for multi-user MIMO especially the orthogonal or semi-orthogonal channel are necessary; otherwise, it is impossible to realize space division. However, it is known that it is difficult to find perfect orthogonal channels, and the multi-users channel estimation is also very complexity. In thi
英文关键词: Physical-layer network coding (PNC);uniquely-decodable coding;sum rate (channel capacity);frequency-selective fading channel;power allocation