利用LLR提高链路自适应准确性的满分集度网络编码策略

Link Adaption Accuracy Enhancement of Network Coding Strategy with Full Diversity Order due to LLR

错误传播严重影响无线网络编码系统的分集阶数。针对此问题,本文设计了一种基于对数似然比的链路自适应智能网络编码策略,利用似然比代替背景噪声的统计特征参数,以提高中继的适应准确性。在等效信源-中继-信道虚拟模型下,利用接收信号的似然比计算信道后验条件误比特率,在此基础上换算得到等效接收信噪比,以此为依据对中继功率进行动态调整。通过理论分析证明了本方案能够获得满分集度。仿真结果表明,本方案能有效避免错误传播,保证系统获得满分集阶数,且性能优于现有链路自适应方案。

Error propagation seriously degenerate the diversity order of wireless network coding（NC） system.To address this problem,a novel link adaptive smart NC algorithm is proposed at relay based on log likelihood ratio（LLR）.The metric of LLR is employed to replace the statistical characteristics of background noise firstly,so as to promote channel adaption accuracy.Under the virtual source-relay-destination channel model,the posteriori conditional bit error probability is calculated according to the LLR of the received signals.Based on it,the equivalent signal-noise ratio is employed to operate dynamic power scaling by relay.It is theoretically proved that the full diversity order can be attained by this scheme.Simulation results show that the proposed algorithm can maintain full diversity order by mitigating error propagation effectively,and outperform the conventional one.