无穷范数松弛的低复杂度超奈奎斯特检测  

Low-complexity Symbol Detection for Faster-than-Nyquist Signaling by Infinity Norm Relaxation

作　　者：来世豪 李明齐[1] LAI Shihao;LI Mingqi(Shanghai Advanced Research Institute,Chinese Academy of Sciences,Shanghai 201210,China;University of Chinese Academy of Sciences,Beijing 100049,China;School of Information Science and Technology,ShanghaiTech University,Shanghai 201210,China)

机构地区：[1]中国科学院上海高等研究院,上海201210 [2]中国科学院大学,北京100049 [3]上海科技大学信息学院,上海201210

出　　处：《电讯技术》2019年第3期249-254,共6页Telecommunication Engineering

基　　金：国家自然科学基金资助项目(6511104204);上海市科委课题(18DZ2203900)

摘　　要：超奈奎斯特(Faster-than-Nyquist,FTN)速率传输可以有效提高频谱效率,但这种非正交传输方式引入的严重码间串扰相应提高了接收端的处理难度。针对该问题,设计了一种基于循环成块传输的低复杂度检测算法。最优检测被建模为无约束的二元二次规划(Boolean Quadratic Program,BQP)问题,为了求解该NP-hard问题,采用无穷范数约束松弛原问题的非凸可行解集,并基于次梯度下降法提出松弛问题的有效优化算法。数值仿真结果表明,所提算法在误比特率(Bit Error Rate,BER)性能上优于频域均衡,且在可接受的性能损失范围内算法执行效率远高于理论最优的最大似然序列估计(Maximum Likelihood Sequence Estimation,MLSE)。As an implementation of non-orthogonal modulation,faster-than-Nyquist(FTN)signaling can increase spectral efficiency by intentional introducing inter-symbol interference(ISI),which incurs extra computational burden for receiver.This paper investigates the low complexity detection method for FTN signaling based on a special circulated block transmission model.First,the optimal detection problem is formulated as a non-convex boolean quadratic program(BQP).Relax technique is then used to get the infinity norm constrained approximate convex problem,which can efficiently solved by the proposed sub-gradient based algorithm.Finally,simulation results show that the proposed scheme can achieve better trade-off between bit error rate(BER)performance and computational complexity than frequency domain equalization(FDE)and maximum likelihood sequence estimation(MLSE).

关 键 词：超奈奎斯特传输 循环块传输 二元二次规划 凸优化松弛 次梯度 

分 类 号：TN911.23[电子电信—通信与信息系统]