Blind Constant Modulus Multiuser Detection With Regularization

被引:1
|
作者
Liao, Bin [1 ]
Guo, Chongtao [1 ]
机构
[1] Shenzhen Univ, Coll Elect & Informat Engn, Guangdong Key Lab Intelligent Informat Proc, Shenzhen 518060, Peoples R China
基金
中国国家自然科学基金;
关键词
Detectors; Dispersion; Newton method; Optimization; Signal to noise ratio; Multiuser detection; Symmetric matrices; constant modulus; regularization; semidefinite relaxation; Newton's method;
D O I
10.1109/LWC.2022.3170321
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
The problem of blind constant modulus multiuser detection in code-division multiple-access (CDMA) systems is investigated in this letter. Different from traditional approaches which either assume the amplitude of the user of interest is known or arbitrarily set the value of dispersion constant, we consider a more general case where the amplitude is unknown and the dispersion constant is deemed as an unknown parameter jointly determined with the weight vector of the detector. Moreover, a regularization term of the weight vector is incorporated into the optimization to stabilize the problem. With these considerations, an iterative framework is thus introduced to tackle the formulated problem. It is shown that for a given weight vector, the dispersion constant can be analytically determined. To handle the more challenging subproblem with respect to the weight vector, we first present a semidefinite relaxation (SDR) based scheme. Then, a Newton's method based approach is developed to release the computational burden. Simulations have been carried out to demonstrate the effectiveness of the proposed methods and superiority over conventional approaches.
引用
收藏
页码:1649 / 1653
页数:5
相关论文
共 50 条