Adaptive kernel methods for CDMA systems

This paper discusses a new adaptive learning approach for code division multiple access (CDMA) systems. The author extends the previous work of Gong et al. (1999) where they applied support vector machines (SVM) for CDMA signal recovery using a modified version of SVM based on a mean squared error criterion called least squares SVM. The least squares SVM solution is found by solving a set of linear equations. An advantage of this formulation is that the algorithm can be implemented adaptively online. The least squares SVM solutions are compared via simulations to other conventional CDMA receivers and found to have comparable performance to standard SVM solutions. The least squares SVM are promising as they offer simple methods of realizing nonlinear receivers, can be implemented adaptively, and can work in time-varying environments that are typical for wireless communications.

[1]  Bernhard E. Boser,et al.  A training algorithm for optimal margin classifiers , 1992, COLT '92.

[2]  M. Honig,et al.  Adaptive techniques for multiuser CDMA receivers , 2000, IEEE Signal Processing Magazine.

[3]  J. Platt.  Sequential Minimal Optimization : A Fast Algorithm for Training Support Vector Machines , 1998 .

[4]  Vladimir Vapnik,et al.  Statistical learning theory , 1998 .

[5]  Johan A. K. Suykens,et al.  Sparse approximation using least squares support vector machines , 2000, 2000 IEEE International Symposium on Circuits and Systems. Emerging Technologies for the 21st Century. Proceedings (IEEE Cat No.00CH36353).

[6]  Tomaso A. Poggio,et al.  A Sparse Representation for Function Approximation , 1998, Neural Computation.

[7]  Sergio Verdu,et al.  Multiuser Detection , 1998 .

[8]  Dimitri P. Bertsekas,et al.  Nonlinear Programming , 1997 .

[9]  Johan A. K. Suykens,et al.  Least squares support vector machine classifiers: a large scale algorithm , 1999 .