Adaptive kernel methods for CDMA systems

This paper discusses a new adaptive learning approach for Code Division Multiple Access (CDMA) systems. The last severaly ears has seen the development of powerful new adaptive learning methods for pattern classification and signal processing applications. We extend previous work where weapplied Support VectorMachines (SVM) for CDMA signal recovery to using a modified version of SVM which uses 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 on-line. The Least Squares SVM solutions are compared via simulations to other conv en tional 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.