Suppression of narrow-band interference in CDMA cellular radio telephone using higher order statistics

In this paper, the performance of a Code Division Multiple Access (CDMA) ce llular communication system in the presence of narrowband interference is i nvestigated when an interference canceller is employed. The interference ca nceller consists of an adaptive transversal filter whose coefficients are d etermined by using linear prediction and the least mean kurtosis (LMK) algo rithm, which is an adaptive algorithm based on higher order statistics (HOS ). The standard LMK algorithm is very responsive to large values of prediction errors; it may quickly become unstable unless a very small adaptation gain parameter is employed. In this study, we first show that the LMK algorithm can simply be viewed as a variable step size least-mean square (LMS) algor ithm where the step size adjustment is controlled by the square of predicti on error and its variance. Second, we provide a modification to the LMK alg orithm to ensure that the mean-square error of the algorithm remains bounde d. Simulation results are presented to demonstrate the effectiveness of the use of such an adaptive filter in mitigating narrowband interference in di rect sequence spread spectrum (DS-SS) systems. Speed of convergence, signal -to-noise ratio improvement (SNRI), and error rate performance of a receive r that employs the modified LMK algorithm are examined. In addition, the re sults obtained by the modified LMK algorithm are compared with the results obtained by the commonly-used LMS algorithm.