Performance Characteristics of Adaptive Complex Linear Prediction Filter

The use of adaptive complex linear prediction filters (ACLPF) is examined as a means of rejecting narrowband interference in a communication system. The performance characteristics of the ACLPFs using various modified least mean square (LMS) algorithms are quantified in a form useful for design applications. The results indicate that the modified algorithms offer extra design freedom in obtaining desired system performance goals. The additional design freedom, in many cases, results in a better overall system design which eliminates implementation difficulties and provides flexibility in selecting design parameters. Numerical results are presented to illustrate the design concept introduced. Computer simulation results are also given corroborating the analytical conclusions. The results obtained can be used to evaluate tradeoffs amongst different algorithm approaches for improving the system performance while maintaining the system implementation practicality.