On the Use of RLS with Covariance Reset in Tracking Scenarios with Discontinuities

Low-complexity algorithms used for adaptive filtering applications have been shown to be ineffective in rapidly changing environments. This work investigates the use of the recursive least-squares algorithm with the addition of covariance resetting (RLS+R) to limit the memory of the algorithm and allow it to react to more rapid variations in the environment, including the sudden appearance of an interferer. A comparison is made of the performance of the least-mean-square (LMS) algorithm with RLS+CR in rural and urban environments with and without interference using measured channel data. The RLS+CR algorithm is shown to outperform the LMS algorithm in urban environments as well as environments with high-powered interference