Blind a Daptive Channel Shortening with a Generalized Lag-Hopping Algorithm Which Employs Squared Auto-Correlation Minimization [GLHSAM]

Abstract

A generalized blind lag-hopping adaptive channel shortening (GLHSAM) algorithm based upon squared auto-correlation minimization is proposed. This algorithm provides the ability to select a level of complexity at each iteration between the sum-squared auto-correlation minimization (SAM) algorithm due to Martin and Johnson and the single lag auto-correlation minimization (SLAM) algorithm proposed by Nawaz and Chambers whilst guaranteeing convergence to high signal to interference ratio (SIR). At each iteration a number of unique lags are chosen randomly from the available range so that on the average GLHSAM has the same cost as the SAM algorithm. The performance of the proposed GLHSAM algorithm is confirmed through simulation studies.