ROBUST BLIND CHANNEL SHORTENING IN IMPULSIVE NOISE ENVIRONMENTS (FriAmOR2)

Author(s) : Rab Nawaz (Cardiff University, UK) Jonathon Chambers (Cardiff University, UK)

Abstract : This paper proposes a robust blind channel shortening algorithm for multicarrier modulation systems. The algorithm is based on the sum-absolute autocorrelation minimization (SAAM) of the effective channel outside a window of desired length. The algorithm approaches the maximum shortening SNR (SSNR) solution of melsa et al., al. under Additive White Gaussian Noise (AWGN) conditions, is computationally less expensive, and robust to non-Gaussian impulsive noise environments than a latest reported blind adaptive channel shortening algorithm (SAM). Due to the mostly minimum phase nature of Asymmetric Digital Subscriber Digital Lines (ADSL) channel impulse responses, a ``left'' initialization scheme is suggested for blind adaptive channel shortening equalizers which further enhances the robustness of SAAM to impulsive noise. Further studies have also been undertaken showing that the bit rates achieved by SAM in AWGN conditions can be improved and approach the maximum SSNR solution of melsa96 et al., too.

Menu