Efficient fast blind equalization with two‐stage single/multilevel modulus and DD algorithm for 64/256/1024qam wired cable communications

Abstract In this paper, an efficient blind equalization with two‐stage single/multilevel modulus and decision directed (DD) least mean square (LMS)‐like algorithm is proposed for high‐order QAM (64/256/1024QAM) cable modem systems. In wire‐lined cable systems, it seems that the cable channel may display a nearly invariant multi‐path effect over a considerable time, and then the blind equalizer can have the capability to recover the received high‐order QAM symbols on wired cables. The proposed blind equalization algorithm applies a two‐stage convergence scheme, which results from the multi‐modulus algorithm (MMA), sometimes called the modified constant modulus algorithm (MCMA), and the DD LMS‐like method. In the first equalization stage, the joint MMA and DD LMS‐like methods are applied for the purpose of fast convergence. When the convergence process reaches an acceptably steady condition, the convergence detector changes the process into the second stage. In the second equalization stage, the enhanced DD LMS‐like method with the generalized multi‐modulus is applied to reduce the mean square error (MSE) of equalizations further. For high‐order QAM cable systems, the proposed method performs faster than the previous well‐known blind methods do. Simultaneously, the proposed algorithm also provides less MSE and lower symbol error rate (SER) than the other blind methods do at the same signal‐to‐noise ratio (SNR).