A pipelined VLSI NEXT canceller for premises applications

Abstract

A near-end crosstalk (NEXT) canceller using a fine-grain pipelined architecture is presented. The architecture is derived by the application of the relaxed look-ahead technique. This technique is an approximation to the look-ahead technique and results in a minimal hardware overhead. Performance of the proposed algorithm is demonstrated for NEXT cancellers used in a 125 Mb/s twisted-pair distributed data interface (TPDDI) and the 155 Mb/s asynchronous transfer mode (ATM) local area network (LAN) applications. It is shown that the proposed pipelined architecture can be clocked at a rate of 107 times faster than the serial architecture with a maximum of 2.0 dB loss in the signal-to-noise ratio (SNR) for both the TPDDI and the ATM LAN applications.