Making Power Line Communications More Reliable by Using Optimal Soft Information

The Power Line Channel (PLC) is a very harsh transmission medium because of its frequency selective and time-varying transfer function that exhibits several notches in the band of interest. Moreover, a strong time varying noise is normally present and poses even more challenges to the modem designer. The time-varying and frequency selective nature of the PLC requires that adaptive and, possibly, nonlinear equalization be performed at the receiver. However, adaptive receivers suffer from severe performance degradation on time varying and Inter-Symbol Interference (ISI)impaired links because of the unreliability of the hard decisions used for channel estimation and tracking. In particular, unreliable hard-decisions are a major issue especially in harsh and time-varying environments. In the present paper, we propose to alleviate this performance loss by the use of optimal soft decisions: the Nonlinear Minimum Mean Square Error (NL-MMSE) filtered and fixed-lag smoothed estimates of the transmitted symbols.