2018/11/21 by Reza Barazideh, Barazideh, Reza, Balasubramaniam Natarajan +5
Engineering · #Advanced Adaptive Filtering Techniques #Electromagnetic Compatibility and Noise Suppression #FOS: Electrical engineering #Power Line Communications and Noise #Signal Processing (eess.SP) #electronic engineering #information engineering
paper · pdf · doi:10.48550/arxiv.1811.08940
openalex publication_date 2018/11/21 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
An Adaptive Nonlinear Differential Limiter (ANDL) is proposed in this paper\nto efficiently alleviate the impact of impulsive noise (IN) in a communication\nsystem. Unlike existing nonlinear methods, the ANDL is implemented in the\nanalog domain where the broader acquisition bandwidth makes outliers more\ndetectable and consequently it is easier to remove them. While the proposed\nANDL behaves like a linear filter when there is no outlier, it exhibits\nintermittent nonlinearity in response to IN. Therefore, the structure of the\nmatched filter in the receiver is modified to compensate the filtering effect\nof the ANDL in the linear regime. In this paper, we quantify the performance of\nthe ANDL by deriving a closed-form analytical bound for the average\nsignal-to-noise ratio (SNR) at the output of the filter. The calculation is\nbased on the idea that the ANDL can be perceived as a time-variant linear\nfilter whose bandwidth is modified based on the intensity of the IN. In\naddition, by linearizing the filter time parameter variations, we treat the\nANDL as a set of linear filters where the exact operating filter at a given\ntime depends upon the magnitude of the outliers. The theoretical average bit\nerror rate (BER) is validated through simulations and the performance gains\nrelative to classical methods such as blanking and clipping are quantified.\n