Azzedine Zerguine2022-05-182022-05-184649https://drepo.sdl.edu.sa/handle/20.500.14154/3234The main idea of equalization is simply to compensate for nonideal characteristics by additional filtering. An equalizer is, then, an adaptive filter that is capable of high-speed digital signaling over slowly varying, band limited channels whose impulse responses are unknown at the receiver. The fractionally-spaced equalizer is superior to the synchronous equalizer, and this is in its ability to be not affected by aliasing problems and independent of the sampler phase. However, for a given time span NT, a synchronous equalizer will need N tap coefficients, while (2N) tap coefficients are needed for a fractionally-spaced equalizer with tap spacing T/2. This will result in a great amount of computations needed to update the tap coefficients every symbol period, an important drawback of the fractionally-spaced equalizer. In this thesis, we introduce an idea where a compromise between performance and complexity, for both equalizers, is presented. The idea consists of setting some of the tap coefficients of a fractionally-spaced equalizer with tap spacing T/2, to zero. The zerotap coefficients are neither updated nor entered in the computational process, thus leading to simplified algorithms.enThesis