Speaker verification (SV) suffers from unsatisfactory performance in\nfar-field scenarios due to environmental noise andthe adverse impact of room\nreverberation. This work presents a benchmark of multichannel speech\nenhancement for far-fieldspeaker verification. One approach is a deep neural\nnetwork-based, and the other is a combination of deep neural network andsignal\nprocessing. We integrated a DNN architecture with signal processing techniques\nto carry out various experiments. Ourapproach is compared to the existing\nstate-of-the-art approaches. We examine the importance of enrollment in\npre-processing,which has been largely overlooked in previous studies.\nExperimental evaluation shows that pre-processing can improve the SVperformance\nas long as the enrollment files are processed similarly to the test data and\nthat test and enrollment occur within similarSNR ranges. Considerable\nimprovement is obtained on the generated and all the noise conditions of the\nVOiCES dataset.\n