Beamformers are used to enhance signals by summing signals across an array of microphones. If we want to transmit a strong signal from point A to point B, we can transmit the signal from all the microphones in an array simultaneously. However, this fails because the receiver first hears the signal closer to it and then the later signals which are further away. However, based on beamforming technique we can send signals from all microphones at different times and as a result they all arrive at the receiver at the same time. This produces the affect of a strong signal from a massive microphone. This algorithm will be beneficial for applications to enhance signals and improve conversation quality for hearing impaired people. In this thesis,the focus is to build an algorithm based on Neural Network to enhance the performance specially in area of speech denoising.