Publications

Abstract : Orthogonal frequency-division multiplexing has become broadly employed in modern communication technology with wireless systems. It subdivides a radio channel together into a significant number of clustered subchannels to provide more reliable data transmission at high rates of speed. Our project is to approximate the communication channel medium using deep learning in of DM systems. Deep learning has been shown to play a critical role in increasing system performance and lowering computing complexity in today’s wireless communication networks. The efficiency of the deep learning model is exploited to conduct channel estimation in the wireless medium. the proposed model is built on the network of Long short-term memory(LSTM) model associated with LS estimates. The profile of the channel estimated using least square (LS) and linear minimum mean square error (LMMSE) through pilot symbols is compared with the proposed model based on LSTM and evaluated using bit error rate (BER) and signal to noise ratio. This study focussed on using fewer pilots to estimate the channel and thereby increasing the spectral efficiency and data rate. The results provide evidence that a deep neural network guarantees a promising channel estimation in comparison with classic algorithms.