Deep Belief Network for Prediction of Rician Fading Channel
Venkatesh P1, Saikat Majumder2
1Venkatesh P, Department of Electronics and Communication Engineering, National Institute of Technology, Raipur (Chhattisgarh), India.
2Saikat Majumder, Department of Electronics and Communication Engineering, National Institute of Technology, Raipur (Chhattisgarh), India.
Manuscript received on 04 December 2019 | Revised Manuscript received on 12 December 2019 | Manuscript Published on 31 December 2019 | PP: 195-199 | Volume-9 Issue-2S December 2019 | Retrieval Number: B11161292S19/2019©BEIESP | DOI: 10.35940/ijitee.B1116.1292S19
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open-access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: In this paper a novel channel prediction scheme is presented for rician fading channel. The channel prediction is done by using a Deep Belief Network (DBN) which is composed of two Restricted Boltzmann Machines (RBMs), this deep learning algorithm can produce fewer predictive errors than echo state networks and other predictive approaches.. Simulation results shows that the DBN channel prediction system has a lower NMSE than the prediction of the echo state network and other conventional prediction methods and the obtained SER gap between the actual CSI and predicted CSI is small.
Keywords: Channel Prediction, Deep Belief Network, Restricted Boltzmann Machine, Rician Fading.
Scope of the Article: Regression and Prediction