A deep learning based solution for imperfect CSI problem in correlated FSO communication channel
arXiv:1909.11002
Abstract
Imperfect channel state information (CSI) at the receiver, which is due to channel estimation error, is one of the main problems toward achieving optimum detection. This paper presents a deep learning based structure for combating this issue. In order to show the effect of using deep learning, the symbol error rate of a simple free space optical (FSO) communication system is simulated over correlated and un-correlated log-normal channel with write/ wrong CSI. Novelties and contributions of this paper, which are done for the first in machine learning for FSO communication include considering deep learning, considering Log-normal channel, considering correlated channel, considering imperfect CSI. The proposed deep learning based structure is compared with maximum likelihood detector, it is shown that in perfect CSI, both perform the same (because maximum likelihood is optimum), but in imperfect CSI, proposed deep learning based structure outperforms maximum likelihood in channels with un-correlation or desired correlation lengths.
References in corpus (6)
- Performance of a Relay-Assisted Hybrid FSO/RF Communication System
- New expression on the performance of a novel multi-hop relay-assisted hybrid FSO / RF communication system
- New expressions on the performance of a novel multi-hop relay-assisted hybrid FSO / RF communication system with receive diversity
- An optimization problem on the performance of FSO communication system
- On the Performance of a Relay-Assisted Multi-Hop Asymmetric FSO/RF Communication System over Negative Exponential atmospheric turbulence with the effect of pointing error
- Novel Suboptimal approaches for Hyperparameter Tuning of Deep Neural Network [under the shelf of Optical Communication]