Deep Learning Enabled Free Space Optical Communication Systems
1Jaypee Institute of Information Technology, Noida, India
2Johannesburg University, Johannesburg, South Africa
3Ajay Kumar Garg Engineering College, Ghaziabad, India
Deep Learning Enabled Free Space Optical Communication Systems
Description
The unprecedented growth of wireless communication systems has led to the advent of all five generations of wireless communication systems, from 1G (first generation) to 5G (fifth generation). However, as per the timeline, the sixth generation (6G) of wireless and mobile communication systems is in the research and development phase. It aims at providing communication services for the enormous data traffic demands supported by extremely large bandwidth (THz band) artificial intelligence (AI). Importantly, the need for self-configuration, context awareness, and cellular aggregation in 6G systems can be achieved with the aid of AI, especially deep learning models.
The THz bandwidth required for massive connectivity cannot be achieved using the conventional radio frequency (RF) system. Therefore, the line-of sight (LoS) free space optical (FSO) communication system, where the optical signal is transmitted in the free space, has been considered a viable solution. However, FSO systems suffer from atmospheric turbulence and pointing errors. Deep learning techniques have the power of executing the required operations, but at a very low cost, low complexity and can provide favorable outcomes. The scintillation of the optical beam has been analyzed in various works in the literature, which is the major cause of pointing error due to LoS nature of FSO systems, the most economical solution to it may be obtained using deep learning algorithms. Besides, channel coding, source coding, channel estimation, noise estimation, interference cancellation, network security, network failure management, and reducing the beam wander are few other applications of deep learning networks that can be proficiently applied to increase the reliability and throughput of FSO communication systems.
Besides, future generation wireless communication systems have also been envisaged to operate in small cells to provide data transfer at higher speeds with very low latency. However, in cellular communication systems, the basic limiting factor is the capacity of the backhauling network. The large data rate requirements of the backhaul networks can be easily met with the help of FSO systems. A heterogeneous formation, combining the RF and FSO forms of communication, called mixed RF/FSO relaying has also been envisaged. This Special Issue aims to welcome original research and review articles on this topic.
Potential topics include but are not limited to the following:
- Mathematical modeling of free space optical systems
- Deep learning for channel estimation of wireless systems
- Deep learning for tracking of optical beams in free space optical systems
- Deep learning based modulation and coding
- Deep learning based network security monitoring of free space optical systems
- Interference management of 6G communication systems
- Supervised/unsupervised Machine Learning Methods for free space optical applications
- Reinforcement Learning Methods for free space optical applications
- Reliability analysis of free space optical systems based on deep learning models
- Energy harvesting using deep learning/machine learning for next generation wireless systems