Special Issue "Reinforcement Learning Based solutions for Next-Generation Wireless Networks Coexistence"

Submission Deadline: 01 March 2021
Submit to Special Issue
Guest Editors
Prof. Yousaf Bin Zikria, Yeungnam University, South Korea.
Prof. Rashid Ali, Sejong University, South Korea.
Prof. Ali Kashif Bashir, Manchester Metropolitan University, United Kingdom.
Prof. Imran Ashraf, Yeungnam University, South Korea.

Summary

The ability of different networks to share the same spectrum is known as network coexistence. The increasing demand for spectrum resources to meet the growing user demands is pushing to use different network technologies operating in different frequency domains. This approach will make it possible to serve more users with improved throughput or desired user requirements. The diverse dissimilar physical, media access (MAC) and network configurations posed an extremely challenging task to propose solutions to achieve an efficient and fair network coexistence. When and how to use the coexistence network to meet the user demand is a trivial task. The difference of transmission, reception, spectrum access and routing of packets demands seamless solutions without starving certain network users. In the recent past, solutions are proposed for coexistence networks such as Wi-Fi and LTE users. Even though it tries to increase fairness, however, Wi-Fi users still face unfairness and starvation. The sub-1 GHz is considered as a low band and can be used to improve the wide-area coverage. Whereas 1-2.6 and 3.5-8 GHz are called mid bands, and 24- 60 GHz are high bands. These bands are used to achieve maximal coverage, capacity, and cell edge performance. Since network traffic growth is most likely expected to exceed the development capabilities of next-generation wireless networks. Further, to achieve low latency and high throughput without compromising user experience demands new solutions and mechanisms. Therefore, The coexistence of next-generation networks is essential to meet up the future user demands.

 

The recent wave of making everything intelligent and self-sustained is the only way forward. Researchers and practitioners are proposing artificial intelligence (AI) based solutions to improve overall network performance. The algorithm and complexity of the solutions greatly vary on the deployment scenario and available resources, such as computational power. Reinforcement Learning (RL) algorithms are used to map situations to actions to maximize a numerical reward signal. In RL, the user is not told which actions to take; instead, it must discover which actions yield the most reward by trying them. RL is computationally less expensive than deep learning (DL) algorithms. It is more suitable for wireless communications and networks. RL algorithms promise to improve overall network performance. Trial and error search and delayed reward are the two intriguing features of RL. The recent deployment of RL based solutions for wireless communications and networks are producing fruitful results. RL based solutions make next-generation wireless networks self-sustainable and robust. However, The advantage of RL based solutions for next-generation wireless networks coexistence is widely unexplored.

 

This special issue will bring together academic and industrial researchers to identify and discuss technical challenges and recent results related to RL-based solutions for next-generation wireless networks coexistence. Specific topics include, but are not limited to:

• RL-based solutions for WiFi-WiFi networks Coexistence

• RL-based solutions for WiFi-Zigbee networks Coexistence

• RL-based solutions for WiFi-LTE-U networks Coexistence

• RL-based solutions for WiFi-Sub 1GHz networks Coexistence

• RL-based solutions for WiFi-Radar (5GHz) networks Coexistence

• RL-based solutions for WiFi- IEEE 802.15.4 networks Coexistence

• RL-based solutions for WiFi- IEEE 802.11ad (60 GHz) networks Coexistence

• RL-based solutions for IEEE 802.15.4- Sub 1GHz networks Coexistence

• RL-based solutions for Zigbee-Sub 1GHz networks Coexistence

• RL-based solutions for 5G-WiFi (2.4-5 GHz) networks Coexistence

• RL-based solutions for 5G-Fixed satellite services (FSS) networks Coexistence

• RL-based physical layer solutions for networks Coexistence

• RL-based MAC layer solutions for networks Coexistence

• RL-based network layer solutions for networks Coexistence

• RL-based privacy and security solutions for networks Coexistence 


Keywords
Reinforcement Learning, Wireless Networks, Networks Coexistence, Next-Generation Networks