Reinforcement learning for multi-metric QoS optimization in 5G/6G networks with Shannon capacity integration

Scientific Reports

卷:16

Article

Hegazi M. Ibrahim^1,2  Ibrahim Elewah³  Turki M. Alanazi⁴  Magda I. EL-Afifi^1,5  Alyaa A. Hamza⁶  Hesham A. Sakr^1,7,8,9 

[1] Department of Electrical Engineering, College of Engineering, University of Hafr Al Batin, Hafr Al Batin, Saudi Arabia

[2] Department of Information Technology, College of Informatics, Midocean University, 98123, Moroni RP, Comoros

[3] Nile Higher Institute for Engineering and Technology, Communications and Electronics Engineering Department, Mansoura, Egypt

[4] Faculty of Information Technology, Department of Computer science, Isra University, Amman, Jordan

[5] College of Engineering and Technology, American University of the Middle East, 15453, Dasman, Kuwait

[6] Elsewedy University of Technology (SU TECH), Cairo, Egypt

[7] Faculty of Engineering Technology, Elsewedy University of Technology (SU TECH), Cairo, Egypt

[8] Faculty of Computing and IT, Sohar University, Sohar, Oman

[9] Faculty of Engineering, New Mansoura University, New Mansourah, Egypt

关键词: Reinforcement learning (RL);    Quality of service (QoS);    Signal-to-noise ratio (SNR);    Bit error rate (BER);    6G networks;   

DOI  :  10.1038/s41598-026-59147-w

 received in 2026-03-07, accepted in 2026-06-18,  发布年份 2026

来源: Springer

【 摘 要 】

This study investigates the use of reinforcement learning (RL) techniques to optimize Quality of Service (QoS) parameters in both 5G and 6G wireless networks. A multi-phase experimental framework was employed to evaluate key QoS metrics including latency, jitter, packet loss, throughput, signal-to-noise ratio (SNR), and bit error rate (BER) across both generations. In the first phase, an RL agent was trained in a 5G environment, achieving notable improvements in latency reduction, jitter control, and throughput enhancement. Comparative analysis in the second phase demonstrated that 6G networks consistently outperformed 5G, yielding higher RL rewards, lower packet loss, and reduced jitter and latency. In the final phase, Shannon Capacity theory was integrated into the RL model, further enhancing transmission reliability and signal quality in the 6G context. Additional testing with video streaming scenarios confirmed 6G’s superior capability in supporting real-time, high-reliability applications. Overall, the findings indicate that 6G networks, when optimized with RL, provide a more intelligent, robust, and efficient solution for future wireless communication systems.

【 授权许可】

CC BY-NC-ND   
© The Author(s) 2026

【 图 表 】

Fig. 4

Fig. 2

Fig. 4

Fig. 2

Fig. 1

Fig. 9

Fig. 8

Fig. 4

Fig. 11

Fig. 4

Fig. 5

Fig. 3

Fig. 7

Fig. 1

Fig. 1

Fig. 6

Fig. 9

Fig. 1

Fig. 7

Fig. 2