Page 136 - ITU Journal Future and evolving technologies Volume 2 (2021), Issue 4 – AI and machine learning solutions in 5G and future networks

ITU Journal on Future and Evolving Technologies, Volume 2 (2021), Issue 4

Fig. 8 – Beam selection results for a BS serving a UAV comparing RL versus a simple baseline algorithm. The optimal result (best beam pair) is also
included. The top plot presents the reward is the magnitude of the equivalent channel for the ‑th beam pair at the time (higher reward values are
better). The bottom plot shows the AoA at the UAV at each time .

REFERENCES

[1] Y.‑G. Lim, Y. J. Cho, M. S. Sim, Y. Kim, C.‑B. Chae, [6] LASSE ‑ UFPA. Radio‑Strike: A Reinforcement Learn‑
and R. A. Valenzuela. “Map‑Based Millimeter‑Wave ing Game for MIMO Beam Selection. 2021. URL:
Channel Models: An Overview, Data for B5G Eval‑ https://aiforgood.itu.int/events/radio-
uation and Machine Learning”. In: IEEE Wire‑ strike - a - reinforcement - learning - game -
less Communications 27.4 (Aug. 2020). Conference for - mimo - beam - selection - in - unreal -
Name: IEEE Wireless Communications, pp. 54–62. engine-3-d-environments/.
ISSN: 1558‑0687. DOI: 10 . 1109 / MWC . 001 . [7] S. Wu, C. Wang, e M. Aggoune, M. M. Alwakeel, and
1900315.
X. You. “A General 3‑D Non‑Stationary 5G Wireless
[2] Hongji Huang, Song Guo, Guan Gui, Zhen Yang, Channel Model”. In: IEEE Transactions on Commu‑
Jianhua Zhang, Hikmet Sari, and Fumiyuki Adachi. nications 66.7 (July 2018). Conference Name: IEEE
“Deep Learning for Physical‑Layer 5G Wireless Transactions on Communications, pp. 3065–3078.
Techniques: Opportunities, Challenges and So‑ ISSN: 1558‑0857. DOI: 10 . 1109 / TCOMM . 2017 .
lutions”. In: IEEE Wireless Communications 27.1 2779128.
(2020), pp. 214–222. DOI: 10.1109/MWC.2019. [8] J. Bian, C.‑X. Wang, X. Gao, X. You, and M. Zhang.
1900027. “A General 3D Non‑Stationary Wireless Channel
[3] Chaoyun Zhang, Paul Patras, and Hamed Haddadi. Model for 5G and Beyond”. In: IEEE Transactions
“Deep Learning in Mobile and Wireless Network‑ on Wireless Communications (2021). Conference
ing: A Survey”. In: IEEE Communications Surveys Name: IEEE Transactions on Wireless Communi‑
& Tutorials 21.3 (2019), pp. 2224–2287. DOI: 10. cations, pp. 1–1. ISSN: 1558‑2248. DOI: 10.1109/
1109/COMST.2019.2904897. TWC.2020.3047973.
[4] Focus Group on Machine Learning for Future Net‑ [9] J. Shen et al. “Natural TTS Synthesis by Condition‑
works including 5G. https://www.itu.int/en/ ing WaveNet on Mel Spectrogram Predictions”. In:
ITU - T / focusgroups / ml5g / Pages / default . https://arxiv.org/abs/1712.05884. 2018.
aspx. (Accessed on 03/19/2020).
[10] Theodore S Rappaport, Yunchou Xing, Ojas Kan‑
[5] C. De Lima et al. “Convergent Communication, here, Shihao Ju, Arjuna Madanayake, Soumyajit
Sensing and Localization in 6G Systems: An Mandal, Ahmed Alkhateeb, and Georgios C Tri‑
Overview of Technologies, Opportunities and chopoulos. “Wireless communications and appli‑
Challenges”. In: IEEE Access 9 (2021), pp. 26902– cations above 100 GHz: Opportunities and chal‑
26925. DOI: 10.1109/ACCESS.2021.3053486. lenges for 6G and beyond” . In: IEEE Access 7 (2019),
pp. 78729–78757.

120 © International Telecommunication Union, 2021