Joint Beamforming, User Association, and Height Control for Cellular-Enabled UAV Communications

• Published in: IEEE transactions on communications Vol. 69; no. 6; pp. 3598 - 3613
• Main Authors: Hou, Jiancao, Deng, Yansha, Shikh-Bahaei, Mohammad
• Format: Journal Article
• Language: English
• Published: New York IEEE 01.06.2021; The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
• Subjects: Algorithms; Array signal processing; Beamforming; Cellular communication; cellular networks; Channel estimation; height control; Iterative algorithms; Iterative methods; MIMO beamforming; Numerical models; Optimization; Path planning; Reliability; spectrum efficiency; UAV association; Unmanned aerial vehicles; Uplink
• ISSN: 0090-6778; 1558-0857
• DOI: 10.1109/TCOMM.2021.3063775

Supporting reliable and seamless connectivity for aerial users, such as Unmanned Aerial Vehicles (UAVs), is one of the key challenges for the next-generation cellular networks. To tackle this challenge, we propose a joint beamforming, user association, and UAV-height control framework for cellular-connected multi-UAV communications. Our objective is to maximize the minimum achievable rate for UAVs subject to co-existing terrestrial users' rate constraints. A hierarchical bi-layer iterative algorithm is devised to solve the problem. By using the projection gradient method in inner layer iterations and the geometric program modeling plus the convex-concave procedure in outer layer iterations, our proposed algorithm is proved to converge to a local optimum. We also examine our proposed algorithm under the practical condition where channel estimation is not perfect. Numerical results show that our proposed joint beamforming, user association, and UAV-height control framework outperforms the conventional nearest UAV association method in terms of UAVs' minimum achievable rate for both perfect and imperfect channel estimation cases. We also observe different UAVs' heights (i.e., between 100m and 300m) do not affect the UAVs' achievable rates. For the case with moving UAV, we also study the trade-off between UAVs' minimum achievable rate and the frequency of updating optimization variables.