Power Allocation and Trajectory Planning for UAV-Assisted PV Plant Inspection Control With URLLC and NOMA

Cellular network is commonly utilized for transmitting inspection control data to distributed photovoltaic (PV) plants, which incurs high costs and low efficiency. To solve this problem, an unmanned aerial vehicle (UAV)-assisted PV plant inspection control system subject to ultra-reliable low latency communication (URLLC) and non-orthogonal multiple access (NOMA) constraints is established. In this system, channel state information (CSI) is difficult to predict in path planning and real-time measurement CSI cannot be used, necessitating the consideration of imperfect CSI. Accordingly, a non-convex optimization problem is put forward to minimize average transmitted power through PV panel clustering, bandwidth allocation and UAV trajectory planning under imperfect successive interference cancellation (SIC). Firstly, aiming at solving PV panel clustering subproblem, a clustering algorithm maximizing PV panel link angle difference budget is proposed. To solve the bandwidth allocation subproblem, a primal-dual iterative solution is proposed which reduces its feasible domain iteratively. Last, as for the UAV trajectory optimization subproblem, a geometric fitting algorithm is proposed considering field of view (FoV) constraint. Simulation results show that, compared with the existing schemes, proposed algorithm can further reduce UAV transmission power in few iteration, thus reducing the computational complexity.