Computationally Efficient Rate-Power Evaluation for Energy-Efficient UAV Inspection Communication Systems

Abstract

Unmanned aerial vehicles (UAVs) are widely used for intelligent inspection, where sensing data must be collected under limited onboard energy and flight time. Multi-antenna reception can improve data-collection capability, but increasing the number of onboard antennas also raises the overall power consumption of the UAV communication system. Motivated by this rate-power tradeoff, this paper studies a four-antenna UAV inspection communication system that employs power-efficient receiver architectures to mitigate the power increase caused by antenna expansion under the low-power constraints of compact inspection platforms. The achievable rate is characterized by mutual information and formulated as a high-dimensional integral within a UAV-oriented rate-power framework. To reduce the cost of repeated rate-curve computation, a multi-fidelity quasiMonte Carlo (MF-QMC) method is developed by correcting a low-fidelity QMC proxy curve with high-fidelity anchor points. Simulation results show that the power-efficient four-antenna onboard reception scheme reduces receiver-chain power consumption by over 80% while maintaining favorable achievablerate performance. Compared with Full-QMC, MF-QMC closely preserves the achievable-rate curve while reducing the number of integrand evaluations by 31.9% and the runtime by 24.9%.