Energy consumption optimization in UAV-assisted multi-layer mobile edge computing with active transmissive RIS

Unmanned Aerial Vehicle (UAV)-assisted edge computing provides low-latency and low-energy consumption computing capabilities for sparsely distributed Internet of Things (IoT) networks. In addition, the assisted UAVs provide line-of-sight links to further improve communication quality. However, the existing offloading strategies have low efficiency and high costs. Motivated by this, we propose a novel UAV-assisted multi-layer mobile edge computing network with active transmissive reconfigurable intelligent surface (RIS). The introduced an active transmissive RIS not only receives data from UAVs but also performs computing functionality. We establish an optimization to minimize the total system energy consumption under delay constraints by jointly planning UAV positions and allocating computing bits, sub-carriers, time slots, transmission power, and RIS transmission coefficient. To tackle this problem, we first use the block coordinate descent (BCD) algorithm to decouple it into four sub-problems. Then, we solve them by adopting successive convex approximation (SCA), difference-convex (DC) programming, and introducing slack variables. Experimental results demonstrate that the proposed network is superior to the other five baselines concerning energy consumption reduction. Also, the influences of system parameters are verified, including the number of IoT devices, the number of RIS elements, and the delay threshold.