Dynamic edge computing empowered by reconfigurable intelligent surfaces

Abstract

In this paper, we propose a novel algorithm for energy-efficient low-latency dynamic mobile edge computing (MEC), in the context of beyond 5G networks endowed with reconfigurable intelligent surfaces (RISs). We consider a scenario where new computing requests are continuously generated by a set of devices and are handled through a dynamic queueing system. Building on stochastic optimization tools, we devise a dynamic learning algorithm that jointly optimizes the allocation of radio resources (i.e., power, transmission rates, sleep mode and duty cycle), computation resources (i.e., CPU cycles), and RIS reflectivity parameters (i.e., phase shifts), while guaranteeing a target performance in terms of average end-to-end delay. The proposed strategy enables dynamic control of the system, performing a low-complexity optimization on a per-slot basis while dealing with time-varying radio channels and task arrivals, whose statistics are unknown. The presence and optimization of RISs helps boosting the performance of dynamic MEC, thanks to the capability to shape and adapt the wireless propagation environment. Numerical results assess the performance in terms of service delay, learning, and adaptation capabilities of the proposed strategy for RIS-empowered MEC.

Power Minimizing MEC Offloading with QoS Constraints over RIS-Empowered Communications

“…In all simulations, we compare the RIS-aided performance with the Lorentzian model and our frequency selective-aware optimization, termed as optimized RIS, with three benchmarks: i) the case without the RIS, termed as direct link, with resources optimized with the proposed method; ii) the frequency flat RIS case, termed as freq-flat RIS, in which the RIS response does not follow the realistic Lorentzian model, but it is flat across all frequencies. In this last case, the RIS is optimized as in [12], while resources are optimized through our method; iii) random RIS: the RIS parameters are randomly selected from the feasible set at each time slot, and resources are accordingly optimized with our method. As a first result, in Fig.…”

“…Therefore, we now propose an efficient heuristic to optimize the RIS configuration. The algorithm to optimize the RIS builds on the one proposed in [12], in which RIS elements responses are subsequently selected with the goal of increasing a weighted sum of channel power gains. However, [12] is limited to frequency selective channels and focuses on a multi-user case.…”

“…The algorithm to optimize the RIS builds on the one proposed in [12], in which RIS elements responses are subsequently selected with the goal of increasing a weighted sum of channel power gains. However, [12] is limited to frequency selective channels and focuses on a multi-user case. Therefore, in this case, we propose to subsequently select RIS parameters (i.e.…”

“…Furthermore, data-driven solutions for RIS-empowered multi-user mobile edge computing is proposed in [11], with the goal of maximizing the total completed task-input bits of all UEs under energy budget constraints. Finally, in [12], a joint optimization of radio, computing, and RIS parameters was proposed considering frequency-flat channels and RIS responses. A similar contribution was presented in [13] for the case of multiple-input-multiple-output communications.…”