Abstract
Recently, the hybrid relay-reflecting intelligent surface (HRRIS) has been introduced as a spectral- and energy-efficient architecture to assist wireless communication systems. In the HRRIS, a single or few active relay elements are deployed along with a large number of passive reflecting elements, allowing it to not only reflect but also amplify the incident signals. In this work, we investigate the potential of the HRRIS in aiding the computation offloading in a single-user mobile edge computing system. The objective is to minimize the offloading latency while ensuring the secrecy of user data against a malicious eavesdropper. We develop efficient solutions to this latency minimization problem based on alternating optimization. Through numerical results, we show that the deployment of the HRRIS can result in a considerable reduction in latency. Furthermore, the latency reduction gain offered by the HRRIS is much more significant than that of the conventional reconfigurable intelligent surface (RIS).
Introduction: The Challenge of Secure MEC
Mobile Edge Computing (MEC) is a transformative technology that allows resource-constrained devices, such as IoT sensors or smartphones, to offload intensive computational tasks to nearby edge servers. However, two critical challenges persist: high latency due to wireless channel fading and security vulnerabilities from malicious eavesdropping. While Reconfigurable Intelligent Surfaces (RIS)—essentially "smart mirrors" for radio waves—have been proposed to mitigate these issues, their purely passive nature limits their ability to overcome severe path loss, as they can redirect but not amplify signals.
System Architecture: The HRRIS Advantage
To address these limitations, this paper proposes an HRRIS-assisted MEC system. The HRRIS architecture integrates a large number of passive reflecting elements with a small number of active relaying elements equipped with power amplifiers. This hybrid approach allows the surface to overcome the "double path loss" effect (User → Surface → Edge Node) that often renders passive RIS ineffective in challenging environments.
The system model considers a single user attempting to offload data to an Edge Node (EN) in the presence of a multi-antenna eavesdropper. By strategically placing active elements, the HRRIS can significantly boost the signal strength at the legitimate receiver while potentially nullifying the signal at the eavesdropper.
Optimization Strategy
The core of the research is a joint optimization problem designed to minimize total latency. The authors optimize three key components:
- Offloading Volume: Determining the optimal split between local processing and offloading.
- Beamforming: Refining how the edge server filters incoming signals.
- HRRIS Coefficients: Adjusting phase shifts for all elements and amplification levels for active elements.
This complex problem is solved using an Alternating Optimization (AO) approach, which iteratively refines each variable. The study also distinguishes between Fixed HRRIS (static active elements) and Dynamic HRRIS (adaptive selection of active elements based on channel conditions).
Performance Evaluation
Numerical simulations using standard wireless parameters (e.g., Rician fading, path-loss models) demonstrate that HRRIS-assisted systems consistently achieve lower latency than both local computing and conventional RIS-assisted schemes.
As shown in Figure 3, the latency reduction is particularly pronounced when the user is far from the edge node. A key insight is that even a single active element can provide enough gain to make offloading viable where passive RIS fails. Furthermore, the Dynamic HRRIS architecture shows remarkable resilience; in scenarios with well-equipped eavesdroppers, it maintained a 9% performance gain over other schemes by intelligently exploiting its power budget.
Conclusion and Future Work
This work establishes that HRRIS can drastically reduce latency in MEC while maintaining robust security. While the current study focuses on a single-user scenario, it provides a foundation for future research into multi-user resource allocation and interference management in complex, multi-eavesdropper environments.
Summary updated