Resilience-Aware Edge Computing

---
Funder: VinUni
Period: 03/2023 - 02/2025
Amount: $72k
Acknowledgment: We would like to thank VinUniversity for supporting our research
---

Brief Description of the Project Concept

This project aims to develop a novel resilience-aware edge computing framework for designing, analyzing, and optimizing future large-scale IoT networks, in the presence of system uncertainties and edge nodes’ failures. The expected outcomes of this project will be innovative learning and distribution resilient resource management algorithms that provide efficient, robust, and secure resource allocation solutions for a very large number of IoT devices across different resource-constrained networks. This will significantly improve real-time responsiveness and deliver seamless and high-quality service anytime and anywhere for networks, producing economic and societal benefits (e.g., remote surgeries, autonomous driving, factory automation, etc.), so crucial to the economic success of Vietnam.

Specific Objectives of the Project

This project will lay a novel theoretical framework for leveraging synergies between communication and computation resources to yield resilient, sustainable and secure resource management schemes, in the presence of uncertainty of users’ demands and edge nodes’ failures. Specifically, the project aims to:

1) develop innovative learning and distributed optimization algorithms uniquely tailored towards massive and secure IoT networks based on the proposed novel resilience-aware edge computing (EC) framework, enabling optimal management of shared common resources under various system uncertainties.

2) develop novel wireless interfaces for next-generation EC networks (e.g. IoT and 6G) based on random access protocols that provide seamless and high-quality wireless services for a massive number of IoT devices under resource-constrained networks.

3) develop novel security solutions to defend against wireless jamming attacks to achieve both authentication and secure data transmission.

4) develop and implement the world’s first prototypes to validate the efficiency of the proposed algorithms.

Impacts

The successful development of truly smart IoT networks requires instilling resilience into the processes that manage common cyber-physical resources to prevent correlated failures. This project aims to address the heterogeneity and uncertainty of demands and ESs’ failures by augmenting recent advances in signal processing, wireless communications, operations research, optimization theory and machine learning to lay the mathematical foundations of resilient EC that enable massive and secure IoT networks.