Blockchain-backed Verifiable Offloading for Edge-assisted Augmented Reality with Energy-aware Scheduling

Abstract

The rapid proliferation of Augmented Reality (AR) applications demands significant computational resources that exceed the capabilities of mobile devices, necessitating efficient offloading strategies to edge computing environments. Traditional offloading frameworks face significant challenges, including security vulnerabilities, a lack of verifiable computation integrity, and suboptimal energy consumption patterns that limit the performance and user experience of AR applications. This paper proposes BEAVEAR (Blockchain Energy-Aware Verifiable Edge AR), a novel framework that integrates blockchain technology with edge-assisted AR computing to address these fundamental limitations.
The proposed system leverages blockchain's immutable ledger capabilities to ensure verifiable computation offloading, where smart contracts automatically validate task execution results and maintain cryptographic proof of computational integrity. The framework incorporates energy-aware scheduling algorithms that optimize resource allocation across heterogeneous edge nodes while considering dynamic voltage and frequency scaling (DVFS) to minimize power consumption without compromising application performance. The blockchain-based trust mechanism eliminates the need for centralized authorities while providing transparent audit trails for all computational transactions between mobile AR devices and edge servers.
Experimental evaluation demonstrates that BEAVEAR achieves superior performance compared to conventional offloading approaches, reducing energy consumption by 23.7% while maintaining computation verification overhead below 4.2%. The system successfully handles dynamic AR workloads with varying computational demands, achieving 98.3% task completion reliability and reducing average response latency by 31.4% compared to cloud-based alternatives. The blockchain integration offers robust security guarantees against malicious edge nodes while maintaining computational efficiency through optimized consensus mechanisms specifically designed for edge environments.
The research contributions include: (1) a novel blockchain-enabled verifiable offloading architecture specifically designed for AR applications; (2) energy-aware scheduling algorithms that dynamically optimize resource allocation considering heterogeneous edge device capabilities; (3) a comprehensive security framework ensuring computation integrity through cryptographic verification; and (4) extensive performance evaluation demonstrating practical feasibility for real-world AR deployments.