Distributed Ledger Technologies for Managing Heterogenous Computing Systems at the Edge

TL;DR

This paper explores using distributed ledger technologies, specifically blockchain smart contracts, to improve edge computing for IoT systems by enabling efficient node selection and task offloading in dynamic, heterogeneous networks.

Contribution

It introduces a novel system integrating blockchain smart contracts with network performance metrics to optimize edge node selection for IoT offloading.

Findings

Demonstrated the feasibility of using blockchain smart contracts for edge node selection.

Showed the system's adaptability to various network topologies.

Validated the approach through experimental results.

Abstract

The increased use of Internet of Things (IoT) devices -- from basic sensors to robust embedded computers -- has boosted the demand for information processing and storing solutions closer to these devices. Edge computing has been established as a standard architecture for developing IoT solutions, since it can optimize the workload and capacity of systems that depend on cloud services by deploying necessary computing power close to where the information is being produced and consumed. However, as the network scale in size, reaching consensus becomes an increasingly challenging task. Distributed ledger technologies (DLTs), which can be described as a network of distributed databases that incorporate cryptography, can be leveraged to achieve consensus among participants. In recent years DLTs have gained traction due to the popularity of blockchains, the most-well known type of implementation. The reliability and trust that can be achieved through transparent and traceable transactions are other key concepts that bring IoT and DLT together. We present the design, development and conducted experiments of a proof-of-concept system that uses DLT smart contracts for efficiently selecting edge nodes for offloading computational tasks. In particular, we integrate network performance indicators in smart contracts with a Hyperledger Blockchain to optimize the offloading on computation under dynamic connectivity solutions. The proposed method can be applied to networks with varied topologies and different means of connectivity. Our results show the applicability of blockchain smart contracts to a variety of industrial use cases.