Performance Evaluation of Wireless PoW Blockchain Networks: A Complex Network Perspective

Abstract

The impact of complex wireless networks on the blockchain has not been well considered since the blockchain system is primarily designed in wired communication environments with enough communication resources. However, deploying blockchain systems in IoT sensors is often accompanied by complex network structures. Therefore, this paper aims to analyze the impact of complex wireless networks on blockchain performance under three classic models: nearest neighbor coupled network, Watts–Strogatz model, and Erdos-Renyi model. We set different transmission rates and network sizes to simulate complex wireless environments and demonstrate the relationship between these parameters and throughput, stale block rate, and average block propagation delay. To better assist the deployment of wireless blockchains, we utilize a Graph Neural Network (GNN) to evaluate the proof-of-work (PoW) blockchain performance under various network topologies quickly.