Proof of learning: Two Novel Consensus mechanisms for data validation using Blockchain Technology in Water Distribution System

This paper proposes an architecture of a data validation system for Water distribution system (WDS) utilising machine learning as two consensus mechanisms instead of the typical consensus mechanism based on the hashing function. The two consensus mechanisms are called Proof-of-single-learning (PoSL) and Proof-of-multiple-learning (PoML) in which the data is validated based on the learning. These two novel methods are compared with the other five hashing-based consensus mechanisms: Proof-of-Work (PoW), Proof-of-Trust (PoT), Proof-of-Vote (PoV), Proof-of-Assignment (PoA), and proof-of-Authentication (PoAuth) for evaluation. Five case studies of WDS are applied and three performance metrics, and two data conversion methods are utilised. Throughput, latency and operations per transaction (OpT are investigated to evaluate the proposed system.