Blockchain technology has become the epitome of security due to its traceability, transparency, immutability, and decentralized nature. To deploy blockchain into any network, a wide variety of mathematical operations are needed to be performed. These operations include hashing, block verification, encryption, and mining. Implementation of these operations is highly complex and requires heavy computational operations to be performed in real-time. Thus, high-performance computing engines are needed to develop blockchain systems, which can be done via the design of parallel processing, pipelining, and VLSI process miniaturization. Apart from high-performance computing, blockchain systems require high-performance communication, which enables seamless node-to-node connectivity for a truly decentralized implementation. Current 4th\documentclass[12pt]{minimal}
\usepackage{amsmath}
\usepackage{wasysym}
\usepackage{amsfonts}
\usepackage{amssymb}
\usepackage{amsbsy}
\usepackage{mathrsfs}
\usepackage{upgreek}
\setlength{\oddsidemargin}{-69pt}
\begin{document}$$_{th}$$\end{document} Generation (4G) communication models do not have enough throughput which is needed for high-performance blockchain-based communications, due to rapid block transfers for verification, mining, and viewing operations. Thus, 5G networks are the best application for deploying truly decentralized blockchain systems due to their low communication overheads, and high throughput. As the length of the blockchain increases, 5G networks will also face scalability issues due to an exponential increase in packet sizes. This will reduce the QoS performance of large length blockchains, thereby making them less useful for real-time large-scale applications like Industrial IoT, medical data privacy, etc. Thus, in this work, a machine learning-based sidechaining model is proposed, that utilizes a modified Genetic Algorithm powered engine. This engine aims at splitting the underlying blockchain into sidechains, thereby reducing mining complexity and reducing the number of packets needed for communication while maintaining true decentralization. The model is compared with standard blockchain & sidechain implementations in terms of access time, reading delay, and writing delay. It is observed that the model when implemented under 5G network emulation, outperforms existing blockchain and sidechain models in terms of these parameters, thereby making it useful for highly scalable blockchain systems.
