Deep deterministic policy gradients for optimizing simulated poa blockchain networks based on healthcare data characteristics

Abstract

Blockchain technology has proven to be the best solution for digital data storage today, which is decentralized and interconnected via cryptography. Many consensus algorithms can be options for implementation. One of them is the PoA consensus algorithm, which is proven to provide high performance and fault tolerance. Blockchain has been implemented in many sectors, including the healthcare sector that has different characteristics of larger and more diverse record sizes. Implementing blockchain costs a lot of money. We used a blockchain network simulator as the best alternative in our research. The main problems with blockchain implementation are having a dynamic characteristic network and providing a blockchain system that is adaptive to network characteristics. Therefore, we propose a method to optimize the simulated PoA blockchain networks using Deep Deterministic Policy Gradients by adjusting the block size and block interval. The simulation results show an increase in effective transaction throughput of up to 9 TPS for AIH and 5 TPS for the APAC data models, and without affecting other important aspects of the blockchain.