Cognitive blockchain and its application to optimize performance in blockchain systems

A distributed ledger called a blockchain is used for logging authenticated cryptographic transactions. The global ledger is updated with transactions using consensus techniques. Consensus algorithms are developed for networks with untrusted nodes to achieve reliability. Academics are paying attention to this technology because it has essential features like decentralization, stability, anonymity, and transparency. Even though the blockchain has some unique features, it has to deal with many challenges and restrictions, such as scalability, security, hidden centrality, and high cost. Artificial intelligence and blockchain are technologies that have been much discussed in the last decade and are developing rapidly. Combining the two to meet the existing challenges can have fascinating results. In this paper, we introduce the novel idea of cognitive blockchain by incorporating intelligent thought into the blockchain. With cognitive capabilities, blockchain technology can perceive the state of the network, analyze the data it has collected, make good decisions, and take appropriate action to improve network performance. We provide an operational framework for cognitive blockchain, which primarily refers to the connections between fundamental cognitive processes, including the perception-action cycle, data analytics, knowledge discovery, intelligent decision-making, and service provision. Then, using learning automata, we provide methods for creating cognitive engines for performance optimization by intelligently adjusting the block size, time interval, and validators in blockchain systems with BFT-based consensus algorithms. Several experiments have been conducted to assess the suggested approaches' effectiveness. The findings demonstrate that the proposed performance optimization approach improves blockchain performance criteria. We introduced a new concept of cognitive blockchain and presented the framework of cognitive blockchain in this paper. Then we propose approaches based on learning automata for designing cognitive engines for performance optimization to select the block size, time interval, and validators in blockchain systems with BFT-based consensus algorithms. image