Treffer: Blockchain and Artificial Intelligence for Big Data Analytics in Networking: Leadingedge Frameworks.

Big Data (BD) Analytics (BDA) in networking involves acquisition, sharing, pre-processing, storage, analysis, interpretation, and decision-making. Blockchain (BC) technology incorporates a progression of bonded blocks that fundamentally upholds credibility, protects unquestionability, and protects the partial-anonymity of its transactions on account of distributed consensus methods and cryptographic protocols. So as to fulfill the deficiency of a review paper catering to individual and combined use of BC and Artificial Intelligence (BCandAI) for BDA in the networking domain, in this work, we recognize 6 sections in the leading edge BCandAI BDA notion and rigorously analyze each stratagem concerning blockchain attributes, blockchain/AI techniques, network attributes, and the like. We piled up an opening sample of 89 publication citations by culling articles for screening requirements tracked down from cyber libraries, availing a comprehensive and protracted systemology. Established upon this exploration, we highlight that Artificial Intelligence (AI) can be involved in BDA by analyzing BD, while blockchain can facilitate secure transmission and storage of BD due to its inherent security features of unchangeability, non-deniability, etc., preventing data poisoning attacks, and facilitating hybrid on- and off-chain storage due to the challenges of high volume by availing techniques just like offloading and partial storage. Moreover, we highlight that there are BCandAI integrated approaches where blockchain-anchored secure BD storage is availed for secure federated learning or blockchain and cloud computing are availed for BD fusion for analysis, availing AI to generate accurate insights from BD. Rigorous analysis discloses that from all studies, 17.5% use BC alone, 20% avail of the combined BCandAI concept, 62.5% use AI alone, 70% address one or more BDA stages, 10% implement PoW consensus, 12.5% avail of deep learning, and 17.5% choose generic or IoT networks. Finally, we express the potentialities and problems of the proposition of BCandAI-anchored BDA concepts and then offer counsel to overpower them. [ABSTRACT FROM AUTHOR]