DESIGN OF AN INTEGRATED MODEL USING MULTIAGENT BIOINSPIRED DEEP REINFORCEMENT LEARNING FOR SCALABLE BLOCKCHAIN-BASED MACHINE LEARNING SYSTEMS
DOI:
https://doi.org/10.11113/aej.v15.24099Keywords:
Consensus Algorithms, Machine Learning, Resource Optimization, Dynamic Sharding, ScenariosAbstract
With the increasing complexity and scale of ML workloads, the necessity to develop secure and efficient consensus mechanisms for blockchain-based ML is now becoming a crucial requirement. Among the two existing consensus protocols, that include PoW and PoS, both are relatively incapable of handling heterogeneous ML tasks because of both high computationally expensive overhead and its energy inefficiency as well as an inability to dynamically adapt itself in varying workloads. In doing so, this work proposes a state-of-the-art suite of consensus designs based on five optimization techniques: the ABC algorithm, MADRL, GA-based sharding, ACO, and DS-RDA. The ABC algorithm optimizes node assignment based on network states and task complexity, which produces 20-30% latency reduction and up to a 25% improvement in energy efficiency. MADRL enables decentralized resource allocation across heterogeneous ML workloads, enhancing throughput by up to 40% and timestamp to complete tasks by 35%. GA-based sharding optimizes the dynamic partitioning process, provides 50% better shard balance and improvements in using resources of up to 30%. ACO can off-chain the computation tasks efficiently; thereby, it reduces 40% of the load from the chain and improves parallel processing by increasing 25%. Last but not the least, DS-RDA improves scalability due to the dynamic nature of reconfiguring shards in line with real-time fluctuations in workload, thus increasing up to 45% throughput under high-load conditions. Collectively, they significantly enhance computation efficiency, resource distribution, and scalability with blockchain security in order to provide a new and robust solution to the present limitations within the existing consensus protocol of machine learning operations.
References
N. Afraz, F. Wilhelmi, H. Ahmadi and M. Ruffini, 2023. "Blockchain and Smart Contracts for Telecommunications: Requirements vs. Cost Analysis," in IEEE Access, 11: 95653-95666, DOI: https://doi.org/10.1109/ACCESS.2023.3309423
C. Xu, Y. Qu, T. H. Luan, P. W. Eklund, Y. Xiang and L. Gao, 2022. "A Lightweight and Attack-Proof Bidirectional Blockchain Paradigm for Internet of Things," in IEEE Internet of Things Journal, 9: 4371-4384, DOI: https: // doi.Org / 10.1109 /JIOT. 2021.3103275
H. Kim and D. Kim, 2024. "Methodological Advancements in Standardizing Blockchain Assessment," in IEEE Access, 12: 35552-35570, DOI: https: // doi.Org /10 .1109 / ACCESS. 2024. 3372578
L. Lu et al, 2023. "iQuery: A Trustworthy and Scalable Blockchain
Analytics Platform," in IEEE Transactions on Dependable and Secure Computing, 20: 4578-4592, DOI: https:// doi.org / 10.1109/TDSC.2022.3228908
M. Touloupou, K. Christodoulou and M. Themistocleous, 2024. "Validating the Blockchain Benchmarking Framework Through Controlled Deployments of XRPL and Ethereum," in IEEE Access, 12: 22264-22277, DOI: https:// doi.org / 10.1109 / ACCESS. 2024.3363833
J. Wang et al, 2024. "LearningChain: A Highly Scalable and Applicable Learning-Based Blockchain Performance Optimization Framework," in IEEE Transactions on Network and Service Management, 21: 1817-1831, DOI: https: // doi.org / 10. 1109 / TNSM.2023.3347789
Y. Liu et al, 2024. "SS-DID: A Secure and Scalable Web3 Decentralized Identity Utilizing Multilayer Sharding Blockchain," in IEEE Internet of Things Journal, 11: 25694-25705, DOI: https://doi.org/10.1109/JIOT.2024.3380068
M. M. Merlec and H. P. In, 2024. "SC-CAAC: A Smart-Contract-Based Context-Aware Access Control Scheme for Blockchain-Enabled IoT Systems," in IEEE Internet of Things 11: 19866-19881, DOI: https://doi.org/10.1109/JIOT.2024.3371504
X. Hao, W. Ren, Y. Fei, T. Zhu and K. -K. R. Choo, 2023. "A Blockchain Based Cross-Domain and Autonomous Access Control Scheme for Internet of Things," in IEEE Transactions on Services Computing, 16: 773-786, DOI: https: // doi.org / 10.1109 / TSC. 2022.3179727
R. H. Kim, H. Noh, H. Song and G. S. Park, 2022. "Quick Block Transport System for Scalable Hyperledger Fabric Blockchain Over D2D Assisted 5G Networks," in IEEE Transactions on Network and Service Management, 19: 1176-1190, DOI: https://doi.org/10.1109/TNSM.2021.3122923
H. J. De Moura Costa, C. A. Da Costa, R. Da Rosa Righi, R. S. Antunes, J. F. De Paz Santana and V. R. Q. Leithardt, 2022. "A Fog and Blockchain Software Architecture for a Global Scale Vaccination Strategy," in IEEE Access, 10: 44290-44304, DOI: https://doi.org/10.1109/ACCESS.2022.3169418
R. A. Mishra, A. Kalla, A. Braeken and M. Liyanage, 2023. "Blockchain Regulated Verifiable and Automatic Key Refreshment Mechanism for IoT," in IEEE Access, 11: 21758-21770, DOI: https://doi.org/10.1109/ACCESS.2023.3251651
Y. Tang, J. Yan, C. Chakraborty and Y. Sun,2023. "Hedera: A Permissionless and Scalable Hybrid Blockchain Consensus Algorithm in Multiaccess Edge Computing for IoT," in IEEE Internet of Things Journal, 10: 21187-21202, DOI: https://doi.org/10.1109/JIOT.2023.3279108
N. Sivaselvan, K. V. Bhat, M. Rajarajan and A. K. Das, 2023. "A New Scalable and Secure Access Control Scheme Using Blockchain Technology for IoT," in IEEE Transactions on Network and Service Management, 20: 2957-2974, DOI: https://doi.org/10.1109/TNSM.2023.3246120
W. Liu, Z. Wan, J. Shao and Y. Yu, 2023. "HyperMaze: Towards Privacy Preserving and Scalable Permissioned Blockchain," in IEEE Transactions on Dependable and Secure Computing 20: 360-376, DOI: https://doi.org/10.1109/TDSC.2021.3133840
M. S. Farooq, Z. Kalim, J. N. Qureshi, S. Rasheed and A. Abid, 2022. "A Blockchain-Based Framework for Distributed Agile Software Development," in IEEE Access, 10: 17977-17995, DOI: https://doi.org/10.1109/ACCESS.2022.3146953
I. Aviv, A. Barger, A. Kofman and R. Weisfeld, 2023. "Reference Architecture for Blockchain-Native Distributed Information System," in IEEE Access, 11: 4838-4851, DOI: https://doi.org/10.1109/ACCESS.2023.3235838
J. Shu, X. Zou, X. Jia, W. Zhang and R. Xie, 2022. "Blockchain-Based Decentralized Public Auditing for Cloud Storage," in IEEE Transactions on Cloud Computing, 10: 2366-2380, DOI: https://doi.org/10.1109/TCC.2021.3051622
M. Usman, M. S. Sarfraz, M. U. Aftab, U. Habib and S. Javed, 2024. "A Blockchain Based Scalable Domain Access Control Framework for Industrial Internet of Things," in IEEE Access, 12: 56554-56570, DOI: https: // doi.org / 10.1109 / ACCESS. 2024. 3390842
O. Kuznetsov, A. Rusnak, A. Yezhov, D. Kanonik, K. Kuznetsova and S. Karashchuk, 2024. "Enhanced Security and Efficiency in Blockchain With Aggregated Zero-Knowledge Proof Mechanisms," in IEEE Access, 12: 49228-49248, DOI: https://doi.org/10.1109/ACCESS.2024.3384705
J. Xi et al. 2023. "A Blockchain Dynamic Sharding Scheme Based on Hidden Markov Model in Collaborative IoT," in IEEE Internet of Things Journal, 10: 14896-14907, DOI: https://doi.org/10.1109/JIOT.2023.3294234
T. Wu, G. Jourjon, K. Thilakarathna and P. L. Yeoh, 2023. "MapChain-D: A Distributed Blockchain for IIoT Data Storage and Communications," in IEEE Transactions on Industrial Informatics, 19: 9766-9776, DOI: https://doi.org/10.1109/TII.2023.3234631
C. T. Nguyen, D. T. Hoang, D. N. Nguyen, Y. Xiao, D. Niyato and E. Dutkiewicz, 2024. "MetaShard: A Novel Sharding Blockchain Platform for Metaverse Applications," in IEEE Transactions on Mobile Computing, 23: 4348-4361, DOI: https://doi.org/10.1109/TMC.2023.3290955
R. Jin, J. Hu, G. Min and J. Mills, 2023. "Lightweight Blockchain Empowered Secure and Efficient Federated Edge Learning," in IEEE Transactions on Computers, 72: 3314-3325, DOI: https://doi.org/10.1109/TC.2023.3293731
V. Agarwal and S. Pal, 2024. "HierChain: A Hierarchical-Blockchain-Based Data Management System for Smart Healthcare," in IEEE Internet of Things Journal, 11:
