Predictive blockchain consensus with real-time failure detection and autonomous recovery for resilient mutual distributed ledgers
- 1Department of Computer Science and Engineering, Sri Shakthi Institute of Engineering and Technology
- 2Department of Computer Science and Engineering, Kongunadu College of Engineering and Technology
Abstract
The emergence of blockchain technologies is changing how we manage data through decentralized, secure systems. In the realm of consensus mechanisms, such as PoW, PoS, and PBFT, several limitations make these technologies inadequate for handling the challenges of IoT-enabled environments and Mutual Distributed Ledgers (MDLs), which require constant and reliable access to their data. These consensus models are reactive, resulting in increased response times (latencies) when a failure or disruption occurs, decreased throughput, and extended recovery periods. The lack of adaptive intelligence to recognize and recover from failures in real-time exacerbates these network failures. This research introduces the Predictive Consensus Algorithm to Blockchain Networks with Failure Detection and Recovery in Real-Time (PCB-FDAR). PCB-FDAR provides a new mechanism by integrating machine learning-based predictive analytics with real-time network monitoring to anticipate future failures and automatically reconfigure the network without human intervention. The framework also enables fault-tolerance across interconnected blockchain environments. PCB-FDAR has been shown through experimentation to outperform traditional consensus mechanisms. When comparing chipsets with an average of 40 blocks, the PCB-FDAR framework achieves an average latency of 1,600 ms, which represents a 42.86% reduction from PoW (2,800 ms) and a 36.00% reduction over PBFT (2,500 ms). In addition, when performing scalability testing, PCBFDAR delivers as high as 1,800 transactions per second (TPS), representing a 450 × improvement over PoW (4 TPS) and a 32.7 × improvement over PoS (55 TPS). Lastly, the PCBFDAR automatic recovery mechanism reduces failure recovery time from 180 to 30 s, resulting in an 83.33% decrease and providing 99% operational availability. Thus, the results of this study demonstrate that PCB-FDAR provides a scalable, reliable, and fault-tolerant consensus framework for real-time distributed applications.
Related
- Secure and explainable fraud detection in healthcare claims using blockchain-based machine learning
- Cross-corpus consensus feature selection for language-invariant Parkinson’s disease detection: a dual-stream acoustic and spectrogram-attention framework
- Consensus and controversy of liquid biopsy using circulating tumour DNA (ctDNA) for molecular residual disease detection in colorectal cancer
- A Scalable Predictive Modeling Approach to Duplicate Detection in Pharmacovigilance: Algorithm Development and Validation Study
- Holistic and proactive machine learning approaches to predict cyberattacks in blockchain networks
BibTeX
@article{kumar2026predictive,
title = {{Predictive blockchain consensus with real-time failure detection and autonomous recovery for resilient mutual distributed ledgers}},
author = {N. M. Saravana Kumar and P. Valarmathi},
journal = {Scientific Reports},
year = {2026},
month = aug,
doi = {10.1038/s41598-026-65170-8},
url = {https://doi.org/10.1038/s41598-026-65170-8},
note = {CC BY},
}
RIS
TY - JOUR
TI - Predictive blockchain consensus with real-time failure detection and autonomous recovery for resilient mutual distributed ledgers
AU - N. M. Saravana Kumar
AU - P. Valarmathi
PY - 2026
DA - 2026/08/01
JO - Scientific Reports
DO - 10.1038/s41598-026-65170-8
UR - https://doi.org/10.1038/s41598-026-65170-8
AB - The emergence of blockchain technologies is changing how we manage data through decentralized, secure systems. In the realm of consensus mechanisms, such as PoW, PoS, and PBFT, several limitations make these technologies inadequate for handling the challenges of IoT-enabled environments and Mutual Distributed Ledgers (MDLs), which require constant and reliable access to their data. These consensus models are reactive, resulting in increased response times (latencies) when a failure or disruption occurs, decreased throughput, and extended recovery periods. The lack of adaptive intelligence to recognize and recover from failures in real-time exacerbates these network failures. This research introduces the Predictive Consensus Algorithm to Blockchain Networks with Failure Detection and Recovery in Real-Time (PCB-FDAR). PCB-FDAR provides a new mechanism by integrating machine learning-based predictive analytics with real-time network monitoring to anticipate future failures and automatically reconfigure the network without human intervention. The framework also enables fault-tolerance across interconnected blockchain environments. PCB-FDAR has been shown through experimentation to outperform traditional consensus mechanisms. When comparing chipsets with an average of 40 blocks, the PCB-FDAR framework achieves an average latency of 1,600 ms, which represents a 42.86% reduction from PoW (2,800 ms) and a 36.00% reduction over PBFT (2,500 ms). In addition, when performing scalability testing, PCBFDAR delivers as high as 1,800 transactions per second (TPS), representing a 450 × improvement over PoW (4 TPS) and a 32.7 × improvement over PoS (55 TPS). Lastly, the PCBFDAR automatic recovery mechanism reduces failure recovery time from 180 to 30 s, resulting in an 83.33% decrease and providing 99% operational availability. Thus, the results of this study demonstrate that PCB-FDAR provides a scalable, reliable, and fault-tolerant consensus framework for real-time distributed applications.
SN - 2045-2322
ER -