ARCHITECTURAL MODEL OF FAULT-TOLERANT INTEGRATION OF PERIPHERAL IOT NETWORKS WITH BLOCKCHAIN LEDGERS BASED ON THRESHOLD CRYPTOGRAPHY

Abstract

The article develops and mathematically justifies a novel architectural model for integrating decentralized

peripheral Internet of Things (Edge IoT) networks with distributed blockchain ledgers. The relevance of the study is driven by the fact that the global digitalization of critical infrastructure and environmental monitoring requires highly reliable interaction between IoT and blockchain technology. However, classical paradigms of direct interaction generate critical systemic problems: the irreversible fixation of compromised data, known as "Garbage In, Garbage Out" (GIGO), the impossibility of horizontal scaling due to strict smart contract limits, and the absolute vulnerability of the system to massive Byzantine manipulations and stealth gradient attacks. The aim of the research is to create a comprehensive model capable of autonomously filtering anomalies on the periphery and guaranteeing the secure recording of a reliable consensus. To overcome the identified limitations, a three-tier architecture is proposed, the core of which is the synergy of the Threshold Signature Scheme (TSS) cryptographic protocol and the robust TWTMOM aggregation algorithm. The proposed model completely abandons centralized vulnerable gateways and utilizes Off-chain aggregation based on Shamir's Secret Sharing polynomial distribution and Feldman's Verifiable Secret Sharing scheme over the secp256k1 elliptic curve. Primary hardware noise filtering is performed using the Mahalanobis distance quadratic form. To confirm the effectiveness of the mathematical model, large-scale empirical modeling using the Monte Carlo method was conducted in a simulated highly competitive environment of 1000 nodes. The experimental results conclusively proved that delegating the verification of the aggregated threshold signature to the standard ecrecover precompile in the EVM virtual machine completely eliminates resource-intensive On-chain cyclic structures, algorithmically compressing thousands of disparate transactions into a single one with a constant computational cost of O(1). The obtained error density distribution graphs confirm that through the application of spatial grouping methods (Median-of-Means) and Markovian trust weighting, the system successfully maintains an asymptotic stability plateau and minimal consensus error dispersion even under simultaneous severe Byzantine compromise of up to 49% of the sensors in the network. Furthermore, the developed model's capacity for immune self-cleaning during prolonged stealth gradient attacks has been mathematically and practically proven due to the integration of the exponential reputation decay mechanism for malicious nodes. Summarizing, the developed architecture forms a stable and financially viable scientific and technological foundation for the secure implementation of Web3 technologies into high-load decentralized critical infrastructure systems, where uncompromised data integrity is the absolute priority.

Keywords: Internet of Things, blockchain, smart contracts, threshold cryptography, TSS, Byzantine fault tolerance, Edge computing, TWTMOM, ECDSA.