Lightweight Consensus for IoT Networks Using Adaptive Proof-of-Probability Models

Abstract

IoT networks require secure coordination but cannot tolerate the heavy computational and energy burden of mainstream blockchain consensus mechanisms. This paper introduces an adaptive Proof-of-Probability (PoP) model designed for ultra-low-power devices. Unlike proof-of-work or stake-based models, PoP assigns block proposal probability based on device reliability, historical behavior, and real-time trust signals. Each node maintains a local trust vector updated through lightweight observations such as uptime, packet integrity, and peer confirmation. We design a probabilistic leader election protocol that minimizes message overhead and supports rapid convergence. Simulations across 10,000-node IoT clusters show PoP reduces energy consumption by 65–78% compared to PoS-lite variants, while maintaining strong resilience against Sybil and eclipse attacks. We also evaluate a real hardware deployment using ESP32 devices to measure runtime impact. Results show near-linear scalability. The paper concludes with security proofs and guidelines for practical deployments.