Journal of Networking Technology Ketmaneechairat, H. Oothongsap, P. Mingkhwan, A 17 2 2025 https://doi.org/10.6025/jnt/2026/17/2/63-82 https://www.dline.info/jnt/fulltext/v17n2/jntv17n2_2.pdf The integration of blockchain technology with Internet of Things (IoT) systems presents significant potential for enhancing security, decentralization, and trust. However, this convergence introduces critical challenges including high computational overhead, scalability limitations, latency constraints in 5G-enabled environments, and the absence of unified frameworks that simultaneously address security, trust management, and energy efficiency. This study proposes a multi layer analytical framework for assessing reliability in IoT-blockchain systems using sensor-driven transaction modeling. The methodology integrates statistical correlation analysis via Point-Biserial coefficients, probabilistic modeling through logistic regression, and comprehensive reliability metrics including Transaction Success Rate (TSR) and Mean Transactions to Failure (MTTF). Analysis of an enhanced IoT dataset augmented with blockchain transaction attributes reveals that, while global linear relationships between sensor values and transaction outcomes remain weak (|r| < 0.07), specific sensor types particularly temperature sensors demonstrate statistically significant elevated failure rates (42.9%, OR = 1.54, p = 0.0087). Logistic regression identifies a critical threshold beyond which failure probability increases substantially, while network and transaction graph analyses highlight the importance of understanding interdependencies among system components. The findings emphasize that transaction reliability is driven by localized, context specific factors rather than broad linear relationships, necessitating targeted interventions including sensor specific calibration, oracle optimization, and smart contract edge case handling. This framework provides actionable insights for enhancing robustness in next-generation 5G-enabled IoT blockchain ecosystems.