Abstract: Wireless sensor networks (WSN) are used for real-time monitoring, but they are also susceptible to cyber-physical system (CPS) attacks, such as tampering and unauthorised access. Conventional privacy and security systems face scalability and resource constraints. To address these challenges, this study proposes a Blockchain-Assisted Random Forest–LSTM (BA-RF-LSTM) model for improving the security and reliability of WSNs. The BA-RF-LSTM uses blockchain (BC) for secure, distributed data transfer. Using lightweight consensus and immutable ledgers, it maintains trust, integrity, and confidentiality. Smart contracts (SC) ensure accurate data transfer and compliance with consensus protocols. BC supports decentralized authentication, data security, and logging. This study developed a simulation-based CPS to investigate secure data transmission in resource-constrained WSN environments. Under the evaluated simulation configuration, the model achieved a mean network throughput of 67.34 kbps and an end-to-end delay of 232 ms. The simulation results indicate that the proposed method can reduce network overhead and data-tampering risks while improving secure data transmission. The proposed model achieved a mean data throughput of 67.34 kbps and an end-to-end network delay of 232 ms under the evaluated configuration. The experimental results show that the proposed method optimises WSN energy consumption, minimises data tampering, and enhances data security.
Keywords: Blockchain; Data transmission; Authentication; WSN; Security; Authorisation
DOI: 10.24874/PES08.03.025
Recieved: 14.05.2026 Revised: 29.06.2026 Accepted: 05.08.2026
UDC:
Reads: 13 