Distributed Blockchain Wireless Sensor Network Architecture for Malicious Node and Sensor Data Detection

Abstract

Wireless Sensor Network (WSN) is a private network with many sensors. It has the ability acquire sensor data with reduced energy consumption but considered vulnerable, primarily due to centralization. Therefore, disruptive blockchain technology capable of hashing blockbased security was recommended to overcome the problem. This research proposes an appropriate architectural model for integrating WSN and distributed blockchain networks to detect suspicious sensor nodes and data. The model is implemented using WSN hardware with a mesh topology and integrated software components, including a blockchain system. The results showed that with 10–500 block node iterations and three distributed Peer servers, the Scrypt hash had the fastest execution time of 469.49 seconds compared to the others. The combination of two consensus mechanism methods, including smart contract (Mac Address and SHA- 256 on the microcontroller) and Proof of Information (PoI) to the three sensor parameters (Temperature or Temp, Humidity or Humd, and gas concentration or MQ2) led to detection and maintenance of sensor nodes and data reaching 97.37% validity. Moreover, sensor control consisting of two Radio Frequency Identification (RFID) nodes and one Quick Response (QR) code node applied to 500 block node iterations led to 592.47 seconds execution time longer than the monitoring sensors. This showed that the separation of monitoring and controlling sensors was the best for the architecture model proposed. The major reason was that the monitoring sensors prioritized automation, speed, and data accuracy in real time. Meanwhile, the controlling sensors focused on the interaction between users and devices, specifically when integrated into a transaction system.

Access to Document

Other files and links

• Link to publication in Scopus
• Open Access Version Available

Fingerprint