Consensus Mechanism of IoT Based on Blockchain Technology

Abstract

Applying blockchain technology to the Internet of Things (IoT) remains a huge challenge. To meet the actual needs of IoT, a lightweight and high-throughput consensus mechanism, combined with blockchain technology, is proposed in this study. Blockchain nodes use the Diffie–Hellman algorithm for key negotiation. Sensors and blockchain nodes can use the shared key to generate HMAC (Hash-based Message Authentication Code) signatures for sensor-aware transactions and use the Verifiable Random Function to implement block nodes. Offline fast election, which is the node that wins the election, becomes the block node. Machine learning methods are also introduced to identify or remove outliers in the sensor data before such data are uploaded to the chain. Experimental results show that the system throughput synchronously increases as the test load increases. Moreover, when the test load is 800 tps, the system throughput reaches the maximum, close to 600 tps. When the test load exceeds 800 tps, the actual system throughput starts to drop, and approximately 90% of transactions have a delay time within 5000 ms. This method can be used in a lightweight IoT system.