Distributed multi-party security computation framework for heterogeneous internet of things (IoT) devices

Abstract

The heterogeneous nature of the Internet of Things (IoT) requires sophisticated and distributed security schemes for providing reliable communication. However, providing a unanimous security measure is a tedious task for heterogeneous devices is a complex task. This function requires multiple communication and information sharing constraints to be adapted by connected users. In this article, a distributed multi-party security computation framework for ensuring the security level of scalable information sharing and communications in heterogeneous IoT. This computation framework uses synchronized security measures through recurrent verification on different communicating ends of the devices. In this verification process, backpropagation learning verifies end-to-end security and the communication session between the connected devices. MPC is generally subject to a significant overhead for correspondence but has the benefit of being confidential even if pursued until a substantial percentage of the participants are aggressive and coordinated. Secret MPC exchanging will protect the specifications of both models and the training/inference information. The multi-party verification relies on both device and service providers for unanimously synchronizing the different security computations in a less complex manner. This framework is reliable in mitigating man-in-middle and information loss adversaries. The performance of the proposed method is verified using the metrics detection ratio, response rate, delay, and overhead.