Study on asynchronous processing of IoT big data-based homomorphic encryption transmission using Raft algorithm

IoT server-layer processing of heterogeneous data alleviates silos but incurs high resource costs. This study proposes shifting computational load to the edge layer to reduce server strain. Focusing on massive IoT deployments in non-ideal environments, it addresses challenges in connectivity, data detection, and channel inference, where traditional frequency offset compensation and synchronization techniques fall short. The research posits that terminals require downlink information transmission for coverage. Combining this with big data demands and data rate considerations, we construct a homomorphic encryption transmission model for asynchronous IoT data transfer. To minimize data rate impacts on reconstruction and performance, a methodology integrating asynchronous data processing and channel inference prediction is developed. We design a multi-node detection and channel inference optimization scheme, merging the Raft algorithm-based homomorphic encryption with convex optimization. Simulation tests compare bit error rate, channel inference quality, and performance against conventional Lasso and OMP (with serial interference cancellation) algorithms. The results verify that the proposed Raft-based approach achieves performance gains of 7.52% over Lasso and 10.64% over OMP, demonstrating significant technical superiority. This innovative application of homomorphic encryption for asynchronous big data transmission advances IoT processing and holds potential for enabling blockchain integration across diverse IoT systems. By comparison with the traditional Byzantine fault tolerance consensus scheme, Raft has more advantages in throughput, delay, and resource consumption and can meet the requirements for efficiency in IoT big data processing while ensuring data transmission security and storage consistency.