New fog computing enabled lossless EEG data compression scheme in IoT networks

Abstract

The rapid evolution of the Internet of Things (IoT) leads to emerging several applications such as smart home, ehealth, smart agriculture and farming, etc. The edge of the IoT networks will receive a huge amount of data generated by these smart applications which regularly require to be delivered to the servers of the remote data centers for additional real-time treatment. However, transmitting the whole of these IoT data across the network toward the data center of the cloud will impose a high overhead on the IoT network. The data exchange and processing long delay have a significant impact on the response speed of the real-time IoT applications. It can reduce the time responsiveness of these applications. The fog computing has provided for the IoT applications as an intermediate layer between the smart end devices and the cloud to reduce IoT data and improve the response time for IoT applications. In this paper, new fog computing enabled lossless EEG data compression scheme is proposed to minimize the amount of IoT EEG data uploaded to the cloud. The EEG data compression scheme consists of two phases: clustering and compression. First, the received data is clustered into clusters using agglomerative hierarchical clustering. The Huffman encoding is applied to each resulted cluster in the second phase. Finally, the compressed files of smaller clusters are combined and uploaded from fog to cloud. Several experiments have been implemented and the comparison results show that the suggested combined Hierarchical Clustering and Huffman Encoding (HCHE) method significantly decreases the volume of EEG data uploaded to the cloud platform. The proposed HCHE method achieves 4.33 of average compression power that is more than twice as much the compression power of some existing methods.