An AI-Based Adaptive Cognitive Modeling and Measurement Method of Network Traffic for EIS

Abstract

Enterprise Information System (EIS) is based on Internet of things (IoT) and aggregates a large amount of data of companies. Real-time reliable data transmission and data processing are very important for EIS. Network traffic of IoT is very important for network management and traffic planning in EIS. However, the measurement overheads and measurement accuracy are a contradiction for the fine-grained traffic measurement requirements. Artificial Intelligence (AI) has long promised to learn the natural feature of network traffic and make some actions about the prediction of traffic. In this paper, we propose an AI-based Lightweight Adaptive Measurement Method (ALAMM) for SDN to reduce the traffic measurement overheads and improve the measurement accuracy. Firstly, we use the AI to model and predict the flow traffic in the network. Based on the traffic prediction results, we propose an adaptive method to decide the traffic sampling frequency. Secondly, we send the measurement primitives to switches to measure the coarse-grained traffic of flows and links. Finally, the matrix filling and optimization method are proposed to recovery the fine-grained measurement and optimize the measurement result. Simulation results show that our approach can obtain network traffic with low overhead and high accuracy.