Traffic prediction for Internet of Things through support vector regression model

Abstract

With the development of the Internet of Things (IoT), the traffic composition in the network has changed greatly. The traffic analysis is the basis for the further tasks in IoT network, such as intrusion detection, abnormal behavior analysis and attack detection. This paper adopts support vector regression (SVR) to predict traffic data in the wireless sensor networks and IoT network. First, the traffic data is represented as the time series form. Then, the sequence of traffic data is processed by logarithmic function to eliminate the fluctuation of the traffic data. Lastly, the processed traffic sequence data is used to learn a SVR model. The learnt SVR model is used to predict the traffic in the future. The experiments on telemedicine, smart agriculture, vending and automatic driving show that the mean square error of proposed traffic prediction method can achieve less than 0.150.

A Hybrid Deep Learning Method Based on CEEMDAN and Attention Mechanism for Network Traffic Prediction

“…In order to capture the complex features of network traffic sequences, researchers use Multilayer perceptron (MLP), Stacked Autoencoder (SAE), Support Vector Machine (SVR) and RNN to predict network traffic. RNNs have gradually become more popular as they have been shown to better capture the long-term dependence of time sequence [24], [25]. With the continuous research, numerous variants of RNN have emerged, among which LSTM and GRU have gradually become mainstream methods due to their ability to better overcome the problems of gradient disappearance and explosion during training.…”

Federated Learning based Spatio-Temporal framework for real-time traffic prediction

“…The model outperforms the baseline models. Chen, X et al use support vector regression (SVR) for predicting traffic parameters in the WSN and IoT networks [15]. Researchers use time-series data processed by logarithmic function to remove any fluctuations in traffic data.…”

A Hybrid Approach by CEEMDAN-Improved PSO-LSTM Model for Network Traffic Prediction

“…erefore, more and more researchers use nonparametric models to predict network traffic data. e Support Vector Regression model (SVR) and its variant MK-SVR are first used to predict network traffic [17][18][19], which effectively predicts the changing trend of network traffic data but lacks the consideration of temporal correlation of time series data leading to a limit of prediction accuracy.…”