Video Streaming Service Identification on Software-Defined Networking

It is well known that video streaming is the major network traffic today. Futhermore, the traffic generated by video streaming is expected to increase exponentially. On the other hand, SoftwareDefined Networking (SDN) has been considered a viable solution to cope with the complexity and increasing network traffic due to its centralised control and programmability features. These features, however, do not guarantee that network performance will not suffer as traffic grows. As result, understanding video traffic and optimising video traffic can aid in control various aspects of network performance, such as bandwidth utilisation, dynamic routing, and Quality of Service (QoS). This paper presents an approach to identify video streaming traffic in SDN and investigates the feasibility of using Knowledge-Defined Networking (KDN) in traffic classification. KDN is a networking paradigm that takes advantage of Artificial Intelligence (AI) by using Machine Learning approaches, which allows integrating behavioural models to detect patterns, like video streaming traffic identification, in SDN traffic. In our initial proof-of-concept, we derive the relevant information of network traffic in the form of flows statistics. Then, we used such information to train six ML models that can classify network traffic into three types, Video on Demand (VoD), Livestream, and no-video traffic. Our proof-of-concept demonstrates that our approach is applicable and that we can identify and classify video streaming traffic with 97.5% accuracy using the Decision Tree model.

TrafficStream: A Streaming Traffic Flow Forecasting Framework Based on Graph Neural Networks and Continual Learning

With the rapid growth of traffic sensors deployed, a massive amount of traffic flow data are collected, revealing the long-term evolution of traffic flows and the gradual expansion of traffic networks. How to accurately forecasting these traffic flow attracts the attention of researchers as it is of great significance for improving the efficiency of transportation systems. However, existing methods mainly focus on the spatial-temporal correlation of static networks, leaving the problem of efficiently learning models on networks with expansion and evolving patterns less studied. To tackle this problem, we propose a Streaming Traffic Flow Forecasting Framework, TrafficStream, based on Graph Neural Networks (GNNs) and Continual Learning (CL), achieving accurate predictions and high efficiency. Firstly, we design a traffic pattern fusion method, cleverly integrating the new patterns that emerged during the long-term period into the model. A JS-divergence-based algorithm is proposed to mine new traffic patterns. Secondly, we introduce CL to consolidate the knowledge learned previously and transfer them to the current model. Specifically, we adopt two strategies: historical data replay and parameter smoothing. We construct a streaming traffic data set to verify the efficiency and effectiveness of our model. Extensive experiments demonstrate its excellent potential to extract traffic patterns with high efficiency on long-term streaming network scene. The source code is available at https://github.com/AprLie/TrafficStream.

Machine Learning based KPI Monitoring of Video Streaming Traffic for QoE Estimation

Quality of Experience (QoE) monitoring of video streaming traffic is crucial task for service providers. Nowadays it is challenging due to the increased usage of end-to-end encryption. In order to overcome this issue, machine learning (ML) approaches for QoE monitoring have gained popularity in the recent years. This work proposes a framework which includes a machine learning pipeline that can be used for detecting key QoE related events such as buffering events and video resolution changes for ongoing YouTube video streaming sessions in real-time. For this purpose, a ML model has been trained using YouTube streaming traffic collected from Android devices. Later on, the trained ML model is deployed in the framework's pipeline to make online predictions. The ML model uses statistical traffic information observed from the network-layer for learning and predicting the video QoE related events. It reaches 88% overall testing accuracy for predicting the video events. Although our work is yet at an early stage, the application of the ML model for online detection and prediction of video events yields quite promising results.

Preventing the Video Leakages from The Traffic Streaming

Video streaming takes up an increasing proportion of network traffic nowadays. Dynamic Adaptive Streaming over HTTP (DASH) becomes the defacto standard of video streaming and it is adopted by YouTube, Netflix, etc.Despite of the popularity, network traffic during video streaming shows identifiable pattern which brings threat to user privacy.In this paper, to proposea video identification method using network traffic while streaming. Though there is bitrate adaptation in DASH streaming, we observe that the video bit rate trend remains relatively stable because of the widely used Variable Bit-Rate(VBR) encoding. Accordingly, we design a robust video feature extraction method for eavesdropped video streaming traffic. Meanwhile, we design a VBR based video fingerprinting method for candidate video set which can be built using downloaded video files. Finally, to propose an efficient partial matching method for computing similarities between video fingerprints and streaming traces to derive video identities. To evaluate our attack method in different scenarios for various video content, segment lengths and quality levels. The experimental results show that the identification accuracy can reach up to 90%using only three minute continuous network traffic eavesdropping.

Visual Vertigo, Motion Sickness, and Disorientation in Vehicles

Environmental circumstances that result in ambiguity or conflict with the patterns of sensory stimulation may adversely affect the vestibular system. The effect of this conflict in sensory information may be dizziness, a sense of imbalance, nausea, and motion sickness sometimes even to seemingly minor daily head movement activities. In some, it is not only exposure to motion but also the observation of objects in motion around them such as in supermarket aisles or other places with visual commotion; this can lead to dizziness, nausea, or a feeling of motion sickness that is referred to as visual vertigo. All people with normal vestibular function can be made to experience motion sickness, although individual susceptibility varies widely and is at least partially heritable. Motorists learn to interpret sensory stimuli in the context of the car stabilized by its suspension and guided by steering. A type of motorist's disorientation occurs in some individuals who develop a heightened awareness of perceptions of motion in the automobile that makes them feel as though they may be rolling over on corners and as though they are veering on open highways or in streaming traffic. This article discusses the putative mechanisms, consequences and approach to managing patients with visual vertigo, motion sickness, and motorist's disorientation syndrome in the context of chronic dizziness and motion sensitivity.