Angular Histogram-Based Visualisation of Network Traffic Flow Measurement Data

Abstract

Knowledge of the traffic that is being carried within a network is critical for ensuring the network’s smooth operation, and network traffic measurement has provided an effective means to achieve this. However, network traffic volume has substantially increased over the last decades. Combine that with the traffic heterogeneity from a diverse range of new, connected devices, we have reached a point where the response to any outage or anomalous event is simply beyond human ability. Network information visualisation is a useful tool to help network administrators deal with this problem. While this approach is not new, traditional approaches do not scale well with the increasing volume and heterogeneity of network traffic. In this paper, we propose the application of angular histogram visualisation to provide an information-rich overview of large network traffic data sets to improve the interpretation and understanding of network traffic flow measurement data. We evaluate our approach experimentally using live network traffic to demonstrate its efficacy and provide suggestions on how it can be further improved.