CloudCap (C2app) : A Cloud-Based Platform for Packet Analysis On The Edge
Kyriazis Kokkinos, Ioannis Polymenidis, Ilias Siniosoglou, Athanasios Liatifis, Panagiotis Sarigiannidis
TL;DR
CloudCap presents a cloud-assisted platform for mobile network traffic analysis, enabling smartphones to capture data and offload heavy processing to a cloud Analysis Engine. The design features a two-component architecture, multiple capture methods (root tcpdump and LocalVPN), and server-side network-flow translation to support scalable, flow-based analysis and visualization on mobile devices. Empirical evaluation shows modest on-device resource usage and responsive cloud-assisted processing, enabling real-time visualizations and on-demand archives. The work demonstrates practical viability for edge-enabled traffic monitoring and points to future enhancements with MEC deployment and machine-learning-based anomaly detection.
Abstract
Data exchange through mobile devices is rapidly increasing due to the high information demands of today's applications. The need for monitoring the exchanged traffic becomes important in order to control and optimize the device and network performance and security. Taking this under consideration, in this paper, we developed a cloud-based system for the analysis of network traffic. The smartphone devices act both as traffic captors and visualization endpoints, enabling the user to get an overview of the network while minimizing resource consumption. In the presented work, we evaluate our system using two test cases and a variety of target devices. Our results prove the usefulness of the proposed system architecture.