International audienceWe propose a SDN-based architecture for distant points in- terconnection through an overlay network. The main goal of the overlay network is to provide with resilient and high- performance interconnection between its nodes, without the need of changing nor handling Internet routers. Our pro- posed architecture benets from the advantages inherited from SDN such as simplicity of management and ow-level Trac Engineering capabilities. In particular, our approach allows to build the overlay network without any tunnelling technology, which promises to provide a gain in terms of performance, and to ease deployment and management. In addition, we address one of the challenges of SDN, by dis- cussing a possible approach for active monitoring in the pro- posed SDN-overlay architecture

No abstract

Internet traffic classification is an essential task for managing large networks. Network design, routing optimization, quality of service management, anomaly and intrusion detection tasks can be improved with a good knowledge of the traffic.Traditional classification methods based on transport port analysis have become inappropriate for modern applications. Payload based analysis using pattern searching have privacy concerns and are usually slow and expensive in computational cost.In recent years, traffic classification based on the statistical properties of flows has become a relevant topic. In this work we analyze the size of the firsts packets on both directions of a flow as a relevant statistical fingerprint. This fingerprint is enough for accurate traffic classification and so can be useful for early traffic identification in real time.This work proposes the use of a supervised machine learning clustering method for traffic classification based on Support Vector Machines. We compare our method accuracy with a more classical centroid based approach, obtaining promising results.

No abstract

The spread of wireless communications has led electronic equipment to include one or more wireless interfaces. If it were possible to use some generic hardware and implement those different interfaces in software, design and cost would be drastically reduced. This is the idea behind Software Defined Radio (SDR). Cognitive radio attempts to make a more efficient use of a crowded specturm by using licensed bands when licensees are idle. SDR and cognitive radio triggered the research work reported in this paper: to provide a fully functional SDR-based wireless network, developed on top of an open source framework, and able to provide data link and upper layers functionalities. This framework was called GNUWiNetwork, is licensed under GNU, and can be extended by others. Prototyping balanced by architectural design led this framework to be structured as a library of independent blocks which implement different functionalities. These blocks can be interconnected to implement existing or experimental protocols. Blocks interchange different types of events, and act accordingly. A graphical user interface makes it easy to put the blocks together, set the required parameters, and make the set run either in simulation mode or through some simple wireless device controlled by GNU Radio. Several types of blocks have been coded and tried in simulation mode. Some real wireless connection were established between two computers equiped with USRP devices, over which remote sessions such as SSH were successfully established. Work done and the results obtained so far show this framework very valuable for teaching and experimenting, but also promising for industrial prototype testing.