Section: New Results

Quality-of-Service

We therefore propose SACS, a caching strategy designed for the CCN architecture that includes social information [11] . CCN is to date the most widely adopted ICN architecture by the research and industrial community. The underlying idea in such strategy is that a small number of users counts a huge amount of social relationships, dominates the activity and receives most attention from other users. We call such users Influential users, and we argue that they produce content that is more likely to be consumed by others, and in consequence their content must be favored and replicated in priority. Our novel caching strategy is therefore prioritizing content from Influential users of the social network. To validate our strategy, we first propose a model of social network over the CCN architecture [30] . Our model has been designed based on the measurement of Pinterest, a web-based OSN system. Extensive simulations of the strategy have been performed, as well as a real implementation on CCNx and deployment over the PlanetLab testbed. Our results with SACS are significant and increase drastically the caching performance of ICN architecture. content

Efficient management of caches is a key success factor in Concent-Centric Networks where multiple (up to every single node in the network) entities act as caches of the shared content in the network. We pursued our investigations towards a common evaluation framework for cache strategies in Content-centric networks and towards the definition of novel cache strategies, exploiting context information available at the service level of today's internet.

Probabilistic end-to-end performance guarantee may be required when dealing with real-time applications. As part of ANR QUASIMODO project, we are dealing with Markov modeling of multi-hop networks running duty-cycled MAC protocols. One of the problems of the existing Markovian models resides in their strong assumptions that may not be directly used to assess the end-to-end delay in practice. In particular, realistic radio channel, capture effect and OS-related implementation factors are not taken into account. We proposed to explore a new approach combining code instrumentation and Markov chain analysis. In [15] we have presented a new approach for extracting empirical Markov chain models from network protocol traces by means of Process Mining techniques. An empirical Markov chain model was obtained for the IEEE 802.15.4 beacon-enabled mode protocol allowing us to estimate the e2e delay for a multi-hop scenario. This approach has also been successfully applied to the case of ContikiMAC [14] .

This year we have focused on implementing and evaluating the used of SDN for managing the QoS in broadband access networks. Unfortunately, application-based QoS on a home network gateway faces significant constraints, as commodity home routers are not typically powerful enough to perform application classification, and many home users are not savvy enough to configure QoS parameters. In [24] we designed FlowQoS, an SDN-based approach where users can specify upstream and downstream bandwidth allocations for different applications at a high level, offloading application identification to an SDN controller that dynamically installs traffic shaping rules for application flows We designed a custom DNS-based classifier to identify different applications that run over common web ports; a second classifier performs lightweight packet inspection to classify non-HTTP traffic flows. We implemented FlowQoS on OpenWrt and demonstrated that it can improve the performance of both adaptive video streaming and VoIP in the presence of active competing traffic.

This work has been carried out as part of a co-supervised PhD thesis between University of Lorraine and SupCom Tunis.

QoS must be guaranteed when dealing with real-time distributed systems interconnected by a network. Not only task schedulability in processors, but also message schedulability in networks should be analysed for validating the system design. In [37] , [36] , [34] , and [35] , we provided an overview of both message scheduling techniques in networks and joint task and message scheduling approaches in closed-loop distributed control systems (networked control systems). Fault-tolerance is another critical issue that one must take into account. In collaboration with an industrial partner, we started a study on the real-time dependability of UAV multi-criticity system interconnected by an embedded mesh network. The future work aims at developing a robust mesh network routing protocol and studying the schedulability under constraints of multi-criticality and graceful degradation during mode change.