Section: New Results

Content-Oriented Systems

Participants : Sara Alouf, Eitan Altman, Konstantin Avrachenkov, Alain Jean-Marie, Philippe Nain, Giovanni Neglia.

Sara Alouf and Nicaise Choungmo Fofack (former PhD student at Maestro , currently at Ingima) have thoroughly revised their study of the modern behavior of DNS caches. In particular the closure properties of the class of distributions called diagonal matrix-exponential are fully derived, hence the analytic models presented in [78] to study tree of caches with general caching durations are extended to the case of polytrees [15] .

In previous years, Alain Jean-Marie and collaborators from the Univ. Montpellier have defined a family of combinatorial designs that minimize the variance in the availability of replicated documents in unreliable infrastructures. Then with Jean-Claude Bermond (Cnrs , with the Inria project-team Coati ), Dorian Mazauric (now with Inria project-team ABS ) and Joseph Yu (UFV Vancouver), it was shown that well-balanced families solve the problem, and such families were constructed for small numbers of replicas. This work is now published in [21] . During the internship of Mikhail Grigorev, several methods for generating at random good solutions have been investigated.

Data offloading from the cellular network to lowcost WiFi has been the subject of several research works in the last years. In-network caching has also been studied as an efficient means to further reduce cellular network traffic. In [49] M. El Chamie (Univ. of Washington, USA), C. Barakat (Inria project-team Diana ) and G. Neglia consider a scenario where mobile users can download popular contents (e.g., maps of a city, shopping information, social media, etc.) from WiFi-enabled caches deployed in an urban area. They study the optimal distribution of contents among the caches (i.e., what contents to put in each cache) to minimize users' access cost in the whole network, and argue that this optimal distribution does not necessarily provide geographic fairness, i.e., users at different locations can experience highly variable performance. In order to mitigate this problem, they propose two different cache coordination algorithms based on gossiping. These algorithms achieve geographic fairness while preserving the minimum access cost for end users.

In [43] K. Avrachenkov in collaboration with V.S. Borkar (IIT Mumbai, India) consider the task of scheduling a crawler to retrieve from several sites their ephemeral content. This is content, such as news or posts at social network groups, for which a user typically loses interest after some days or hours. Thus development of a timely crawling policy for ephemeral information sources is very important. The authors first formulate this problem as an optimal control problem with average reward. The reward can be measured in terms of the number of clicks or relevant search requests. The problem in its exact formulation suffers from the curse of dimensionality and quickly becomes intractable even with moderate number of information sources. Fortunately, this problem admits a Whittle index, a celebrated heuristics which leads to problem decomposition and to a very simple and efficient crawling policy. The Whittle index is derived, together with its theoretical justification.