Section: New Results

Wireless Networks

Participants : Sara Alouf, Eitan Altman, Giovanni Neglia, Alina Tuholukova.

Control of Delay–Tolerant Networks

In [5] E. Altman and G. Neglia, in collaboration with F. De Pellegrini (Create-Net, Italy) and D. Miorandi (U-Hopper, Italy), study optimal stochastic control of delay tolerant networks. First, the structure of optimal two-hop forwarding policies is derived. In order to be implemented, such policies require knowledge of certain global system parameters such as the number of mobiles or the rate of contacts between mobiles. But, such parameters could be unknown at system design time or may even change over time. In order to address this problem, adaptive policies are designed that combine estimation and control: based on stochastic approximation techniques, such policies are proved to achieve optimal performance in spite of lack of global information. Furthermore, the paper studies interactions that may occur in the presence of several DTNs which compete for the access to a gateway node. The latter problem is formulated as a cost-coupled stochastic game and a unique Nash equilibrium is found. Such equilibrium corresponds to the system configuration in which each DTN adopts the optimal forwarding policy determined for the single network problem.

Performance Evaluation of Train Moving-Block Control

In moving block systems for railway transportation a central controller periodically communicates to the train how far it can safely advance. On-board automatic protection mechanisms stop the train if no message is received during a given time window. In [45], [63] G. Neglia, S. Alouf, and A. Tuholukova in collaboration with A. Dandoush (SME Sudria, France, formerly engineer with Maestro ) and S. Simoens, P. Dersin, J. Billion and P. Derouet (all from Alstom Transport) consider as reference a typical implementation of moving-block control for metro and quantify the rate of spurious Emergency Brakes (EBs), i.e. of train stops due to communication losses and not to an actual risk of collision. Such unexpected EBs can happen at any point on the track and are a major service disturbance.

The general formula for the EB rate found in [45] requires a probabilistic characterization of losses and delays. Calculations are surprisingly simple in the case of homogeneous and independent packet losses. More complex loss scenarios are studied in [59]. The approach is computationally efficient even when emergency brakes are very rare (as they should be) and can no longer be estimated via discrete-event simulations.

The analytical models have also been validated using ns-3 simulations [35].

Speed estimation

After several years of cooperation with Nokia (formerly Alcatel-Lucent) Bell Labs in developing tools for speed estimation from measurement of the radio channel, we have now started to publish our joint patented work. This includes the work on mobility state estimation in LTE by D.-G. Herculea, V. Capdevielle, C. S. Chen, N. Ben Rached and F. Ratovelomanana from Nokia-Bell Labs in collaboration with E. Altman and M. Haddad (UAPV), see [38].

Sonorous cartography for sighted and blind people

E. Altman has been invited by D. Josselin from UMR Espace in UAPV to co-advise a Master project and later a thesis financed by the CNRS on Sonorous cartography. Other persons with whom we collaborate are D. Roussel, S. Boularouk, A. Saidi, M. Driss (from UAPV) and O. Bonin (Laboratoire Ville, Mobilité, Transport) all coauthors of [40] which won the best short paper award in the AGILE conference. In this article, we test the usability of a cartographic tool mixing maps and sounds. This tool is developed within QuantumGIS as a plugin prototype. We first present some theoretical reflections about synesthesia. Secondly, we explain the way we “sonificate” the images, by associating colors and recorded chords and sounds. Then we present the results of several usability tests in France with different users, including blind people.

To help blind people compensate visual perception and to better understand their outdoor environment, S. Boularouk and D. Josselin from UAPV in collaboration with E. Altman, proposed in [49] a method using human-computer interaction via Text-to-Speach. It helps visually impaired people to know surrounding places from OpenStreetMap data by hearing. The principal idea is to convey spatial information by voice synthesis and receive requests from blind people by voice recognition.

Scheduling for mobile users with non-stationary mobility

H. Zaaraoui and Z. Altman from Orange Labs in collaboration with T. Jiménez (UAPV) and E. Altman have studied scheduling in an environment with non-stationary mobility (cars are moving on a road and may have to stop at red lights). They propose scheduling schemes for such mobility patterns and study their performance in in [55] and in [48].

User Association in Multi-user MIMO Small Cell Networks

Dense Networks and large MIMO are two key enablers to achieve high data rates towards next generation 5G networks. In this context, S. Ramanath (Lekha Wireless Solutions and IIT Mumbai) and M. Debbah (Huawei) in collaboration with E. Altman study in [47] user association in an interference limited Multiuser MIMO Small Cell Network. Extending on previous findings, they derive explicit expressions for the optimal ratio of the number of antennas at the base station to the number of users that can associate to a base station in such a Network. The expressions are used to compute the actual number of users that can associate for a given interference level and other system parameters. Simulation results and numerical examples are provided to support our theoretical findings.