François Mériaux

Stochastic Differential Games and Energy-Efficient Power Control

One of the contributions of this work is to formulate the problem of energy-efficient power control in multiple access channels (namely, channels which comprise several transmitters and one receiver) as a stochastic differential game. The players are the transmitters who adapt their power level to the quality of their time-varying link with the receiver, their battery level, and the strategy updates of the others. The proposed model not only allows one to take into account long-term strategic interactions, but also long-term energy constraints. A simple sufficient condition for the existence of a Nash equilibrium in this game is provided and shown to be verified in a typical scenario. As the uniqueness and determination of equilibria are difficult issues in general, especially when the number of players goes large, we move to two special cases: the single player case which gives us some useful insights of practical interest and allows one to make connections with the case of large number of players. The latter case is treated with a mean-field game approach for which reasonable sufficient conditions for convergence and uniqueness are provided. Remarkably, this recent approach for large system analysis shows how scalability can be dealt with in large games and only relies on the individual state information assumption.

Predicting a user's next cell with supervised learning based on channel states

Knowing a users next cell allows more efficient resource allocation and enables new location-aware services. To anticipate the cell a user will hand-over to, we introduce a new machine learning based prediction system. Therein, we formulate the prediction as a classification problem based on information that is readily available in cellular networks. Using only Channel State Information (CSI) and handover history, we perform classification by embedding Support Vector Machines (SVMs) into an efficient pre-processing structure. Simulation results from a Manhattan Grid scenario and from a realistic radio map of downtown Frankfurt show that our system provides timely prediction at high accuracy.

Mean field energy games in wireless networks

This work tackles the problem of energy-efficient distributed power control in wireless networks with a large number of transmitters. The problem is modeled by a dynamic game. Each transmitter-receiver communication is characterized by a state given by the available energy and/or the individual channel state and whose evolution is governed by certain dynamics. Since equilibrium analysis in such a (stochastic) game is generally difficult and even impossible, the problem is approximated by exploiting the large system assumption. Under an appropriate exchangeability assumption, the corresponding mean field game is well defined and studied in detail for special cases. The main contribution of this work is to show how mean field games can be applied to the problem under investigation and provide illustrative numerical results. Our results indicate that this approach can lead to significant gains in terms of energy-efficiency at the resulting equilibrium.

Robust IP and UDP-lite header recovery for packetized multimedia transmission

Recently, Joint Source-Channel Decoding (JSCD) techniques have been proposed to improve the reception of multimedia contents transmitted over error-prone channels. These techniques take advantage of the redundancy left by the source coder and of bit reliability measures (soft information) provided by channel decoders to correct transmission errors. To be put at work, protocol stacks have to be made permeable to transmission errors in order to allow soft information to reach the upper protocol layers. For that purpose, headers have to be reliably estimated at each protocol layer. First results have been obtained for lower protocol layers (PHY and MAC) protected by CRCs. The aim of this paper is to extend these results to upper protocol layers (IP and UDP-lite) protected by checksums. As for CRCs, trellis-based decoding techniques may be employed for data protected by checksums. Nevertheless, specific tools have been proposed in this paper to reach a complexity-efficiency trade-off.

View-dependent geometry coding of 3D scenes

A view-dependent geometry coding of 3D scenes defined by sets of semi-regular meshes is presented. The objective is to reduce the quantity of significant data to store when visualizing static 3D scenes from a specific point of view. The proposed coding scheme combines a segmentation for determining the visible regions, and an allocation process for improving the visual quality of the encoded scene.

Achievability of Efficient Satisfaction Equilibria in Self-Configuring Networks

In this paper, a behavioral rule that allows radio devices to achieve an efficient satisfaction equilibrium (ESE) in fully decentralized self-configuring networks (DSCNs) is presented. The relevance of ESE in the context of DSCNs is that at such state, radio devices adopt a transmission/receive configuration such that they are able to simultaneously satisfy their individual quality-of-service (QoS) constraints. An ESE is also an efficient network configuration, i.e., individual QoS satisfaction is achieved by investing the lowest possible effort. Here, the notion of effort refers to a preference each radio device independently establishes among its own set of actions. In particular, the proposed behavioral rule requires less information than existing rules, as in the case of the classical best response dynamics and its variants. Sufficient conditions for convergence are presented in a general framework. Numerical results are provided in the context of a particular uplink power control scenario, and convergence from any initial action profile to an ESE is formally proved in this scenario. This property ensures the proposed rule to be robust to the dynamic arrival or departure of radio devices in the network.

Long-term energy constraints and power control in cognitive radio networks

When a long-term energy constraint is imposed to a transmitter, the average energy-efficiency of a transmitter is, in general, not maximized by always transmitting. In a cognitive radio context, this means that a secondary link can re-exploit the non-used time-slots. In the case where the secondary link is imposed to generate no interference on the primary link, a relevant issue is therefore to know the fraction of time-slots available to the secondary transmitter, depending on the system parameters. On the other hand, if the secondary transmitter is modeled as a selfish and free player choosing its power control policy to maximize its average energy-efficiency, resulting primary and secondary signals are not necessarily orthogonal and studying the corresponding Stackelberg game is relevant to know the outcome of this interactive situation in terms of power control policies.

A stochastic game formulation of energy-efficient power control: Equilibrium utilities and practical strategies

Frequency non-selective time-selective multiple access channels in which transmitters can freely choose their power control policy are considered. The individual objective of the transmitters is to maximize their averaged energy-efficiency. For this purpose, a transmitter has to choose a power control policy that is, a sequence of power levels adapted to the channel variations. This problem can be formulated as a stochastic game with discounting for which there exists a theorem characterizing all the equilibrium utilities (equilibrium utility region). As in its general formulation, this theorem relies on global channel state information (CSI), it is shown that some points of the utility region can be reached with individual CSI. Interestingly, time-sharing based solutions, which are usually considered for centralized policies, appear to be part of the equilibrium solutions. This analysis is illustrated by numerical results providing further insights to the problem under investigation.