Overall Objectives
Scientific Foundations
Application Domains
New Results
Contracts and Grants with Industry
Other Grants and Activities

Section: New Results

Radio Link

Body Area Networks (BANETs):

It is expected that Body Area Networks will be deployed in various applications to instrument and support humans in many aspects of their life. Different standards actually compete to fulfill the constraints associated with possible applications. Among all possible applications, we focus in [49] , [23] on multi-nodes monitoring applications. The body is thus equipped with a set of sensors transmitting in real-time their measures to a common sink. The underlying network topology is thus classical in the field of wireless sensor networks (WSN) and referred to as the star topology. The classical super-frame structure as proposed in 802.15.3 seems to comply with the needs of such an application. However, the specificities of the BAN channel can reduce the performance of such protocol. Indeed, the channel time variations make the star structure unstable and subject to a high level of packet loss. We investigate some cooperative mechanisms to address this problem and we evaluate their implementation in the superframe structure.

Multi-hop transmission with unreliable links:

During the last decade, many works were devoted to improving the performance of relaying techniques in ad hoc networks. One promising approach consists in allowing the relay nodes to cooperate, thus using spatial diversity to increase the capacity of the system. However, this approach introduces an overhead in terms of information exchange, increasing the complexity of the receivers. A simpler way of exploiting spatial diversity is referred to as opportunistic routing. In this scheme, a cluster of nodes still serves as relay candidates but only a single node in the cluster forwards the packet. [39] proposes a thorough analysis of opportunistic routing efficiency under different realistic radio channel conditions. The study aims at finding the best trade-off between two objectives: energy and latency minimizations, under a hard reliability constraint. We derive an optimal bound, namely, the Pareto front of the related optimization problem, which offers a good insight into the benefits of opportunistic routing compared with classical multi- hop routing. In [50] , we emphasize the problem of energy consumption only. We put forward an analytical framework to optimize the opportunistic communication scheme in order to minimize the energy consumption. Meanwhile, the optimizations of node density and the transmission power are analyzed for different ranges of relay candidates in Rayleigh block fading and Additive White Gaussian Noise (AWGN) channels. In the energy consumption viewpoint, the analysis indicate that the opportunistic communication is more efficient in Rayleigh block fading channel than that in AWGN channel. These results were obtained during the thesis of Ruifeng Zhang [3]

[25] , [38] address the problem of reliable transmission of sensed data through a vast field of small and vulnerable sensors towards a sink node. We concentrate in this paper on networks deployed rapidly in harsh environments as needed for instance in disaster-relief scenarios. Hence, emphasis has to be put on the minimization of the global energy consumption of the network and on providing both fast data transmissions and a rapid network setup. Therefore, we introduce a new gradient broadcasting routing algorithm for wireless sensor networks, U-GRAB, where the broadcasting decision is taken according to a utility-based policy. This policy favors the broadcasting of packets for nodes that experience non-congested channels and have a satisfactory energy level. Our simulation results show that this new forwarding strategy greatly improves the robustness/energy/delay trade-off of GRAB, the current state-of the art solution in gradient broadcasting techniques.

[11] , [15] studies the impact of fountain codes in the context of wireless sensor networks. This impact has been previously studied for wireless network in which the MAC layer is considered perfect. In [11] the power consumption is considered in the case of realistic MAC layer: the acknowledgment messages used in fountain codes can be lost. We show that fountain codes provide more resiliency to the lost of acknowledgment messages than protocols without coding. Moreover, the use of fountain codes allow to reduce the energetic cost of point to point connection. In [15] , a strategy is proposed to improve the classical decode and forward strategy. From a given set of received packet, a new set of packet is generated respecting the Soliton distribution. This strategy reduces the overall number of messages exchanged. These results are part of the thesis of Anya Apavatjrut.

Performance on wireless links:

Recent years have witnessed a tremendous growth of research in the field of wireless systems and networking protocols. Consequently, simulation has appeared as the most convenient approach for the performance evaluation of such systems and several wireless network simulators have been proposed in recent years. However, the complexity of the wireless physical layer (PHY) induces a clear tradeoff between the accuracy and the scalability of simulators. Thereby, the accuracy of the simulation results varies drastically from one simulator to another. In [4] , we focus on this tradeoff and we investigate the impact of the physical layer modeling accuracy on both the computational cost and the confidence in simulations. We first provide a detailed discussion on physical layer issues, including the radio range, link and interference modeling, and we investigate how they have been handled in existing popular simulators. This work describes the backbone of WSNet.

Distributed consensus in WSN:

Gossip averaging algorithms are used in distributed systems to let the nodes sharing a common value. This year, we introduce an alternative to standard gossip averaging algorithms for wireless sensor networks [29] . The proposed algorithm takes opportunity of the intrinsic broadcasting nature of the wireless medium and copes well with the unreliability of radio links. Indeed, this algorithm works asynchronously and in a distributed manner using one-way exchanges (acknowledgements are not required). It relies on the regularization formalism as used in image processing, and we show that it provides a tunable approximation of distributed averaging. An extension of this approach was proposed in [27] . Sensor networks aim at monitoring events and phenomena occurring in their environment and providing useful information to one or several end users.When a global knowledge of sensed data is needed, techniques from data gathering, statistical estimation and parametric modeling can be used. While the two first methods require respectively a large amount of energy and a knowledge of statistical dependencies between measurements, a new simple algorithm for fitting a parametric model to sensed data was proposed. The novelty of this approach stands in its intrinsic robustness to packet losses and asynchronous data exchanges. Moreover, this algorithm is intuitively efficient as it uses the broadcasting nature of the wireless medium. These contributions were developed in strong cooperation with CEA LETI, and led to the PhD defense of Nicolas Maréchal [2] .

MAC/PHY UWB for WSN design:

This work is done in the context of a Cooperation with Orange Labs. They are the subject of the thesis of Benoit Miscopein. We propose a MAC layer for sensor networks, including the physical layer is capable of UWB impulse. We propose in [33] , [32] a specification and the performance study for a procedure for detecting UWB signal pulse. From the block detection signal, we then designed a protocol MAC based on preamble sampling ([31] ). This type of protocol requires the sending of a preamble before a useful package to wake-up the destination node. However, the length of this preambule can exceed the maximum emission time defined in an UWB equipment, according to European regulation. To circumvent this problem, we propose a protocol that enables collaborative relay of a burst of preambule by neighboring recipient, so that the sum of the relay is sufficiently long for the recipient and to wake-up that each relay node not violating the regulation. This protocol was the object of the deposit of a patent and a talk within the standardization group 802.15.6. We have also studied the background interfering, and we derive a non-coherent receiver optimal in a multi-path and in the presence of interference.


Logo Inria