Thèse Protocoles de Diffusion de Données de Télédétection à Tolérance de Pannes pour le Changement Climatique et le Développement Durable H/F Doctorat.Gouv.Fr

Établissement : Ecole Nationale de l'Aviation Civile École doctorale : SYSTEMES Laboratoire de recherche : ENAC-LAB - Laboratoire de Recherche ENAC Direction de la thèse : Guthemberg DA SILVA SILVESTRE ORCID 0000000202737212 Début de la thèse : 2027-09-01 Date limite de candidature : 2026-11-23T23:59:59 La collecte de données de télédétection environnementale est essentielle pour quantifier et mieux comprendre les impacts du changement global. La collecte de données dans des écosystèmes éloignés, tels que les estuaires de mangroves, les forêts tropicales humides, les habitats en eaux profondes et les cours d'eau, nécessite une couverture étendue et une grande précision spatio-temporelle. Si les systèmes satellitaires permettent une couverture étendue, ils ne permettent pas une collecte de données spatio-temporelles précises. Les essaims de drones sont apparus comme un moyen prometteur d'améliorer la précision de la collecte de données. Cependant, la conception d'essaims de drones en tant que plateformes de télédétection tolérantes aux pannes reste un défi. Ce projet de thèse aborde un enjeu de recherche majeur pour ces plateformes : la diffusion de données tolérante aux pannes. Le projet étudiera et proposera des techniques de diffusion orientées mission afin de permettre une collecte adaptative de données de télédétection, allant de la collecte rapide à faible latence à des applications de télédétection tolérantes aux pannes de partition et aux retards. Effective environmental data collection and dissemination in drone swarms relies heavily on distributed synchronisation mechanisms that allow nodes to exchange state, maintain a consistent view of shared data, and support collective decision making. This requirement is particularly challenging in mobile settings, where communication links are intermittent, network topology changes frequently due to mobility, and available bandwidth and energy resources are limited. Additionally, the CAP Theorem [Gilbert and Lynch 2012] states that a distributed system can simultaneously satisfy at most two of the following properties: consistency, availability, and partition tolerance. In highly dynamic swarm networks, network partitions are frequent and largely unavoidable, which makes it difficult for traditional strongly consistent systems to maintain high availability. As a result, such systems are hard to deploy efficiently in drone swarms, where continuous operation and timely coordination are essential.

Multicast primitives [Janic 2005] are fundamental for information dissemination in networked systems. These primitives allow drones to synchronise their collected
data and to enable fault tolerance through replication. As drones communicate through a sparse, wireless ad hoc network, we assume that state synchronisation messages are exchanged through unreliable channels to reduce synchronisation costs.

Ad-hoc routing protocols maintains reliable connectivity in highly mobile swarm networks. The first paradigm is based on routing, and is prominently represented by the Better Approach To Mobile Ad hoc Networking (batman) protocol [Johnson et al. 2008]. batman is a proactive layer-two routing protocol designed for mobile ad hoc networks, where each node maintains next hop information rather than complete end-to-end paths. By continuously exchanging small originator messages, nodes are able to estimate link quality and select the most reliable neighbor for forwarding traffic. This localized decision process limits routing state complexity and allows batman to react efficiently to topology changes caused by mobility. Comparative studies in Flying Ad hoc Network (FANET) scenarios have evaluated batman alongside other routing protocols under high mobility conditions [Guillen-Perez et al. 2021], making it a relevant communication substrate for drone swarm deployments.

Gossip-based dissemination improves scalability by reducing redundant transmissions. An alternative paradigm relies on gossip-based communication. The theoretical foundations of gossip protocols are well established in [Shah 2009], and a number of practical designs have been proposed to adapt these ideas to resource constrained and dynamic networks. Protocols such as RAPID [Drabkin et al. 2007] and Trickle [Levis et al. 2008] implement so-called polite gossip mechanisms, in which broadcast activity is regulated according to local utility and network conditions. Rather than blindly flooding messages, nodes exchange lightweight metadata to detect missing information and trigger selective retransmissions when needed. This approach reduces redundant transmissions while preserving a high probability of delivery. More advanced techniques, such as MIXER [Herrmann et al. 2018], further improve dissemination efficiency by introducing network coding into the gossip process.

References

Gilbert, S. and Lynch, N. (2012). Perspectives on the cap theorem. Computer, 45(2):30-36.

Janic, M. (2005). Multicast in network and application layer. Dissertation (tu delft), Delft University of Technology.

Johnson, D., Ntlatlapa, N. S., and Aichele, C. (2008). A simple pragmatic approach to mesh routing using BATMAN. In 2nd IFIP International Symposium on Wireless Communications and Information Technology in Developing Countries.

Guillen-Perez, A., Montoya, A.-M., Sanchez-Aarnoutse, J.-C., and Cano, M.-D. (2021). A comparative performance evaluation of routing protocols for flying ad-hoc networks in real conditions. Applied Sciences, 11(10):4363.

Shah, D. (2009). Gossip algorithms. Foundations and Trends in Networking, 3(1):1-125.

Drabkin, V., Friedman, R., Kliot, G., and Segal, M. (2007). Rapid: Reliable probabilistic dissemination in wireless ad-hoc networks. In 2007 26th IEEE International Symposium on Reliable Distributed Systems (SRDS 2007), pages 13-22. IEEE.

Levis, P., Brewer, E., Culler, D., Gay, D., Madden, S., Patel, N., Polastre, J., Shenker, S., Szewczyk, R., and Woo, A. (2008). The emergence of a networking primitive in wireless sensor networks. Communications of the ACM, 51(7):99-106.

Herrmann, C., Mager, F., and Zimmerling, M. (2018). Mixer: Efficient many-to-all broadcast in dynamic wireless mesh networks. In Proceedings of the 16th ACM Conference
on Embedded Networked Sensor Systems. This project has three main goals. Firstly, we aim to survey data dissemination techniques for environmental monitoring. Secondly, we aim to design novel, adaptive protocols for drone swarms based on the ecosystems to be monitored. Thirdly, we expect to validate the techniques with a proof of concept. The main steps of our methodology are listed below:

1) Extensive state-of-the-art research.

2) Reproduction and comparison of remarkable dissemination solutions through simulation and emulation frameworks;

3) Formal specification and verification of novel dissemination protocols;

4) Implementation and performance evaluation of a proof-of-concept prototype using a small drone swarm.