Associate Professor HDR

My research interests include distributed estimation, graph filtering, wireless sensor networks, traffic prediction and estimation, filtering, identification, and tensor analysis. My current research activities invlove network systems and graph analysisdistributed estimation, tensor analysis, and transportation networks and smart mobility. Some other contributions on nonlinear systems identification, nonlinear filtering, and signal processing for communications can be found in my publication list.

Network systems and graph analysis

This is a research topic at the boundaries between graph theory and dynamical systems theory.  A first main line of research is to study complex systems whose interactions are modeled with graphs, and to unveil the effect of the graph topology on system-theoretic properties such as observability, controllability and input and state observability. Significant recent results include: average observability of large scale networks, structural and strongly structural characterization of input and state observability for linear network systems being linear time invariant, time-varying or topology time-varying, and characterization of observability of consensus networks over distance-regular graphs.
A second line of research concerns graph discovery, namely algorithms aiming at reconstructing some properties of the graph (graph robustness through Laplacian eigenvalues estimation and network topology reconstruction). In particular, distributed algorithms are devised for estimating Laplacian eigenvalues using transient measurements of consensus protocols and for reconstruction the network topology.

Selected papers:
  • S. GRACY, F. GARIN, and A.Y. KIBANGOU, "Input and state observability of network systems with time-varying topology". IEEE Transactions on Control of Network Systems, In press. [DOI: 10.1109/TCNS.2018.2880304]
  • T.-M.-D. TRAN and A.Y. KIBANGOU, "Distributed estimation of Laplacian eigenvalues via constrained consensus optimization problems", Systems and Control letters, 2015.[ DOI: 10.1016/j.sysconle.2015.04.001]
  • A.Y. KIBANGOU and C. COMMAULT, "Observability in connected strongly regular graphs and distance regular graphs", IEEE Trans. on Control of Network Systems, Vol. 1, No 4, pp. 360-369,  2014. [DOI: 10.1109/TCNS.2014.2357532]
  • F. MORBIDI and A.Y. KIBANGOU, "A distributed solution to the network reconstruction problem", Systems and Control Letters, No. 70, pp. 85-91, 2014. [DOI: 10.1016/j.sysconle.2014.05.008]

Distributed estimation and data fusion in network systems

This research topic concerns distributed data combination from multiple sources (sensors), and related information fusion, to achieve more specific inference than that could be achieved by using a single source (sensor). It plays an essential role in many networked applications, such as networked control, monitoring, and surveillance. One key ingredient for distributed algorithms is the well known consensus issue. We have been mainly focused on synthesis of finite-time average consensus protocols and also on the joint state and unkown input estimation problem in a distributed setting. 

Selected papers:

Tensor decompositions and applications

This issue concerns the analysis and the processing of multidimensional data, organized in multiway array. Precisely, we devise distributed algorithms for tensor decompositions (PARAFAC) and study decomposition of large tensors, a crucial issue in our big data era. Applications concern distributed detection in complex networks and distributed signal processing for collaborative networks and traffic data imputation. This topic is developed in strong collaboration with Federal University of Ceara (Brazil) and I3S (Sophia Antipolis).

Selected papers:

Transportation networks and smart mobility

This is currently the main application domain of the NeCS team. We are working on robust traffic estimation and prediction issues. The developped methods are assessed using real data from the Grenoble south ring through the city lab GTL. Other results concern pedestrian navigation through attitude estimation.

Selected papers:


Grenoble Images Parole Signal Automatique laboratoire

UMR 5216 CNRS - Grenoble INP - Université Joseph Fourier - Université Stendhal