TY - JOUR
T1 - Distributed Inference of the Multiplex Network Topology of Complex Systems
AU - Lombana, Daniel Alberto Burbano
AU - Freeman, Randy A.
AU - Lynch, Kevin
N1 - Funding Information:
Manuscript received December 6, 2018; accepted February 14, 2019. Date of publication March 8, 2019; date of current version March 18, 2020. This work was supported in part by the Office of Naval Research under Grant N00014-13-1-0331 and in part by the Army Research Lab. Recommended by Associate Editor Y. Wan. (Corresponding author: Daniel Alberto Burbano Lombana.) D. A. Burbano Lombana is with the Department of Mechanical and Aerospace Engineering, New York University, New York, NY 11201 USA (e-mail:,daniel.burbano@nyu.edu).
PY - 2020/3
Y1 - 2020/3
N2 - Many natural and engineered systems can be modeled as a set of nonlinear units interacting with each other over a network of interconnections. Often, such interactions occur through different types of functions giving rise to so-called multiplex networks. As an example, two masses can interact through both a spring and a damper. In many practical applications, the multiplex network topology is unknown, and global information is not available. In this paper, we propose a novel distributed approach to infer the network topology for a class of networks with both nonlinear node dynamics and multiplex couplings. In our strategy, the estimators measure only local network states but cooperate with their neighbors to fully infer the network topology. Sufficient conditions for stability and convergence are derived using appropriate Lyapunov functions. Applications to networks of chaotic oscillators and multirobot manipulation are presented to validate our theoretical findings and illustrate the effectiveness of our approach.
AB - Many natural and engineered systems can be modeled as a set of nonlinear units interacting with each other over a network of interconnections. Often, such interactions occur through different types of functions giving rise to so-called multiplex networks. As an example, two masses can interact through both a spring and a damper. In many practical applications, the multiplex network topology is unknown, and global information is not available. In this paper, we propose a novel distributed approach to infer the network topology for a class of networks with both nonlinear node dynamics and multiplex couplings. In our strategy, the estimators measure only local network states but cooperate with their neighbors to fully infer the network topology. Sufficient conditions for stability and convergence are derived using appropriate Lyapunov functions. Applications to networks of chaotic oscillators and multirobot manipulation are presented to validate our theoretical findings and illustrate the effectiveness of our approach.
KW - Adaptive control
KW - distributed algorithms
KW - network control
KW - network reconstruction (NR)
UR - http://www.scopus.com/inward/record.url?scp=85082299117&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85082299117&partnerID=8YFLogxK
U2 - 10.1109/TCNS.2019.2903907
DO - 10.1109/TCNS.2019.2903907
M3 - Article
AN - SCOPUS:85082299117
VL - 7
SP - 278
EP - 287
JO - IEEE Transactions on Control of Network Systems
JF - IEEE Transactions on Control of Network Systems
SN - 2325-5870
IS - 1
M1 - 8663306
ER -