TY - GEN
T1 - Agent based optimally weighted kalman consensus filter over a lossy network
AU - Alam, S. M.Shafiul
AU - Natarajan, Balasubramaniam
AU - Pahwa, Anil
N1 - Funding Information:
The authors would like to thank the NSF for providing support for this research through award No. CNS-1136040.
Publisher Copyright:
© 2015 IEEE.
PY - 2015
Y1 - 2015
N2 - We consider a distributed agent based state estimation problem where each agent has a local underdetermined observation space and is interested in a subset of states. At a particular time instant, each agent predicts its state and makes intermediate correction based on its local measurements. Information about the corrected state elements are then exchanged among neighboring (i.e., agents who share at least one state element) agents. Based on the final processing of the exchanged information, an agent based optimally weighted Kalman consensus Filter is formed in the light of well-established theory of distributed Kalman filtering. The optimal weighting of consensus is derived through Lyapunov function based stability analysis of corresponding estimation error. The effect of communication is also investigated by introducing random failures in the communication link among neighboring agents. The corresponding bounds on the degree of consensus and inter-agent link failure rate are also derived for stable implementation of the agent based dynamic estimation. The proposed filter is applied to a custom built 2- agent system to conform the desired optimal limit for the degree of consensus based information exchange both under perfect and lossy communication network.
AB - We consider a distributed agent based state estimation problem where each agent has a local underdetermined observation space and is interested in a subset of states. At a particular time instant, each agent predicts its state and makes intermediate correction based on its local measurements. Information about the corrected state elements are then exchanged among neighboring (i.e., agents who share at least one state element) agents. Based on the final processing of the exchanged information, an agent based optimally weighted Kalman consensus Filter is formed in the light of well-established theory of distributed Kalman filtering. The optimal weighting of consensus is derived through Lyapunov function based stability analysis of corresponding estimation error. The effect of communication is also investigated by introducing random failures in the communication link among neighboring agents. The corresponding bounds on the degree of consensus and inter-agent link failure rate are also derived for stable implementation of the agent based dynamic estimation. The proposed filter is applied to a custom built 2- agent system to conform the desired optimal limit for the degree of consensus based information exchange both under perfect and lossy communication network.
KW - Binary projection matrix
KW - Kalman consensus filter
KW - Lyapunov stability
KW - Random link failure
UR - http://www.scopus.com/inward/record.url?scp=84964862138&partnerID=8YFLogxK
U2 - 10.1109/GLOCOM.2014.7417484
DO - 10.1109/GLOCOM.2014.7417484
M3 - Conference contribution
AN - SCOPUS:84964862138
T3 - 2015 IEEE Global Communications Conference, GLOBECOM 2015
BT - 2015 IEEE Global Communications Conference, GLOBECOM 2015
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 58th IEEE Global Communications Conference, GLOBECOM 2015
Y2 - 6 December 2015 through 10 December 2015
ER -