A new approach to maneuvring target tracking in passive multisensor environment

Ammar Cherchar (1) , Adel Belouchrani (2) , Thierry Chonavel (3, 4)

1 EMP - École Militaire Polytechnique [Alger]
2 ENP - École nationale polytechnique [Alger, Algérie]
3 Lab-STICC_TB_CID_TOMS
4 SC - Département Signal et Communications

This paper present a new approach to the multisensor Bearing-Only Tracking applications (BOT). Usually, a centralized data fusion scheme which involves a stacked vector of all the sensor measurements is applied using a single estimation filter which copes with the non-linear relation between the states and the measurements. The aforementioned approach is asymptotically optimal but suffers from the computational burden due to the augmented measurement vector and transmission aleas like delays generated by the bottleneck that occurs at the fusion center. Alternatively, since the Cartesian target positions can be determined by fusing at least 2 infrared sensor measurements in 2D case, one can use a local linear filter to estimate the target motion parameters, then a state fusion formula based on the Likelihood of the expected overall local measurements is applied to obtain the global estimate. The simulation results show that the proposed approach performance is equivalent to the centralized fusion schema in terms of tracking accuracy but exhibits the advantages of the decentralized fusion schema like parallel processing architecture and robustness against transmission delays. In addition, the low complexity of the obtained algorithm is well suited for real-time applications.

Mots clés: Target tracking; Bearing only tracking; Kalman filtering; Data fusion

Domaines: Traitement du signal et de l'image [eess.SP]

Ammar Cherchar, Adel Belouchrani, Thierry Chonavel. A new approach to maneuvring target tracking in passive multisensor environment. Passive 2010, Jun 2010, Brest, France. ⟨hal-00725143⟩