MULTIPLE-TARGET TRACKING USING MAXIMUM-LIKELIHOOD PRINCIPLE

We propose a method (tracking algorithm (TAL)) based on the maximum li kelihood (ML) principle for multiple target tracking in near-field usi ng outputs from a large uniform linear array of passive; sensors, The targets are assumed to be narrowband signals and modeled as sample fun ctions of a Gaussian stochastic process, The phase delays of these sig nals are expressed as functions of both range and bearing angle (''tra ck parameters'') of respective targets, A new simplified likelihood fu nction for ML estimation of these parameters is derived from a second- order approximation on the inverse of the data covariance matrix, Maxi mization of this likelihood function does not involve inversion of the M x M data covariance matrix, where M denotes number of sensors in th e array, Instead, inversion of only a D x D matrix is required, where D denotes number of targets, In practice, D much less than M and, henc e, TAL is computationally efficient, Tracking is achieved by estimatin g track parameters at regular time intervals wherein targets move to n ew positions in the neighborhood of their previous positions, TAL pres erves ordering of track parameter estimates of the D targets over diff erent time intervals. Performance results of TAL are presented, and it is also compared with methods in papers by Sword and by Swindlehurt a nd Kailath. Almost exact asymptotic expressions for the Cramer-Rao bou nd (CRB) on the variance of angle and range estimates are derived, and their utility is discussed.