Abstract : Tracking an unknown, nonstationary signal in time and frequency against a noisy, nonstationary background on an intensity-modulated sensor display is a difficult problem. Traditional techniques involve thresholding the sensor data and treating the exceedences as point measurements that are subsequently fed to a tracking algorithm. Choosing this threshold is a challenge in itself, and even then it is typically subject to a prescribed probability of detection and probability of false alarm. In reference 1, Streit derives a new tracking algorithm that uses all of the senor data, and thus avoids thresholding entirely. The fundamental premise of this tracking algorithm is that losses due to thresholding the sensor data can be eliminated completely if all of the sensor data are used by the tracking algorithm.
[1]
R. Streit,et al.
Probabilistic Multi-Hypothesis Tracking
,
1995
.
[2]
D. Rubin,et al.
Maximum likelihood from incomplete data via the EM - algorithm plus discussions on the paper
,
1977
.
[3]
N. L. Johnson,et al.
Discrete Multivariate Distributions
,
1998
.
[4]
Roy L. Streit,et al.
Tracking on Intensity-Modulated Data Streams
,
2000
.
[5]
Alan V. Oppenheim,et al.
Discrete-time signal processing (2nd ed.)
,
1999
.
[6]
Y. Bar-Shalom.
Tracking and data association
,
1988
.