Abstract

The formula for calculating the theoretical probability of false association for the technique of track-to-track association (TTTA) using multiscan data is presented. The multiscan TTTA technique and the single-scan TTTA technique are studied by analysis and simulation. The conclusion is that the theoretical performance of the multiscan method is better than that of the single-scan method; however the practical performance of the former evaluated in our simulation is much worse than its theoretical performance, yet could still be more or less better than that of the latter. The reason is that estimation errors of the sensor-level tracks are strongly dependent across time.

Keywords: Track-to-track association; Multisensor tracking; Information fusion

Nomenclature

estimates of n-dimensional state vectors from sensor i,j

n-dimensional true state vectors

Δ^ij(k)

difference of the true states

difference of two estimates

error in the difference between two estimates

T^ij(k)

covariance of

D(k)

test statistic for single-scan track-to-track association technique

C(k)

test statistic for multiscan track-to-track association technique

B

distance vector between tracks of two different targets

W

normalized squared distance

N_n(B,T)

n-dimensional multivariate Gaussian density function with mean B and covariance T

chi-square distribution with n degrees of freedom

noncentral chi-square distribution with n degrees of freedom and noncentral parameter B^′B

α

significance level

value of at significance level α

λ

decision threshold

p(C)

probability density function of C

ψ_C(k)(μ)

characteristic functions of C(k)

P_CA

probability of correct association

P_FA

probability of false association

P_FAS1

simulated probability of false association for single-scan track-to-track association technique

P_FAS2

simulated probability of false association for multiscan track-to-track association technique

P_FA1

theoretical probability of false association for single-scan track-to-track association technique

P_FA2

theoretical probability of false association for multiscan track-to-track association technique

P_CAS1

simulated probability of correct association for single-scan track-to-track association technique

P_CAS2

simulated probability of correct association for multiscan track-to-track association technique

P_CA1

theoretical probability of correct association for single-scan track-to-track association technique

P_CA2

theoretical probability of correct association for multiscan track-to-track association technique

Article Outline

Nomenclature

1. Introduction

2. Single-scan TTTA technique and its performance

3. Multiscan TTTA technique and its performance

4. Simulation

5. Conclusion

References