Abstract
In this paper, a recent novel Bayesian deghosting algorithm for the Multi-Static Primary Surveillance Radar (MSPSR) systems is analysed. The previous research was primarily concerned with the verification of the approach based on the Indian Buffet Process (IBP) and its usage as the apriori probability distribution for the bistatic-to-cartesian association matrix. In this paper, the degree of dependence of the algorithm performance on the specific choice of the hyperparameters is explored and assessed. Together with the analysis, the methodology for the choice of hyperparameters without any apriori knowledge is developed. All analyses are verified using simulations derived from a realistic MSPSR system. The results obtained suggest a rather low sensitivity of the association results concerning the hyperparameters computation method.
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call