There is no best universal algorithm. Each algorithm has some flaws and advantages. If your problem is nonlinear and you know the initial state error and noise covariances, then use the extended Kalman filter. If you do not know the initial conditions and noise covariances (as it usually is), try the extended UFIR algorithm:

Y. S. Shmaliy, Suboptimal FIR filtering of nonlinear models in additive white Gaussian noise, IEEE Trans. Signal Process., vol. 60, no. 9, pp. 5519-5527, Oct. 2012.

This is not a remedy as well against all errors. But this filter ignores noise and initial errors and is practically more robust than the Kalman filter.

Initial position estimation, if known, current algorithms can converge.

We may look at how to triangulate from different sensors, removal of ghosts, and then also look at TMA. In this case Ristic book on Beyond Kalman Filter is a classic.

Any other book or specific algorithm we can start discussion.

There is no best universal algorithm. Each algorithm has some flaws and advantages. If your problem is nonlinear and you know the initial state error and noise covariances, then use the extended Kalman filter. If you do not know the initial conditions and noise covariances (as it usually is), try the extended UFIR algorithm:

Y. S. Shmaliy, Suboptimal FIR filtering of nonlinear models in additive white Gaussian noise, IEEE Trans. Signal Process., vol. 60, no. 9, pp. 5519-5527, Oct. 2012.

This is not a remedy as well against all errors. But this filter ignores noise and initial errors and is practically more robust than the Kalman filter.

Thank you very much for all advices and comments

Prof. Shmaliy I have downloaded your book (Continuous time systems) it is really help full to understand the systems (modeling,analysis and realization)

Abdul Momin and Raj i think the Kalman is good in tracking using in radar system , because it is depend on target maneuvering , but before that we know the radar is an active system ,if we want using the case That the receiver is passive and the target is active ,so what do you think about this case is the Kalman filtering is accurate, how could it catch the target ???

Hi Hesh, thanks for good words about my book.

About the target tracking. What do you mean ¨the target is active¨? If you mean that the receiver locates the target based on the reflections from the target as in passive radars, then yes, the extended or unscented Kalman filter is now a basic solution for line-tracking and target state estimation.Tracking and estimation errors strongly depend on noise. So, if you do not know the target and measurement noise covariances, then read my first comment.

Dear Abdul, also thanks for good words about my book! It was written as a textbook and it is a pleasure to know that Hesh and you use it.

There would be rather than one algorithm various types of approach depending on the philosophy behind conception of your model. Fundamentally: Deterministic and stochastic-statistical approaches. First one is based in any case on "source footprint" knowledge and therefore you look for the closest match for your model in the signal and the second one relies in the forever beloved, simple yet mighty correlation -can you perhaps guess my taste from my comments? haha. In earth sciences for us is really impossible to have source prediction in the strictest of senses so we estimate source by correlation and extrapolation algorithms. Yet in NDT techniques quite often you model the source.