J. K. Wolf, A. M: Viterbi, S.G. Dixon, "Finding the Best Set of K paths through a trellis with Applications to Multitarget Tracking", IEEE Trans. AES 25, pp. 287-296, 1989.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....through a trellis graph can be defined as a Minimum Cost Network Flow (MCNF) problem. He also showed that the worst case computation time for this problem is bounded by n 3 logn, where n is the number of nodes in the trellis. His O(n 3 logn) time algorithms are much faster than those proposed in [15]. Note that the computational complexity of our algorithm is currently under investigation, and will be reported elsewhere. It is important to note that, for K 2 the best set of K paths is not found in general by finding the best path, then the next best path that is completely disjoint with ....

.... shortest link disjoint paths [3] The reason for this is that these paths are not independent of one another because of the requirement that the paths be completely disjoint so that it might be better to use one or more of the branches that are part of the best single path for other paths [15]. 3. Problem Transformation We have formulated the path selection and rerouting problem in addressing network survivability as a graph theoretic, K best path, problem with an efficient minimum cost network flow solution. For any given network topology, our objective is to transform it to a ....

J.K. Wolf, A.M. Viterbi and G.S. Dixon, "Finding the Best Set of K Paths Through a Trellis with Applications to Multitarget Tracking", IEEE Trans. AES 25, pp. 287-296, 1989.

....association, the Viterbi Data Association (VDA) algorithm [16] an extension of an earlier Viterbi based algorithm [17] is a single track algorithm in which measurement gating is incorporated for an over the horizon radar application. A K track Viterbi algorithm has been proposed for MHT [18], but this is only optimum with respect to a simple nonprobabilistic cost function (i.e. the energy required to get from one measurement to the next) A more efficient computational algorithm for the same cost was suggested in [19] As a hypothesis pruner, the Viterbi algorithm assures the optimum ....

.... same cost was suggested in [19] As a hypothesis pruner, the Viterbi algorithm assures the optimum hypothesis is kept only if all hypothesis costs satisfy a finite state machine or Markov condition [8] Although this condition is satisfied for the simple nonprobabilistic energy cost used in [18], it is not for the maximum likelihood and MAP costs which are standard for MHT. For MAP based MHT, the single track Viterbi algorithms [16] 17] implement suboptimum sequential pruning, and are prone to loosing tracks for high clutter and missed detection scenarios, especially when extended to ....

[Article contains additional citation context not shown here]

J. K. Wolf, A. M. Viterbi and G. S. Dixon, "Finding the best set of K paths through a trellis with application to multitarget tracking," IEEE Trans. on Aerospace & Elect. Sys., Vol. 25, No. 2, pp. 287-296, Mar. 1989.

....which is optimum over the entire processing interval may not be found, and in particular in sequential track estimation valid tracks can be dropped after trackto measurement association fails for several measurement times. This research was supported by ONR under Grant N00014 98 1 0892. In [6], an alternative approach to computation reduction is proposed. Therein, a K path extension of the Viterbi algorithm [7] is used to find the best K nonintersecting tracks according to a deterministic energy cost function. A trellis diagram is defined where at each stage (measurement time) states ....

....the problems associated with feasible tracks and gating volumes. Use of this K path extension of the Viterbi algorithm is predicated on a finite state machine structure or Markov condition in the cost function, which is satisfied with the particular deterministic energy cost function used in [6]. However, an ML or Bayesian cost function will not satisfy this requirement since track measurements are correlated throughout the history of a track. One way to think of this is that, in implementing an ML or Bayesian multitrack estimator, Kalman filters are used in the cost computation to ....

[Article contains additional citation context not shown here]

A. M. Viterbi J. K. Wolf and G. S. Dixon. Finding the best set of k paths through a trellis with application to multitarget tracking. IEEE Trans. on Aerospace & Elect. Systems., Vol. 25, No. 2, pp. 287-296, March 1989.

....trellis graph can be defined as a Minimum Cost Network Flow (MCNF) 4 problem. He also showed that the worst case computation time for this problem is bounded by n 3 logn, where n is the number of nodes in the trellis. His O(n 3 logn) time algorithms are much faster than those proposed in [15]. Note that the computational complexity of our algorithm is currently under investigation, and will be reported elsewhere. It is important to note that, for K 2 the best set of K paths is not found in general by finding the best path, then the next best path that is completely disjoint with the ....

.... shortest link disjoint paths [3] The reason for this is that these paths are not independent of one another because of the requirement that the paths be completely disjoint so that it might be better to use one or more of the branches that are part of the best single path for other paths [15]. 3. Problem Transformation We have formulated the path selection and rerouting problem in addressing network survivability as a graph theoretic, K best path, problem with an efficient minimum cost network flow solution. For any given network topology, our objective is to transform it to a ....

J.K. Wolf, A.M. Viterbi and G.S. Dixon, "Finding the Best Set of K Paths Through a Trellis with Applications to Multitarget Tracking", IEEE Trans. AES 25, pp. 287-296, 1989.

....associate the detections over time to form multiple target tracks. For example, maximum likelihood (ML) based data association algorithms have been developed based on track splitting [2] and integer programming [3] In [1] and [4] algorithms based on a Bayesian formulation have been proposed. In [5], a K path extension of the Viterbi algorithm is used to find the best K tracks according to a deterministic energy cost function. These approaches identify a set of multiple tracks but do not directly address the possibility that there may be alternative yet reasonable sets of tracks that fit ....

....to problems for which all other paths can not be optimum. It has been used extensively in digital communication and other application problems for which the sequence can be modeled as a Markov process. It has also been proposed for multitarget tracking based on a deterministic energy cost [5], where the incremental path costs from a state j m Gamma1 at stage m Gamma 1 to a state j m at stage m is a function of only the j m Gamma1 and j m measurements. In general, multitarget tracking problems, couched in a trellis framework, can not be solved using the Viterbi algorithm. For the ML ....

[Article contains additional citation context not shown here]

A. M. Viterbi J. K. Wolf and G. S. Dixon. Finding the best set of k paths through a trellis with application to multitarget tracking. IEEE Trans. on Aerospace & Elect. Systems., 25, No. 2:287--296, March 1989.

No context found.

J. K. Wolf, A. M: Viterbi, S.G. Dixon, "Finding the Best Set of K paths through a trellis with Applications to Multitarget Tracking", IEEE Trans. AES 25, pp. 287-296, 1989.

No context found.

A. M. Viterbi, J. K. Wolf and G. S. Dixon, Finding the best set of k paths through a trellis with application to multitarget tracking IEEE Transactions on Aerospace and Electronic Systems 25 2 (1989) 287--296.

No context found.

J.K. Wolf, A.M. Viterbi and G.S. Dixon, "Finding the Best Set of K Paths Through a Trellis with Applications to Multitarget Tracking", IEEE Trans. AES 25, pp. 287-296, 1989.

No context found.

J. K. Wolf, A. M. Viterbi, and G. S. Dixon, "Finding the best set of K paths through a trellis with application to multitarget tracking," Trans. on Aerospace and Elect. Sys. 25, pp. 287--296, Mar 1989.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC