TBD_Viterbi: Enhanced Viterbi Algorithm for Track-Before-Detect Dynamic Programming

To achieve a more accurate track-before-detect algorithm, we have implemented an enhanced Viterbi algorithm for dynamic programming. This algorithm represents a classical optimization problem-solving approach widely applied in computer science, artificial intelligence, and mathematics. By integrating this algorithm into our track-before-detect framework, we can significantly improve both accuracy and computational efficiency while achieving superior target tracking performance. The implementation involves modifying the traditional Viterbi decoder to handle state transitions in tracking scenarios, where path metrics are calculated using likelihood functions based on sensor measurements. Key enhancements include adaptive thresholding for state pruning and memory-efficient path history management to handle extended tracking sequences. The algorithm maintains multiple hypothesis paths through a trellis structure, with each node representing a potential target state and edges representing feasible transitions between consecutive time steps.