Hyperbola Tracing Algorithm Based on Particle Filter Approach Within a Half-Quadrant Space for Signal Source Localization

Signal source localization has become an interesting research problem due to its wide applications. Time Difference of Arrival is one of the popular models to estimate the location of signal source using multiple sensors. Closed-form and exact-solution algorithms are commonly utilized to solve non-linear distance equations in the model through linearization process. As an alternative, hyperbola-tracing algorithm to find intersection point between two hyperbolas is proposed in this paper. This algorithm starts with reduction of possible solution space into half of a quadrant in 2D problem. It is carried out through comparison of time of signal arrival at each sensor. Two pairs of sensors are then selected to build two hyperbolas. The second step consists of particle-filter-based approach to seek the intersection of these two non-linear functions. A simulation is presented to show the performance of this algorithm.