Detailed Simulation of Signal-Level Sensor Data Using Monte Carlo Path Tracing and Photon Mapping

Abstract

Simulated sensor data from active and passive sensors has numerous applications in target detection and tracking. Simulated data is particularly useful in performance evaluation of target tracking algorithms where the ground truth of a scenario must be known. For real sensor data it is impossible to know the ground truth so simulated data must be used.

This paper discusses existing methods for simulation of data from active sensors and proposes a method that builds on existing techniques from the field of computer graphics. An extension to existing methods is proposed to accommodate the simulation of active sensor data for which timing and frequency information is required in addition to intensity. Results from an existing method of active sensor data simulation are compared to the results of the proposed method. Additionally, a cloud computing framework is proposed and its scalability evaluated to address the fairly large computational load of such a simulation.