Synthetic Inline Generation of Modeled Atmosphere

To address the MDAs need for an innovative, low-cost process to perform inline environment generation for real-time hardware-in-the-loop testing and constructive digital simulations, Physical Optics Corporation (POC) proposes to develop a new Synthetic Inline Generation of Modeled Atmosphere (SIGMA) system based on deep modeling and GPU acceleration. The innovation in using multi-layer stacked deep network modeling and GPU accelerators enables close replication of existing environment models and data, without their time-consuming computationally expensive processes, to provide inline generation of credible environment data, with up to 3 orders of magnitude speed improvement, for Tier 1 and Tier 2 modeling and simulation (M&S). As a result, SIGMA eliminates the need for pre-generating environment data and allows more test cases to be executed with environments turned on, with no code modification needed. In Phase I, POC will demonstrate the feasibility of SIGMA by designing its high-level information architecture and componentized system design, developing prototype system, testing it using SHARC/SAMM Atmosphere Generator, and demonstrating its total test case runtime increase by less than 10%. In Phase II, POC plans to develop and deliver a full-scale prototype system, incorporating additional models, and demonstrate its total test case runtime increase by less than 5%. Approved for Public Release | 16-MDA-8917 (15 November 16)