A Simulation Toolset for Adaptive Remote Sensing (STARS)
Presenting Author: Graeme Smith
Organization: The Ohio State University
Co-Author(s): Andrew O'Brien, Christopher D. Ball, Stephen Musko, Jakob Delong, Mohammad Abu Shattal, Ryan Linnabary, and Joel T. Johnson
Future Earth Science missions will use constellations of small satellites equipped with fully-adaptive sensors. Typically, these small satellites will be required to operate within significant resource limitations (i.e. power, data rate, and on-board processing). Achieving the full benefits of small satellite constellations will therefore require the management and optimization of the complete distributed sensing system within these limitations. Optimizing missions that utilize fully-adaptive sensors requires sophisticated algorithms that are difficult to design and validate but are expected to yield greatly improved science outcomes. A new software toolset is needed to help users optimize these adaptive algorithms in next generation observing system simulation experiments (OSSEs). Under AIST-2016, the Simulation Toolset for Adaptive Remote Sensing (STARS) Project is developing the new software libraries needed by future OSSEs to support constellations of adaptive sensors. Three distinct libraries comprise STARS: ADAPT provides functions and classes to support the simulation of fully-adaptive sensors, MANAGE provides functions and classes to support the simulation of resource management algorithms, and COLLABORATE provides functions and classes to support the simulation of constellation communications. Details of these object-oriented C++ libraries will be presented at the conference along with results from example simulations. Several case studies are being conducted to demonstrate features of the software. One such example involves the use of a radar measuring cloud profiles from low Earth orbit that can adapt waveform parameters to optimize the value of its measurements. Adaptation of the sensor occurs within the constraints of management of battery and data storage resources within a CubeSat platform. Because the system cannot operate at 100% duty cycle, the measurements taken per orbit must be scheduled to maximize utility. This case study is further extended to include the cueing of measurements within a satellite constellation. Results from the case study will be presented as an example of the capabilities of the STARS toolset as well as the potential of future adaptive and collaborative remote sensing systems.