Cubesat Heterogeneous-processor Array for Multi-Parametric Sensing (CHAMPS)

ABSTRACT: CubeSats, small modular satellite platforms that range from 1U (10x10x10cm) to 3U (10x10x30cm) size, are becoming highly regarded platforms for significant advancements in space systems. One main reason is that because of typical lifetime with low earth orbit (LEO) missions CubeSats can leverage commercial off-the shelf (COTS) electronics with minimal device enhancements to survive radiation effects, both total ionizing dose (TID) and single event effects (SEE). This use of COTS electronics opens up CubeSats to state-of-the-art advances in computing technology which can revolutionize the sensor data processing power of these small spacecraft despite their very limited electronics Size, Weight and Power (SWaP) budgets. One very promising new data processing technology consists of the use of multiple processor arrays such as General-Purpose Graphics Processing Units (GPGPUs). These processor arrays can be combined with FPGA fabric and standard processors in order to provide three levels of processing capability: general management functions in a user-friendly Operating System environment; sensor data calculations performed in parallel by the processor array; and FPGA-based hardware acceleration for key performance bottlenecks. For the proposed project, Innoflight will compare available devices and complete a preliminary design for this type of flexible space-based processor. BENEFIT: Innoflight"s CubeSat Heterogeneous-processor Array for Multi-Parameter Sensing (CHAMPS) system that is proposed for this SBIR effort will have a very broad range of possible applications for space missions. Most spacecraft can benefit from improved processing capability as defined by the ratio of performance to power. The CHAMPS system will be designed from the ground up to be scalable for missions with varying processing requirements and available power. This scalability will make CHAMPS useful not only for CubeSats but for larger platforms as well. Therefore, it should be an interesting processing option for a wide range of DoD, NASA, and commercial spacecraft missions, most of which involve the reduction and fusion of sensor data as one of their key functions.