ERSA’2008
The International Conference on
Engineering of Reconfigurable Systems and Algorithms

Reconfigurable Computing for Space Based Applications

Chairman: Dr. Aravind R Dasu, Utah State University, USA

Embedded stand-alone microprocessors have long been the workhorse of space-based computers. But the demands to explore deeper into space and obtain greater return in science per mission has spurred the need to explore more powerful alternatives that can support a suite of algorithms related to both data processing as well as command and control, which are far too complex for traditional space based microprocessors.

Modern FPGAs with a complex and rich fabric consisting of embedded processors, memory units, DSP blocks and offering features such as partial-dynamic reconfiguration, selective SEU/MEU fault protection circuits for routing/logic, are becoming increasingly attractive as the future of space based computing systems.

But their very richness of components and features, make porting applications onto them a challenging task, opening up possibilities of optimizations and novel techniques at a multitude of levels, some of which are listed below (and serve as guidelines for topics of the technical session):

• Software/algorithms customized for space based applications to be ported onto FPGAs:

  • Data compression in domains such as image, hyper-spectral data etc.,
  • Mission planning/autonomy oriented algorithms, Navigation/guidance etc.
  • Design automation tools for VLSI architecture optimization, power reduction etc.

• Hardware customized for space based FPGAs:

  • Hardware-software co-design
  • Partial-dynamic reconfiguration
  • Selective fault tolerance/mitigation circuits
  • Low power designs
  • Non-standard low complexity floating-point arithmetic circuits
  • Highly parallel/pipelined VLSI-architectures.
  • Multi-FPGA systems for scalable performance