Optimization of satellite’s onboard data processing workload and related system resources, by offloading specific CPU-intensive tasks onto external computational nodes

This study presents an innovative way of supporting Low Earth Orbit satellites -in particular Earth Observation missions- by deploying a LEO infrastructure made up of high-performing computational nodes which can assist clients in their most demanding computational needs. In particular, this work investigates the viability of such a solution, proposing, for specific categories of missions, an approach to reduce the use of some onboard resources and to save on satellite system’s costs by offloading computationally-intensive tasks from the onboard systems, with a contextual optimization of the radio link utilization and of the wait-time for the availability of the final data that result from the overall processing.