Session

Technical Session 8: Advanced Technologies II

Location

Utah State University, Logan, UT

Abstract

For science payloads on nanosatellite missions, there is a great interest in cost-effective, reliable and state-of-the-art computing performance. Highly integrated system architectures combine reconfigurable System-on-Chip (SoC) devices, memory and peripheral interfaces in a single System-on-Module (SoM) and offer low resource requirements regarding power and mass, but moderate to high processing power capabilities. The major advantages of these architectures are flexibility, (re)programmability, modularity and module reuse.

However, it is a challenge to use SoM with COTS based memories devices in a radiation sensitive environment. In order to achieve these requirements, mitigation measures, such as the use of redundant or alternative memory devices and in-flight reconfiguration, are important in terms of reliability. Reprogramming strategies e.g. partial dynamic reconfiguration and scrubbing techniques are published in the past.

With a remote sensing instrument for atmospheric temperature measurements using a SRAM-based Xilinx Zynq-7000 SoM, we combine some of these techniques with supervisor circuits to select the boot image from alternative memory devices. The approach distinguishes between programmable logic and processing system reconfiguration, and enables in-flight firmware updates in the case of Single Event Effect (SEE) hazards or changing measurement conditions.

Available for download on Saturday, August 07, 2021

Share

COinS
 
Aug 11th, 11:00 AM

In-Flight Reconfiguration with System-On-Module Based Architectures for Science Instruments on Nanosatellites

Utah State University, Logan, UT

For science payloads on nanosatellite missions, there is a great interest in cost-effective, reliable and state-of-the-art computing performance. Highly integrated system architectures combine reconfigurable System-on-Chip (SoC) devices, memory and peripheral interfaces in a single System-on-Module (SoM) and offer low resource requirements regarding power and mass, but moderate to high processing power capabilities. The major advantages of these architectures are flexibility, (re)programmability, modularity and module reuse.

However, it is a challenge to use SoM with COTS based memories devices in a radiation sensitive environment. In order to achieve these requirements, mitigation measures, such as the use of redundant or alternative memory devices and in-flight reconfiguration, are important in terms of reliability. Reprogramming strategies e.g. partial dynamic reconfiguration and scrubbing techniques are published in the past.

With a remote sensing instrument for atmospheric temperature measurements using a SRAM-based Xilinx Zynq-7000 SoM, we combine some of these techniques with supervisor circuits to select the boot image from alternative memory devices. The approach distinguishes between programmable logic and processing system reconfiguration, and enables in-flight firmware updates in the case of Single Event Effect (SEE) hazards or changing measurement conditions.