Determining the necessity of fault tolerance techniques in FPGA devices for space missions

Functionality of electronic components in space is strongly influenced by the impact of radiation induced errors which may interfere with the proper operation of the equipment. In space missions, FPGA implementations are generally protected using computationally expensive radiation-error mitigation techniques such as error co rrecting codes (ECC) and triple modular redundancy (TMR). For high-performance systems, such fault tolerance techniques can prove problematic due to both the added computational requirements and their resulting power overhead. As such it is important to make a proper assessment of the expected error rates to make a proper selection of mitigation techniques. This paper provides an extensive overview of the techniques used for determining the necessity of such mitigation techniques in space missions and other situations where a large radiation dose will be encountered. Given the presented study and radiation analysis, in this paper an experimental example is presented in the form of a case study on the Digital Receiver System (DRS) in the Netherlands-China Low-frequency Explorer (NCLE) mission, which is implemented using a Xilinx Kintex-7 SRAM FPGA. Fault rates are estimated for a five-year mission to the second Earth-Moon Lagrange point (L2) and the chosen fault mitigation strategy as implemented in NCLE-DRS is presented. The effect of potential upsets on the functionality of DRS has been taken into account in order to make error estimations more precise. Thus, two test-benches are developed and presented to experimentally evaluate the effect of upsets in FPGA configuration memory and the data on the DRS final outputs. The approach provided in this paper should generalize well to other space missions, as long as a general estimate of the expected radiation environment is available.