Protecting Image Processing Pipelines against Configuration Memory Errors in SRAM-Based FPGAs

Image processing systems are widely used in space applications, so different radiation-induced malfunctions may occur in the system depending on the device that is implementing the algorithm. SRAM-based FPGAs are commonly used to speed up the image processing algorithm, but then the system could be vulnerable to configuration memory errors caused by single event upsets (SEUs). In those systems, the captured image is streamed pixel by pixel from the camera to the FPGA. Certain local operations such as median or rank filters need to process the image locally instead of pixel by pixel, so some particular pixel caching structures such as line-buffer-based pipelines can be used to accelerate the filtering process. However, an SRAM-based FPGA implementation of these pipelines may have malfunctions due to the mentioned configuration memory errors, so an error mitigation technique is required. In this paper, a novel method to protect line-buffer-based pipelines against SRAM-based FPGA configuration memory errors is presented. Experimental results show that, using our protection technique, considerable savings in terms of FPGA resources can be achieved while maintaining the SEU protection coverage provided by other classic pipeline protection schemes.

[1]  R.C. Baumann,et al.  Radiation-induced soft errors in advanced semiconductor technologies , 2005, IEEE Transactions on Device and Materials Reliability.

[2]  Marco D. Santambrogio,et al.  TMR and Partial Dynamic Reconfiguration to mitigate SEU faults in FPGAs , 2007, 22nd IEEE International Symposium on Defect and Fault-Tolerance in VLSI Systems (DFT 2007).

[3]  Richard Szeliski,et al.  Computer Vision - Algorithms and Applications , 2011, Texts in Computer Science.

[4]  Stergios I. Roumeliotis,et al.  Vision-Aided Inertial Navigation for Spacecraft Entry, Descent, and Landing , 2009, IEEE Transactions on Robotics.

[5]  F. Irom Guideline for ground radiation testing of microprocessors in the space radiation environment , 2008 .

[6]  Mark F. Flanagan,et al.  An Efficient Technique to Protect Serial Shift Registers Against Soft Errors , 2013, IEEE Transactions on Circuits and Systems II: Express Briefs.

[7]  Stefan Mangard,et al.  Protecting the Control Flow of Embedded Processors against Fault Attacks , 2015, CARDIS.

[8]  Pedro Reviriego,et al.  A Comparison of Dual Modular Redundancy and Concurrent Error Detection in Finite Impulse Response Filters Implemented in SRAM-Based FPGAs Through Fault Injection , 2018, IEEE Transactions on Circuits and Systems II: Express Briefs.

[9]  Ricardo Reis,et al.  Multiple fault injection platform for SRAM-based FPGA based on ground-level radiation experiments , 2015, 2015 16th Latin-American Test Symposium (LATS).

[10]  Tomás Krajník,et al.  Accelerating embedded image processing for real time: a case study , 2016, Journal of Real-Time Image Processing.

[11]  Jong-Yeol Lee,et al.  Distributed CRC Architecture for High-Radix Parallel Turbo Decoding in LTE-Advanced Systems , 2015, IEEE Transactions on Circuits and Systems II: Express Briefs.

[12]  Mehdi B. Tahoori,et al.  An Analytical Approach for Soft Error Rate Estimation of SRAM-Based FPGAs , 2004 .

[13]  Pedro Reviriego,et al.  Error Detection Technique for a Median Filter , 2017, IEEE Transactions on Nuclear Science.