REFLEX: Reconfigurable logic for entropy extraction

True Random Number Generators are cryptographic primitives that harness on-chip physical phenomena to generate random bits. Variations in the fabrication process and operating conditions can decrease reliability of these circuits by decreasing the randomness of their output. Conventional lightweight post-processing techniques use an XOR tree to combine the outputs of two or more TRNGs and improve the net entropy. We present REFLEX: Reconfigurable Logic for Entropy Extraction; a post-processing technique using reconfigurable logic. REFLEX replaces each stage of XOR tree with a reconfigurable logic block that is programmed based on the entropy of TRNGs. The output of TRNGs are further reordered before entering the logic tree to provide optimal entropy extraction. Up to 55% more chips pass NIST randomness tests when using REFLEX as compared to conventional XOR tree. When implemented in IBM 32nm SOI process, REFLEX has a maximum synthesized area of 620μm2. It consumes a leakage power of 70μW and an energy overhead of 0.04pJ/bit when operating at 2GHz. The design is scalable to multiple TRNGs and provides energy efficient entropy extraction for low power applications.

[1]  Wayne P. Burleson,et al.  Entropy extraction in metastability-based TRNG , 2010, 2010 IEEE International Symposium on Hardware-Oriented Security and Trust (HOST).

[2]  H. P. Hofstee,et al.  True hardware random number generation implemented in the 32-nm SOI POWER7+ processor , 2013, IBM J. Res. Dev..

[3]  R. Thewes,et al.  A low-power true random number generator using random telegraph noise of single oxide-traps , 2006, 2006 IEEE International Solid State Circuits Conference - Digest of Technical Papers.

[4]  Jing Wang,et al.  A 1.04 µW Truly Random Number Generator for Gen2 RFID tag , 2009, 2009 IEEE Asian Solid-State Circuits Conference.

[5]  K. Marton,et al.  Randomness in Digital Cryptography : A Survey , 2011 .

[6]  Hashimoto Masanori,et al.  An oscillator-based true random number generator with jitter amplifier , 2011 .

[7]  Dennis L. Goeckel,et al.  Wireless physical-layer security performance of UWB systems , 2010, 2010 - MILCOM 2010 MILITARY COMMUNICATIONS CONFERENCE.

[8]  Himanshu Kaul,et al.  2.4 Gbps, 7 mW All-Digital PVT-Variation Tolerant True Random Number Generator for 45 nm CMOS High-Performance Microprocessors , 2012, IEEE Journal of Solid-State Circuits.

[9]  Berk Sunar,et al.  A Provably Secure True Random Number Generator with Built-In Tolerance to Active Attacks , 2007, IEEE Transactions on Computers.

[10]  Daniel E. Holcomb,et al.  Power-Up SRAM State as an Identifying Fingerprint and Source of True Random Numbers , 2009, IEEE Transactions on Computers.

[11]  Masanori Hashimoto,et al.  Jitter amplifier for oscillator-based true random number generator , 2011, 16th Asia and South Pacific Design Automation Conference (ASP-DAC 2011).

[12]  Wayne P. Burleson,et al.  Robust metastability-based TRNG design in nanometer CMOS with sub-vdd pre-charge and hybrid self-calibration , 2012, Thirteenth International Symposium on Quality Electronic Design (ISQED).

[13]  Paolo Ienne,et al.  An architecture-independent instruction shuffler to protect against side-channel attacks , 2012, TACO.

[14]  David Blaauw,et al.  A true random number generator using time-dependent dielectric breakdown , 2011, 2011 Symposium on VLSI Circuits - Digest of Technical Papers.

[15]  Laurent Fesquet,et al.  A Self-Timed Ring Based True Random Number Generator , 2013, 2013 IEEE 19th International Symposium on Asynchronous Circuits and Systems.