3.8 Proceedings Paper

Novel Concept of Hardware Security in Using Gate-switching FinFET Nonvolatile Memory to Implement True-Random-Number Generator

Publisher

IEEE
DOI: 10.1109/IEDM13553.2020.9371993

Keywords

-

Funding

  1. Ministry of Science and Technology, Taiwan [MOST-105-2221-E-009-130 MY3]
  2. Research of Excellence program [MOST-109-2639-E-009-001]

Ask authors/readers for more resources

For the first time, we use a gate-switching resistance memory to implement the TRNG (True random number generator). First, a resistance memory was built on a FinFET platform, named RG-FinFET(resistive-gate) RRAM which was simply an MIM(metal-insulator-metal) integrated on top of the FinFET gate. It performed as a novolatile memory (NVM) which used resistance switching to distinguish 0 and 1 states through the drain current of the FinFET. The experimental results show that RG-FinFET exhibits high SET speed of 50ns at 3.4V/RESET speed of 10ns at 2.4V, read time as small as 16ns at 1.1V. Furthermore, excellent 10(7) cycles endurance and data-retention under 125 degrees C for over one month can be achieved. The array-level performance is also analyzed, showing well disturbance-immune during SET, RESET and read. Secondly, a TRNG was developed based on the drain current variation of RG-FinFET. In terms of the security, this TRNG exhibits ideal un-biased normal distribution of hamming distance, and narrow distribution of hamming weight. Moreover, we introduced the concept of XNOR-enhanced operation to TRNG at high temperature to enhance its uniformity. In NIST test, this TRNG passed all items. More importantly, this work is ready for an embedded FinFET technology to develop TRNG with full logic CMOS compatibility.

Authors

I am an author on this paper
Click your name to claim this paper and add it to your profile.

Reviews

Primary Rating

3.8
Not enough ratings

Secondary Ratings

Novelty
-
Significance
-
Scientific rigor
-
Rate this paper

Recommended

No Data Available
No Data Available