IEEE Consumer Electronics Magazine - July/August 2021 - 104

Physical Unclonable Function
Table 1. NIST STS for TRNG performance evaluation.
Test name
Frequency
Block frequency
Cumulative sums (forward)
Cumulative sums (reverse)
Runs
Longest run
Rank
FFT
Overlapping template
Universal
Exhaustive test
p-value Proportion Result
0.845629
0.451279
0.152695
0.847926
0.562478
0.384567
0.747956
0.859674
Nonoverlapping template (148) 0.501324
0.569541
Approximate entropy
Random excursions (8)
Serial-1
Serial-2
0.659841
0.356947
0.846259
Random excursions variant (18) 0.395846
0.756185
Linear complexity
Result is " pass, " if p-value > 0.01.
calculating " p-value, " which indicates the confidence
of randomness. A test with a p-value of 5
0.01 indicates the test is passed and with 99%
confidence. The nature of each random test is
different and hence, the criteria for each test are
also different. We collected a total of 100 random
number sequences, where each sequence consists
of 1 million random bits with light varied at
the random interval to simulate the real-world
scenario. Furthermore, we exposed some sensors
to light and some sensors were blocked during
the data collection procedure. The minimum
pass rate for each test other than random excursion
and random excursion variant is 96 out of
100. The criteria to pass the random excursion
and random excursion variant test are 65 out of
69 random bit sequences.
Table 1 lists the results of each NIST test in
terms of the p-value, the proportion of the test
passed and the result of the test as either pass/fail.
The number in the bracket next to each test
104
0.869416
0.231567
0.9900
1.0000
0.9900
0.9900
1.0000
0.9900
1.0000
1.0000
0.9937
0.9800
0.9900
1.0000
0.9927
0.9943
0.9900
1.0000
1.0000
Pass
Pass
Pass
Pass
Pass
Pass
Pass
Pass
Pass
Pass
Pass
Pass
Pass
Pass
Pass
Pass
Pass
denotes, number of subtests. The proportion simply
indicateshowmany randomsequences passed
for the test, with 1 indicating all 100 or 69 random
bit sequences have cleared the test. The proposed
prototype passes all tests with a very high p-value,
with the lowest 0.231567 and the highest 0.845629.
The average p-value for all the tests is 0.45.
CONCLUSION
The research work in this article proposes an
integrated design of TRNG and PUF using PV
solar cells and microcontroller. We have shown
that the voltage response of PV solar cells can be
engineered in static (stable) and dynamic (large
variation) response. The segregation is based on
dividing the PV solar cell histogram within or
outside one SD (s) around the mean voltage
value (m). The proposed prototype uses IVN
structure which has 33.69% better throughput
to generate true random bits. The proposed prototype
achieves an average 92.13% reliability and
50.91% uniformity in PUF response. The integrated
TRNG-PUF architecture can be beneficial
in space-limited CPS.
ACKNOWLEDGMENTS
This work was supported by National Science
Foundation Grant: 1738662.
&REFERENCES
1. S. P. Mohanty, " Advances in transportation cyberphysical
system (t-CPS), " IEEEConsum. Electron.
Mag., vol. 9, no. 4, pp. 4-6, Jul. 2020.
2. S. P. Mohanty, " Consumer electronics is the driver of
smart cars notes from the editor, " IEEEConsum.
Electron. Mag., vol. 7, no. 5, pp. 3-55, Sep. 2018.
3. R. Alguliyev, Y. Imamverdiyev, and L. Sukhostat,
" Cyber-physical systems and their security issues, "
Comput. Industry, vol. 100, pp. 212-223, 2018.
4. M. Stipcevic and C¸ . K. Koc¸ , " True random number
generators, " in Open Problems in Mathematics and
Computational Science. Berlin, Germany: Springer,
2014, pp. 275-315.
5. G. Honan, N. Gekakis, M. Hassanalieragh, A. Nadeau,
G. Sharma, and T. Soyata, " Energy harvesting and
buffering for cyber physical systems: A review, "
Cyber-Phys. Syst., Comput. Perspective, 2015,
pp. 191-217.
IEEE Consumer Electronics Magazine
IEEE Consumer Electronics Magazine - July/August 2021

Table of Contents for the Digital Edition of IEEE Consumer Electronics Magazine - July/August 2021
