Ijett V68i1p204

Download as pdf or txt
Download as pdf or txt
You are on page 1of 4

International Journal of Engineering Trends and Technology (IJETT) – Volume 68 Issue 1- Jan 2020

Analysis and Synthesis of RSA Algorithm


using VHDL and Matlab
Ashwini Araballi, Zuhi Subedar
Assistant Professor& Electronics and Communication Department & VTU
Jain College of Engineering, Belagavi, Karnataka

Abstract — This paper represents the simulation


and synthesis of RSA algorithm using MATLAB and
VHDL. RSA is considered as one of the hardest
algorithms as it is difficult to crack the standards
used in the encryption for securely storing and
transmitting the information. The analysis of this
algorithm is carried out for the parameters like
overall execution time, operating frequency and
Fig 1: Cryptosystem
resource consumed during simulation. The result
analysis depicts the comparison between RSA For example, consider two users Bob and
implementation using Matlab and VHDL. Alice. Let [PRa, PUa] and [PRb, PUb] be the private
keys and public keys of Alice and Bob respectively.
Keywords — Cryptography, Encryption, Decryption, There are two ways in which the Public Key
RSA, Security, Matlab, VHDL. Cryptography can be represented namely encryption
with private key and encryption with public key. If
I. INTRODUCTION Bob uses Alice’s public key to encrypt the data to be
The amount and means of data transmitted over transmitted, then at the receiver end, Alice can
the transmission channel has changed a lot since the decrypt the message using its own private key and
last few decades. Especially the information the scenario is depicted in Fig. 2. If Bob uses its
transmitted over the media electronically has grown private key to encrypt the message then Alice can
fast and even growing exponentially [1][2]. As a decrypt the message using Bob’s public key and the
result of this, it becomes very essential to find new scenario is depicted in Fig. 3.
ways for guaranteeing the information security
during communication. An art of keeping the RSA algorithm is one of the public key
information secure is called cryptography and the cryptographic algorithms invented by Rivest, Adi
basic model for the same is depicted in Fig. 1. This Shamir and Leonard Adleman [5]. It is considered
scheme helps the people to carry out their business as a highly secured algorithm when sufficiently long
routines electronically such as emails, e-commerce, keys are used such as 1024 bits and above [4]. The
ATM machines etc without the fear of deception and basic motto of this paper is to implement RSA
deceit. It not only assures message integrity but also algorithm in MATLAB and VHDL and analyse the
ensures the authenticity of the sender [3]. performance of the same in both the cases using
softwares/platforms.
In order to provide high performance security, it
becomes necessary to use encryption schemes such
as symmetric with private key encryption algorithms
and asymmetric with public key encryption
algorithms. A symmetric encryption algorithm
makes use of one common key for both encryption
and decryption at both sender and receiver sides. On
the contrary, public key cryptosystem makes use of
two keys namely private key and public key; one of
which is used for encryption and other for
decryption i.e two pairs of keys are generated for
both the users.
Fig 2: Encryption with Public Key

ISSN: 2231-5381 http://www.ijettjournal.org Page 30


International Journal of Engineering Trends and Technology (IJETT) – Volume 68 Issue 1- Jan 2020

exponentiation performs the mathematical operation


to generate ciphertext. It incorporates encryption
with public key cryptographic scheme [9][11].

Fig 5: Encryption
c) Decryption Unit
Fig 3: Encryption with Private Key This unit again is same as encryption unit as shown
in Figure 6. The block does same mathematical
II. METHODOLOGY operation as encryption unit i.e. modular
exponentiation. It generates the plaintext back at the
A. RSA Algorithm receiver end.

In 1977 Ron Rivest, Adi Shamir and Adleman


proposed this method based on the Diffie Hellman
key exchange algorithm. The difficulty in factorizing
the large integers along with the use of sufficient
length of the key assures highest security to this
algorithm. Also, difficulty in guessing the prime Fig 6: Decryption
numbers and even choosing large prime numbers TABLE I: STEPS OF RSA ALGORITHM
makes factorization very difficult[6][7]. 1 Select two prime numbers p and q
RSA structure combines three following elements: 2 Calculate n=p * q
 Key generation
 Encryption 3 Calculate ∅(n)=(p-1)*(q-1)
 Decryption 4 Let (Public Key) e so that gcd (∅(n),e)=1

a) Key Generation Unit 5 Calculate d*e=1 mod ∅(n) we get (private key)d
6 Compute ciphertext C=M e mod n for plaintext M
This module is used for the generation of the private
key and public key that are to be used in Public Key 7 Compute plaintext M=C d mod n using ciphertext C
C cryptographic algorithms. The prime numbers, The above Table 1 depicts the overall steps included
denoted by p and q are multiplied by the multiplier in RSA algorithm.
and mod of the product is calculated. Then
subtractor and the second multiplier block calculate III. IMPLEMENTATION
Euler’s totient function. GCD calculator checks The RSA algorithm is implemented in Matlab
whether е and ∅(n) are prime or not relatively. coding and in VHDL coding by using Matlab tool
Another calculator, called multiplicative inverse is and Xilinx ISE 13.2.The parameters such as
used to calculate d, which is nothing but private key encryption decryption time i.e, Overall execution
[8]. time, operating frequency and memory (CPU time)
are taken for the analysis in both the cases.

Case i: MATLAB Implementation of RSA


In this case the RSA algorithm is analysed and
simulated using Matlab. The parameters like, overall
execution time which includes encryption and
decryption time, the operating frequency and total
memory used are analysed and the readings are
tabulated as shown in Table 2. For some fixed length
of prime numbers the factorization is a little simple
Fig 4: Key Generation process. But for random / larger lengths of primes
b) Encryption Unit the factorization becomes complex and consumes
relatively more time compared to fixed ones. In-
This unit makes use of modular exponentiation to short, larger the prime number selected, the complex
perform encryption as shown in Figure 5. Modular the factorization becomes.

ISSN: 2231-5381 http://www.ijettjournal.org Page 31


International Journal of Engineering Trends and Technology (IJETT) – Volume 68 Issue 1- Jan 2020

TABLE II: PARAMETRIC RECORDS USING MATLAB


FOR FIXED KEY LENGTH

Figure 7: Plot for message length v/s Total execution


time

From the above results it is concluded that the


execution time taken for Matlab is very less as
compared to VHDL and the analysis plot is
represented by Figure 8. Similarly it is known to us
that the operating frequency is inversely proportional
to the execution time i.e, as the execution time
decreases operating frequency increases. Also from
Figure 9 it is concluded that the memory consumed
Figure 6: Plot of Message length v/s Execution time
in Matlab is very high as compared to VHDL
with fixed and variable key sizes
synthesis report. Figure 10 depicts the simulation
As the length of message increases, the time output of the algorithm with some initialization
required for the encryption and decryption also values.
increases as illustrated in Figure 6. However,
frequency of operation decreases and memory
consumption increases.

Case ii: Xilinx ISE 13.2


In this case, the RSA algorithm is analysed and
synthesized using Xilinx ISE 13.2[10]. The
parameters like overall execution time, operating
frequency and total memory used are observed and
the readings are represented in Table 3. As the size
of message increases the time required for the
encryption and decryption increases as illustrated in
Figure 7, frequency of operation decreases and
memory consumption increases. Here the execution Figure 8: Plot for Total execution time analysis
time increases because of the Modular Exponentiator
and multipliers during encryption and decryption
processes. It follows radix -2 Montgomery
multiplications.

TABLE III: PARAMETRIC RECORDS USING XILINX ISE


(VHDL)

Figure 9: Plot for Memory used analysis

ISSN: 2231-5381 http://www.ijettjournal.org Page 32


International Journal of Engineering Trends and Technology (IJETT) – Volume 68 Issue 1- Jan 2020

networks, Communication and Mobility, Procedia


Computer Science 34, pp. 647 – 654, 2014.
[9] Vibhor G, Aruna c, “Architectural analysis of RSA
cryptosystem on FPGA “, International Journal of
Computer Applications, vol. 26, 2011.
[10] Chiranth E, Chakravarthy H.V.A, Nagamohanareddy P,
Umesh T.H, Chethan Kumar M,“Implementation of RSA
Cryptosystem Using Verilog”, International Journal of
Scientific & Engineering Research, vol. 2, Issue 5, pp.1-
7,2011.
[11] Amer, K.M, Sharif, S.M, Ashur, A.S., "Enhancement of
hardware modular multiplier radix-4 algorithm for fast
RSA cryptosystem," International Conference on
Computing, Electrical and Electronics Engineering, vol. 1,
pp. 692-696, Aug. 2013.
[12] S.Syamkumar Ch.Umasankar, “A Novel methodology for
Implementation RSA algorithm using FFT in Galois
Field”, International Journal of Engineering Trends and
Figure 10: Snapshot of VHDL simulation output
Technology (IJETT), vol. 30, December 2015.
[13] Fathima Nizar, Fathima Latheef, Alfina Jamal, “RSA
IV. CONCLUSION Based Encrypted Data Embedding Using APPM”,
The Public key cryptographic algorithm RSA is International Journal of Engineering Trends and
Technology (IJETT), vol. 9, Mar 2014.
synthesized and simulated successfully in MATLAB
and XILINX ISE 13.2 using VHDL. From the
analysis and synthesis reports it is concluded that
Matlab execution of RSA takes very less time as
compared to Xilinx because VHDL Xilinx makes
use of modular exponentiation and multiplication
blocks during encryption and decryption. Also the
operating frequency is high as compared to Xilinx
and the memory consumed by Matlab is high
compared to Xilinx because the number of processes
followed in the implementation of RSA is very
lengthy. Almost 83% of the CPU space is consumed
by Matlab and 17% is consumed by Xilinx. The
speed of implementation of RSA in MATLAB is
97% and in case of Xilinx is only 3%.

REFERENCES
[1] Sushanta Kumar Sahu and Manoranjan Pradhan, “FPGA
Implementation of RSA Encryption system”, International
Journal of Computer Applications, vol. 19, Issue 9, 2011.
[2] Ari Shawkat Tahir, “Design and Implementation of RSA
Algorithm using FPGA”, Research Gate, 2015.
[3] Sandeep Singh, Parminder Singh Jassal, “Synthesis and
Analysis of 32-Bit RSA Algorithm Using VHDL”,
International Journal of Engineering Sciences, vol. 1, Jan.
2016.
[4] Ankit A. and Pushkar P, “Implementation of RSA
Algorithm on FPGA”, International Journal of Engineering
Research & Technology (IJERT), ISSN: 2278-0181, vol.
1, Issue 5, 2012.
[5] Amit Thobbi, Shriniwas Dhage, Pritesh Jadhav and
Akshay Chandrachood,“Implementation of RSA
Encryption Algorithm on FPGA”, American Journal of
Engineering Research (AJER), vol. 4, Issue 6, pp-144-
151,2015.
[6] Khaled Shehata, Hanady Hussien, Sara Yehia, “FPGA
Implementation of RSA Encryption Algorithm for E-
Passport Application”, World Academy of Science,
Engineering and Technology International Journal of
Computer and Information Engineering, vol. 8, 2014.
[7] Gurpreet K. and Vishal A, “An Efficient Implementation
of RSA Algorithm using FPGA and Big Prime Digit”,
International Journal of Computer & Communication
Engineering Research (IJCCER), Volume 1 - Issue 4,
2013.
[8] M. A. Smadi, Qasem A. and Abdullah A, “Efficient FPGA
Implementation of RSA Coprocessor Using Scalable
Modules”, International Symposium on Emerging Inter-

ISSN: 2231-5381 http://www.ijettjournal.org Page 33

You might also like