Analysis of RC6-Lite Implementation for Data Encryption

Amin Subandi

, Maya Silvi Lydia

and Rahmat Widia Sembiring

Faculty Computer and Information Technology, Universitas Sumatera Utara, Medan - Indonesia

Politeknik Negeri Medan, Medan – Indonesia

Keywords: Cryptography, RC6, RC6-Lite, Implementation, Encryption.

Abstract: RC6 is a simple modern cryptographic algorithm but can provide sufficient security for data encryption and

decryption, RC6 is a parameterized algorithm with RC6-w / r / b, where the suggested w values are 16, 32 and

64 and r values (rotation of encryption processing ) is 20, while b is the user-specified key length. In this

study, the authors are interested to implement with the Microsoft Visual Basic 2005 Express Edition

programming language and to determine the effectiveness of RC6 with the lowest parameters of the suggested

RC6 8/5 / b, and we call it with RC6-Lite. From the result of the research, it can be concluded that for the

change in plaintext, the average value of avalanche effect produced is 26.49% and for the changes made on

the key, the average value of avalanche effect produced better that can reach 52.38%. And this is of course

still better than the classic Playfair and Vigenere cipher algorithms which each have only avalanche effect of

10.9% and 3.1% (Saeed & Rashid, 2010).

1 INTRODUCTION

One of the techniques for securing data that can be

used is cryptographic techniques, cryptography is the

science and art to maintain the confidentiality and

security of information on a data using mathematical

techniques that are poured into the form of

algorithms.

Cryptography is only considered as a security of

electronic data communications, whereas since 2000

B.C. (Paar, 2010) where electronic data

communication has not existed, this cryptography

science has been applied.

In the days of ancient Rome, it was told that Julius

Caesar sent a secret message to a General on the

battlefield, so that to secure the information in the

message from the others during the delivery process,

the message was randomized so that only the General

was aware of the information in the message (Ariyus,

2008).

From its time, cryptography is divided into

modern and classical cryptographic algorithms.

Classical cryptography works in alphabet mode while

modern cryptography works in bit mode. While the

key, the cryptographic algorithm is divided into key

symmetric cryptography algorithms and asymmetric

keys. The symmetric key encrypts and decrypts a

message with the same key while the asymmetric uses

a different key in the encryption and decryption

process.

Rivest Code 6 (RC6) is a modern block cipher

cryptography algorithm which is the closest rival

algorithm of the Rijndael algorithm, Rijndael is the

winning algorithm of the new Advanced Encryption

Standard (AES) algorithm search competition held by

the National Institute of Standards and Technology

(NIST) this is held to look for new algorithms to

replace the DES algorithm. RC6 filed by Ronald Linn

Rivest et al of RSA Security Inc. In 1998. Famously

simple but has a pretty good security. The algorithm

is simple because in this algorithm contains six basic

primitive operations such as addition, subtraction,

XOR, multiplication and shifting bit either right or left

(Fishawy & Zaid 2007).

This algorithm has three processes, the process of

key expansion, encryption and the decryption

process. The key expansion process is the process to

generating S-box keys with user keys, while

encryption is the process of encoding messages with

S-box keys that have been generated from the key

expansion process, as well as the decryption process.

The RC6 algorithm is a parameterized code block

algorithm with the parameter RC6-w / r / b where w

is the size of the word block, r is the number of

encryption rotations whereas b is the key length

specified by the user. The suggested w values are 16,

32 and 64. The standard RC6 filed on NIST has RC6-

Subandi, A., Lydia, M. and Sembiring, R.

Analysis of RC6-Lite Implementation for Data Encryption.

DOI: 10.5220/0010037500420047

In Proceedings of the 3rd International Conference of Computer, Environment, Agriculture, Social Science, Health Science, Engineering and Technology (ICEST 2018), pages 42-47

ISBN: 978-989-758-496-1

32 / 20 / b parameters, each block having 32 bits and

encrypting 20 rounds. For the maximum quality of

safety parameters RC6 32/20/16 is recommended

(Ahmed et al, 2007) In the encryption process this

algorithm does it on every four plaintext blocks if in

the last phase of the encryption process is not

sufficient then padding bits on plaintext to four the

last plaintext block is fulfilled.

Much research has been done in implementing

this RC6 algorithm, and with the above parameters,

this algorithm is able to provide enough security.

However, for cryptanalysis against RC6 this is best

attacked by X2 (Terada & Ueda, 2009), and to

overcome this, Terada & Ueda (2009) do research by

modifying RC6 to RC6T, where they add T ( )

function, a simple additional swap on each round.

In this study, the authors are interested to analyze

and implement the RC6 algorithm on text data with

the lowest parameter and the author call it as RC6-

Lite is RC6 with a variant of parameter RC6-8 / 5 / b.

Thus, RC6-Lite will encrypt the characters (8 bits)

and 4 characters (32 bits) of processing not per four

characters (32 bits) per block and 16 characters (128

bits) per process as in the standard RC6, 5 times the

encryption process rotation is not 20 times. With the

RC6-Lite parameter how is its effectiveness ?, to

measure its effectiveness, the author will calculate the

value of avalanche effect generated from this RC6-

Lite algorithm.

2 METHODS

2.1 Key Expansion Procedure

This procedure is used for generating internal keys

based on the secret key K to populate the S table, the

number of S variables is 2 (r + 2). In populating the S

table, RC6 uses two magical constants Pw and Qw.

𝑃𝑤  𝑂𝑑𝑑



𝑒2



2



 (1)

𝑄𝑤  𝑂𝑑𝑑



∅1



2



 (2)

Where :

w = 16, 32 and 64

𝑒 = 2.718281828459 ... (logarithmic base) and

∅ = 1.6180339887... (golden ratio)

the Odd(x) function will produce the odd integer

closest to x.

Pada penelitian ini, nilai w = 8, sehingga nilai

konstanta Pw = 185 = b9 dan Qw = 159 = 9f

The key expansion procedure can be seen in the

following pseudocode:

S[0] = Pw

for i = 1 to (2r+3) do

{ S[i] = S[i-1] + Qw }

x, y, i, j = 0

for k = 1 to (3 x (2r+4)) do

{ S[i] = (S[i] + x + y) <<< 3

x = S[i]

L[j] = (L[i] + x + y) <<< 3

y = L[j]

i = (i+1) mod (2r+4)

j = (j+10 mod c

}

From the pseudocode above, in this study with the

value of r = 5, it will generate as many as 14 pieces

of the key S. In the standard RC6, the above process

will produce S key of 44 pieces. While c is the number

of keys entered by the user. This S key will be used

for the encryption and decryption process, where the

first two keys will be used for the initial whitening

process, and the last two keys are used for the final

whitening process and the rest is used for encryption

of r times.

2.2 Encryption and Decryption

Procedure

RC6-Lite algorithm in this research will break the

block data of 32 bit into 4 pieces of 8 bit block, then

this algorithm work with 4 8-bit registers of A, B, C

and D. In the process will be obtained (A, B, C, D) =

(B, C, D, A) which means that the value located on the

right side comes from the register on the left side. The

encryption and decryption process of the RC6

algorithm begins and ends with a whitening process

that aims to disguise the first and last iteration of the

encryption and decryption process, and in the process

requires 14 subkeys that have been generated in the

key expansion process and named with S [0 ] to S

[13] each of which is 8-bit in length. In each iteration

in the encryption process used 2 (two) subkeys. The

sub-keys in the first iteration use S [2] and S [3], the

next iteration uses sub-advanced subkeys.

Analysis of RC6-Lite Implementation for Data Encryption

For the encryption process can be clearly seen in

the following figure:

Figure 1: Encryption Scheme of RC6.

Pseudocode for encryption process can be seen as

follows:

B = B + S[ 0 ]

D = D + S[ 1 ]

for i = 1 to r do

{ t = ( B x ( 2B + 1 ) )<<< 5

u = ( D x ( 2D + 1 ) )<<< 5

A = (( A t )<<< u )+ S[ 2i ]

C = (( C u )<<< t )+ S[ 2i + 1 ]

(A, B, C, D) = (B, C, D, A)

}

A = A + S[ 42 ]

C = C + S[ 43 ]

Pseudocode for decryption process is described

below:

C = C - S[ 43 ]

A = A - S[ 42 ]

for i = r downto 1 do

{ (A, B, C, D) = (D, A, B, C)

u = ( D x ( 2D + 1 ) )<<< 5

t = ( B x ( 2B + 1 ) )<<< 5

C = ((C - S[ 2i + 1 ] )>>> t ) u

A = ((A - S[ 2i ] ) >>> u ) t

}

D = D - S[ 1 ]

B = B - S[ 0 ]

2.3 System Scheme

In implementing this RC6-Lite algorithm, the authors

use Microsoft Visual Basic 2005 Express Edition,

with the system scheme to be created can be seen in

the following figure:

Figure 2: System scheme in this research.

Plaintext and user-keys entered by the user, then

performed the process of key expansion to generate

the S-Key, then do the encryption or decryption

process with RC6 Lite to generate ciphertexts.

2.4 Comparison of Avalanche Effect

Avalanche effect is one of the ways to measure the

effectiveness of an algorithm, avalanche effect is the

bit changes that occur when plaintext or key changes,

it will produce different ciphertext bits, the greater the

level of its avalanche effect, the more difficult for the

cryptanalyst to do the attack.

The algorithm is good if the avalanche effect

value is 45 to 60%, to calculate the avalanche effect

value using the following formula:

𝑎𝑣𝑎𝑙𝑎𝑛𝑐ℎ𝑒 𝑒𝑓𝑓𝑒𝑐𝑡 

.   

  

𝑥 100 (3)

In this research will be adding and subtraction of

characters both on plaintext and key, and also

replacement of one character either plaintext or key.

3 RESULT AND DISCUSSIONS

As the research data, plaintext and key used are as

follows:

Plaintext :

THE KEY IS UNDER THE TABLE

Key :

THE KEY

Plaintext and Key

Key Ekspansion

Encrypt/Decrypt

RC6

Lite

Ciphertext

ICEST 2018 - 3rd International Conference of Computer, Environment, Agriculture, Social Science, Health Science, Engineering and

Technology

Based on the application that the authors created

using the RC6-Lite algorithm, the resulting ciphertext

is as follows:

Ciphertext :

™¢‹ FäƒÇ«<+Þ‰Ì.Æ¯ÕÔbZä> Ú&£l

Viewed from the resulting ciphertext, it appears

that the length of the ciphertext is longer than

plaintext, this is due to the addition of padding made

to plaintext which is not a multiple of 4 (RC6-Lite

encrypts every 4 characters), padding in this study is

done by adding the character " * " On plaintext. More

details can be seen in the following figure:

Figure 3: Encryption Application.

Figure 4: Comparison experiment.

The first test made a change to plaintext, that is by

removing the character "U" and still with the same

Key, the test results as shown there are images and

described as follows:

The first test:

Key : THE KEY

Plaintext1 : THE KEY IS UNDER THE TABLE

Bit Ciphertext1 :

10011001101000101000101110001101010

00110111001001000001111000111101010

11001111000010101111011110100010011

10011000010111011000110101011111101

01011101010001100010010110101110010

00011111010001101110110100010011010

10001101101100

Plaintext2 : THE KEY IS NDER THE TABLE

Bit Ciphertext2 :

10011001101000101000101110001101010

00110111001001000001111000111111010

00111111100100110001100010101001110

01111011101101001101011100110011010

00101000101110001101000011001100100

11011011000000001110101100110110001

00000111110000

The resulting bit length is 224 bits, with the

number of bits split is 84 bits thus the avalanche effect

percentage is 37.5%.

The second test:

Key :

THE KEY

Plaintext1 :

THE KEY IS UNDER THE TABLE

Bit Ciphertext1 :

10011001101000101000101110001101010

00110111001001000001111000111101010

11001111000010101111011110100010011

10011000010111011000110101011111101

01011101010001100010010110101110010

00011111010001101110110100010011010

10001101101100

Plaintext2 :

THE KEY ISS UNDER THE TABLE

Bit Ciphertext2 :

10011001101000101000101110001101010

00110111001001000001111000111110100

11111011111101011110000100000111000

11011000001011111011011110101111100

10010110101111111000101000100110010

00111111100110001000100110011110011

00001011111111

Analysis of RC6-Lite Implementation for Data Encryption

The resulting bit length is 224 bits, with the

number of bits split is 78 bits thus the avalanche effect

percentage is 34.82%.

The third test :

Key :

THE KEY

Plaintext1 :

THE KEY IS UNDER THE TABLE

Bit Ciphertext1 :

10011001101000101000101110001101010

00110111001001000001111000111101010

11001111000010101111011110100010011

10011000010111011000110101011111101

01011101010001100010010110101110010

00011111010001101110110100010011010

10001101101100

Plaintext2 :

THE REY IS UNDER THE TABLE

Bit Ciphertext2 :

10011001101000101000101110001101110

00101101100101011111100101110101010

11001111000010101111011110100010011

10011000010111011000110101011111101

01011101010001100010010110101110010

00011111010001101110110100010011010

10001101101100

The resulting bit length is 224 bits, with the

number of bits split is 16 bits thus the avalanche effect

percentage is 7.14%.

The fourth test:

Key1 :

THE KEY

Key2 :

THE EY

Plaintext1 :

THE KEY IS UNDER THE TABLE

Bit Ciphertext1 :

10011001101000101000101110001101010

00110111001001000001111000111101010

11001111000010101111011110100010011

10011000010111011000110101011111101

01011101010001100010010110101110010

00011111010001101110110100010011010

10001101101100

Bit Ciphertext2 :

00000101101001010100000100010010111

01001100101100011101001011100011110

00010100100100001101010101011111000

10001100001101000101111100011111110

10101100101000010111000101101000011

00010010111100110111000011111000001

11000111000100

The resulting bit length is 224 bits, with the

number of bits split is 119 bits thus the avalanche

effect percentage is 53.12%.

The fifth test:

Key1 :

THE KEY

Key2 :

THE KEYS

Plaintext1 :

THE KEY IS UNDER THE TABLE

Bit Ciphertext1 :

10011001101000101000101110001101010

00110111001001000001111000111101010

11001111000010101111011110100010011

10011000010111011000110101011111101

01011101010001100010010110101110010

00011111010001101110110100010011010

10001101101100

Bit Ciphertext2 :

11101000101011100011101111100110011

11001101111111101011100011110011001

10001100011010011000000011100011101

10011010110110010010011010010101011

01100111010100011111101010001101000

10010100001101000100010000010000001

01101110101111

The resulting bit length is 224 bits, with the

number of bits split is 110 bits thus the avalanche

effect percentage is 49,11%.

The sixth test:

Key1 :

THE KEY

Key2 :

THE REY

Plaintext :

THE KEY IS UNDER THE TABLE

Bit Ciphertext1 :

10011001101000101000101110001101010

00110111001001000001111000111101010

11001111000010101111011110100010011

10011000010111011000110101011111101

01011101010001100010010110101110010

00011111010001101110110100010011010

10001101101100

Bit Ciphertext2 :

11001010111001000001111110010011011

01100001110010001111000000101000001

00011001111001010011110101111001110

00011100110000000001110011100101000

00111100000101110010011001011010111

10010000000100110011001011100000100

10110010111101

ICEST 2018 - 3rd International Conference of Computer, Environment, Agriculture, Social Science, Health Science, Engineering and

Technology

The resulting bit length is 224 bits, with the

number of bits split is 123 bits thus the avalanche

effect percentage is 54.91%.

To prove that the application in this study can

restore plaintext as before, the simulation can be seen

in the following figure:

Figure 5: Decryption application.

4 CONCLUSIONS

From the result of research can be concluded that:

 Implementation done in this research using

Microsoft Visual Basic 2005 Express edition

application can encrypt and decrypt the data,

provided that the key entered by the user is

equal.

 For changes to plaintext, such as deletion,

addition and replacement of one character, the

average avalanche effect yielded 26.49%, this

result is even smaller than the resulting of

54.6% DES in the Saeed & Rashid (2010)

research and lost thinner than Blowfish 28.71%

in Ramanujam & Karuppiah (2011) research,

this may be caused by RC6 algorithm to block

each process, so the effect that occurs only on

the block of each process, in RC6-Lite only do

the process of encryption and decryption at 32

bits each, and RC6 encrypts every 128 bits.

Nevertheless, RC6-Lite is still better than the

7.8% SDES algorithm and from the classical

Playfair cryptography of 10.9% and Vigenere

3.1% (Saeed & Rashid, 2010).

 For changes to the Key, such as deletion,

addition, and replacement of one character, the

results show that the resulting avalanche effect

reaches an average value of 52.38%, this

proves that the key expansion process in this

RC6-Lite can play an important role in

encryption and decryption on this algorithm.

REFERENCES

Ahmed. H.E.H., Kalash, H.M. & Farag Allah, O.S., 2007.

Encryption Efficiency Analysis and Security Evaluation

of RC6 Block Cipher for Digital Images. Institute of

Electrical and Electronics Engineers (IEEE), DOI

10.1109/ICEE.2007.4287293.

Ariyus, D. 2008. Pengantar Ilmu Kriptografi, Teori

Analisis dan Implementasi. Penerbit ANDI:

Yogyakarta.

Fishawy, N. E. & Zaid, O. M. A. 2007. Quality of

Encryption Measurement of Bitmap Images with RC6,

MRC6 and Rijndael Block cipher algorithm.

International Journal of Network Security, 5(3): 241-

251.

Paar, C & Pelzl, J. 2010. Understanding Cryptography.

Springer: Verlag Berlin Heidelberg.

Ramanujam, S. & Karuppiah, M. 2011. Designing an

algorithm with high Avalanche Effect. International

Journal of Computer Science and Network Security

(IJCSNS). 11(1): 106-111.

Saeed, F. & Rashid, M. 2010. Integrating Classical

Encryption with Modern Technique. International

Journal of Computer Science and Network Security

(IJCSNS). 10(5): 280-285.

Terada, R. & Ueda, E. T. 2009. A New Version of RC6

algorithm, stronger against X

cryptanalysis.

Conference in Research and Practice in Information

Technology (CRPIT). Vol.98.

Analysis of RC6-Lite Implementation for Data Encryption