One to many (new scheme for symmetric cryptography)

TELKOMNIKA Telecommunication, Computing, Electronics and Control
Vol. 21, No. 4, August 2023, pp. 762∼770
ISSN: 1693-6930, DOI: 10.12928/TELKOMNIKA.v21i4.24789 ❒ 762
One to many (new scheme for symmetric cryptography)
Alz Danny Wowor1
, Bambang Susanto2
1Department of Informatics Engineering, Faculty of Information Technology, Satya Wacana Christian University, Salatiga, Indonesia
2Department of Mathematics, Faculty of Science and Mathematics, Satya Wacana Christian University, Salatiga, Indonesia
Article Info
Article history:
Received Jun 19, 2020
Revised Oct 07, 2020
Accepted Feb 08, 2021
Keywords:
Cryptanalyst
One-to-many scheme
Symmetric cryptography
ABSTRACT
Symmetric cryptography will always produce the same ciphertext if the plaintext and the
given key are the same repeatedly. This condition will make it easier for cryptanalysts
to perform cryptanalysis. This research introduces a one-to-many cryptography scheme,
which can produce different ciphertexts even if the input given is the same repeatedly.
The one-to-many encryption scheme can produce several ciphertexts with differences of
up to 50%. The avalanche effect test obtained an average of 52.20%, better than mod-
ern cryptography Blowfish by 25.46% and 6% better than advanced encryption standard
(AES). One-to-many can produce different n-ciphertexts, which will certainly make it
more difficult for cryptanalysts to perform cryptanalysis and require n-times longer to
break than other symmetric cryptography.
This is an open access article under the CC BY-SA license.
Corresponding Author:
Alz Danny Wowor
Department of Informatics Engineering, Faculty of Information Technology, Satya Wacana Christian University
Jl. Notohamidjojo 1-10, Blotongan, Salatiga, Indonesia
Email: alzdanny.wowor@uksw.edu
1. INTRODUCTION
The need for information security becomes an extraordinary transformation in the field of cryptography.
Cryptographers create algorithms to secure information by emphasizing the strength of mathematics in taking
advantage of the complexity of algorithms and key secrecy. The more complex the algorithm design, the better the
information security. On the other hand, the speed of the process in algorithm is also taken into account. This is
the focus among cryptographers, namely how to create a powerful algorithm against the attack of cryptanalysts yet
has a fast processing time.
Symmetric cryptography is still an interesting study and is still very feasible to study [1]. One of them is
the advanced encryption standard (AES) has become a popular criterion and is widely used commercially. There
are several studies [2]-[14] that use algorithms recommended by the National Institute of Standards and Technology
(NIST) as the information security, although several studies have been claimed to break the algorithm [15]-[20].
Today’s advancement of computer technology also has created a challenge for the development of cryp-
tography. The idea of quantum computation can enhance various cryptanalysis techniques because quantum me-
chanical system calculation will produce a much less time complexity than that of classical calculation model [21],
[22]. Computational advancement certainly encourages cryptographers to review the cryptographic schemes that
will be used.
There a lot of methods that can be done to make an algorithm to be more resistant to attacks. Henriques
and Vernekar [23] combined symmetric and asymmetric algorithms to secure communication between devices in
an IoT system. In [24] proposed the use of acoustic echo cancellation (AEC) and elliptic curve cryptography
(ECC) encryption techniques with a combination of symmetric and asymmetric algorithms. In [25] modified shift
row and mix column in AES, so that the processing time can be faster. In [26] developed a cryptographic technique
Journal homepage: Journal homepage: http://telkomnika.uad.ac.id

TELKOMNIKA Telecommun Comput El Control ❒ 763
based on quantum key distribution (QKD). In [27] designed and implemented a real-time cryptographic algorithm
to secure medical devices that provide a high level of security for remote health monitoring systems.
The existing symmetrical cryptography is the type of one-to-one encryption scheme, which means that
every input plaintext and key always produces the same ciphertext every time a process is carried out. In solving
of cryptography, a criptanalyst will certainly try to look for patterns from percipext and makes an analysis of keys
and plaintexts. The existing cryptography algorithms sometimes at certain parts are very vulnerable to extreme
testing and as a result the patterns in ciphertext will be easier to be found.
This research designs a new scheme that can answer that problem, which is the one-to-many scheme.
This scheme will create different ciphertext even though the input plaintext and keys are always the same. Any
changes that occur in ciphertext will take longer time and certainly will also make it difficult for cryptanalysts to
see the relationship between plaintext and ciphertext. Thus, this scheme certainly is more secure than one-to-one
encryption scheme. The initial design of this scheme is more on how to get several balanced ciphertext than on the
complexity of algorithm in the encryption-decryption process.
2. PROPOSED RESEARCH
2.1. One to many scheme
One-to-many is an encryption scheme that can create different chipertexts, although using the same plain-
text and key, and decryption process that can retrieve the plaintext. Etymologically, one-to-many consists of two
words, “one” is a plaintext, and “many” is a process generating a different ciphertext for each time it is encrypted.
EK(pa) = cai
(1)
The one-to-many scheme encryption is described in (1). The encryption (Ek) uses plaintext (pa) and key
as inputs, which can create ciphertext cai , i ∈ Z+
. All encryption processes can create different cai although with
the same input.
DK(cai ) = pa (2)
In contrast, the process of DK decryption function in (2) uses a key and ciphertext as inputs. A one-to-
many scheme should be able to return each different ciphertext (cai
) resulting from the encryption process to be
a plaintext. This study uses a random number generator to show that the one-to-many scheme can be applied in
cryptographic processes as shown in Figure 1.
2.2. Implementation of one-to-many scheme
The design of one-to-many algorithm scheme is shown in Figure 1. The plaintext input (P) and key
(K) are in the form of text. K is processed into seed (x0) in cryptographically secure pseudo random number
generator (CSPRNG) chaos which will produce n-number of random numbers sequence {A1, A2, · · · , An}. The
chose key process scheme will manage and select the compatible key for the encryption-decryption process so that
it produces the appropriate plaintext.
K
Examination
Scheme
F(x0)
Chose Key Scheme
Ai,
i ∈ {1, 2, · · · , n}
A∗
i ⊕ P C
P
x0 A∗
i
Figure 1. General scheme of one to many encryption
K = {k1, k2, · · · , kn} is the symbol of the key and P = {p1, p2, · · · , pn} is the symbol of plaintext.
Each ki ∈ K plays role as input to the examination scheme [28], and produces x0 as the seed for the CSPRNG
chaos function (F).
F(x0) = {A1, A2, · · · An} (3)
Each Ai, i = 1, 2, · · · , n is the data set of CSPRNG chaos random numbers, where A1 = {a11, a12, · · · , a1k},
A2 = {a21, a22, · · · , a2k}, · · · , An = {an1, an2, · · · , ank}. From (3), using the chosen key A∗
n from the key
selection scheme, the encryption process is obtained as:
EK → P ⊕ A∗
i = {c1 = p1 ⊕ a∗
1r, c2 = p2 ⊕ a∗
2r, · · · , cn = pn ⊕ a∗
nr} (4)
One to many (new scheme for symmetric cryptography) (Alz Danny Wowor)

764 ❒ ISSN: 1693-6930
In the general, the following equation is obtion:
ci = pi ⊕ a∗
ir (5)
The decryption process can also be done by generating each random number using key input as well.
Let K = {k1, k2, · · · , kn} as the key, and ciphertext C = {c1, c2, · · · , cn}. Each random number from
CSPRNG chaos is obtained from F(K) = {A1, A2, · · · An}. The one-to-many scheme will select A∗
i as the
chosen key for the decryption process.
DK → C ⊕ A∗
i = {p1 = c1 ⊕ a∗
1r, p2 = c2 ⊕ a∗
2r, · · · , pn = cn ⊕ a∗
nr} (6)
So that plaintext will be recovered as a decryption process.
pi = ci ⊕ a∗
ir (7)
2.3. Key generation
The key generation starts from the examination process [28], each key input (ki) and the index constant
(bi), where ki, bi ∈ Z256
for i = 1, 2, · · · , 8 which is used to find the ratio x0 = pa/qa, where pa and qa are
obtained from (8) and (9).
pa = (k1 + k2 + k3 + k4 + k5 + k6)/6 =
6
X
i=1
ki/6 (8)
qa = k1b1 + k2b2 + k3b3 + k4b4 + k5b5 + k6b6 + k7b7 + k8b8 =
8
X
i=1
kibi (9)
The value of x0 is used as the seed in (10) that generates a CSPRNG chaos-based number serial. This
process is in accordance with the search function F(x0) as shown in Figure 1. The random number will be used
as a key in the encryption or decryption process.
xn+1 = txn(1 − xn) (10)
Each chaotic number is in decimal form and cannot be used as a key, so it is converted to an integer using
the truncation function in (11).
T(x, size) = ∥x × 10count
∥, x ̸= 0 (11)
The “chose key scheme” is the most important process, because here algorithm will choose the key avail-
ability and set which key will be selected as the chosen key A∗
n to be used in the encryption process in (4) or
decryption in (6). The designed algorithm can produce several different and balanced key sets {A1, A2, · · · , An}
as shown earlier in (3). After selecting a set which will be used as the key, algorithm will provide a marker that is
included in the ciphertext, so that in the process of decryption, algorithm can recognize the marker and will refer
to the set of keys used.
3. RESULT AND ANALYSIS
3.1. Test of the key changes
Test of key changes in one-to-many scheme is carried out to see how well an algorithm produces ciphertext
if there is a small change (1-bit change) in the key. The plaintext “satya wacana” and two keys “fti uksw” and “ftj
uksw” are selected, where the change from the letter i to j is a 1-bit difference. This test also uses the correlation
statistical test and the mean difference, to see the difference between each ciphertext produced.
TELKOMNIKA Telecommun Comput El Control, Vol. 21, No. 4, August 2023: 762–770

3. RESULT AND ANALYSIS
3.1. Test of The Key Changes
i xi bil-1 bil-2 bil-3 bil-4 bil-5
1 0.099998932494109 099 998 932 494 109
2 0.359996583976590 359 996 583 976 590
3 0.921596174007104 921 596 174 007 104
4 0.289026664250287 289 026 664 250 287
5 0.821961006410555 821 961 006 410 555
6 0.585364441404410 585 364 441 404 410
7 0.970851648574852 970 851 648 574 852
(a)
0 100 200
0
0.5
1
i
x
i
(b)
Figure 2. Logistic function iteration with “fti uksw” key. (a) Data of the first 7 iteration, (b) The first 200 Iteration
On to Many Encryption (Alz Danny Wowor)
Figure 2. Logistic function iteration with “fti uksw” key: (a) data of the first 7 iteration and (b) the first 200
iteration
The first test uses the key “fti uksw”, and the index value B = 1, 2, · · · , 8 is selected. Based on (8) and
(9), x0 = 0.0256580696623239 is obtained. To generate a random value of CSPRNG chaos, value t = 4 is used
in (10). The results of the first 7 iteration values are shown in Figure 2(a) and the results of the first 200 iterations
are shown in n Figure 2(b). In cryptography, keys must be in integer number. Therefore, the result of iteration is
divided to be five integers, i.e., set-i, {1, 2, 3, 4, 5} respectively. Each set-i is Ai in (3), which is used as the key for
encryption-decryption processes in (5) and (7).
The scatter plot in Figure 2(b) is a visualization of the first 200 iterations, which proves that the logistic
function with the key input “fti uksw” can generate random numbers. Based on the table in Figure 2(a), there are
five datasets that can be used as keys. Using (3), we obtain A1 = set-1, A2 = set-2, · · · , A5 = set-5.
Table 1. Key test “fti uksw”
i-th
Ciphertext-i
Value Category Mean
encryptions correlation correlation difference
1 D6C80D9A966941D2F422035C0720232E20 0.393295 Fair 62.06
2 5945C893228CCA237235F29822B3B46C21 0.086143 Very low 68.69
3 17A8221167D9FFE5653E3A00D8F456B922 −0.515177 Fair 99.81
4 61317B73FBB4B5EAD47BFB9FAF479B2523 0.232599 Low 77.00
5 E0AFDC998EB4CFCA719639F391581FC824 0.343136 Low 82.06
Based on (4), the key selection process can be carried out. The selection scheme will choose the key that
will be used in the encryption process. Table 1 shows the result of five encryptions using the key “fti uksw”. These
results indicate that one-to-many scheme can produce different ciphertext for each encryption process. The differ-
ence between each ciphertext and plaintext can be seen using the correlation value, which respectively there are
two ciphertexts that have relationship with plaintexts in moderate and low categories, and one ciphertext that has a
relationship with the very low category.
Mean difference is the average absolute difference between plaintext and ciphertext. The greater the
distance between plaintext and ciphertext, the better it will be. The results obtained in Table 1 show the mean
difference value of different ciphertexts. These results provide information that one-to-many scheme can create
different ciphertext for each encryption. Thus, cryptanalysts will need more work to be able to solve this algorithm.
4 r ISSN: 1693-6930
i xi bil-1 bil-2 bil-3 bil-4 bil-5
1 0.099998932494109 099 998 932 494 109
2 0.359996583976590 359 996 583 976 590
3 0.921596174007104 921 596 174 007 104
4 0.289026664250287 289 026 664 250 287
5 0.821961006410555 821 961 006 410 555
6 0.585364441404410 585 364 441 404 410
7 0.970851648574852 970 851 648 574 852
(a)
0 100 200
0
0.5
1
i
x
i
(b)
On to Many Encryption (Alz Danny Wowor)
The first test uses the key “fti uksw”, and the index value B = 1, 2, · · · , 8 is selected. Based on Equation
8 and Equation 9, x0 = 0.0256580696623239 is obtained. To generate a random value of CSPRNG chaos, value
t = 4 is used in Equation 10. The results of the first 7 iteration values are shown in Figure 3(a) and the results of
the first 200 iterations are shown in n Figure 3(b). In cryptography, keys must be in integer number. Therefore, the
result of iteration is divided to be five integers, i.e. set-i, {1, 2, 3, 4, 5} respectively. Each set-i is Ai in Equation 3,
which is used as the key for encryption-decryption processes in Equation 5 and Equation 7.
The scatter plot in Figure 3(b) is a visualization of the first 200 iterations, which proves that the logistic
function with the key input “fti uksw” can generate random numbers. Based on the table in Figure 3(a), there are
five datasets that can be used as keys. Using Equation 3, we obtain A1 = set-1, A2 = set-2, · · · , A5 = set-5.
Table 1. Key Test “fti uksw”
i-th
Ciphertext-i
Value Category Mean
encryptions Correlation Correlation Difference
1 D6C80D9A966941D2F422035C0720232E20 0.393295 Fair 62.06
2 5945C893228CCA237235F29822B3B46C21 0.086143 Very Low 68.69
3 17A8221167D9FFE5653E3A00D8F456B922 0.515177 Fair 99.81
4 61317B73FBB4B5EAD47BFB9FAF479B2523 0.232599 Low 77.00
5 E0AFDC998EB4CFCA719639F391581FC824 0.343136 Low 82.06
Based on Equation 4, the key selection process can be carried out. The selection scheme will choose
the key that will be used in the encryption process. Table 1 shows the result of five encryptions using the key
“fti uksw”. These results indicate that one-to-many scheme can produce different ciphertext for each encryption
process. The difference between each ciphertext and plaintext can be seen using the correlation value, which
respectively there are two ciphertexts that have relationship with plaintexts in moderate and low categories, and
one ciphertext that has a relationship with the very low category.
Mean Difference is the average absolute difference between plaintext and ciphertext. The greater the
distance between plaintext and ciphertext, the better it will be. The results obtained in Table 1 show the mean
difference value of different ciphertexts. These results provide information that one-to-many scheme can create
different ciphertext for each encryption. Thus, cryptanalysts will need more work to be able to solve this algorithm.
0 2 4 6 8 10 12 14 16
0
100
200
Index of character
ASCII
equivalent
value
plaintext
ciphertext 1
ciphertext 2
ciphertext 3
ciphertext 4
ciphertext 5
Figure 4. Comparison plot of plaintext and ciphertext for key “fti uksw”
Visualization of each ciphertext of the key “fti uksw” in Cartesian coordinates in Figure 5 shows a sequence
of different numbers and do not form a pattern that facilitates direct guessing. It appears that each ciphertext ob-
tained is different although with the same input. This result indicates that one-to-many scheme can make plaintext
and ciphertext that do not have a strong relationship statistically.
TELKOMNIKA Telecommun Comput El Control, Vol. x, No. x, February 202x : xx – xx
Visualization of each ciphertext of the key “fti uksw” in Cartesian coordinates in Figure 3 shows a se-
quence of different numbers and do not form a pattern that facilitates direct guessing. It appears that each ciphertext

766 ❒ ISSN: 1693-6930
obtained is different although with the same input. This result indicates that one-to-many scheme can make plain-
text and ciphertext that do not have a strong relationship statistically.
0 2 4 6 8 10 12 14 16
Index of character
i xi set-1 set-2 set-3 set-4 set-5
1 0.100089355076193 100 089 355 076 193
2 0.360285904306499 360 285 904 306 499
3 0.921919885858189 921 919 885 858 189
4 0.287934439669651 287 934 439 669 651
5 0.820112792487100 820 112 792 487 100
6 0.590111200344443 590 111 200 344 443
7 0.967519886289935 967 519 886 289 935
(a)
0 100 200
0
0.5
1
i
x
i
(b)
Figure 4. Logistic function iteration with “ftj uksw” key. (a) Data of the first 7 iteration, (b) The first 200 Iteration
Figure 4. Logistic function iteration with “ftj uksw” key: (a) data of the first 7 iteration and (b) the first 200
iteration
The second test using the key “ftj uksw”, and using the same index value B = 1, 2, · · · , 8, x0 =
0.0256818986893376 is obtained. The results of the first 7 iteration values are shown in Figure 4(a) and the
results of the first 200 iterations are shown in n Figure 4(b). Taking the value of t = 4, the iteration results are
obtained as shown in Figure 4(a). Five datasets A1 = set-1, A2 = set-2, · · · , A5 = set-5 are used as keys in the
encryption-decryption processes.
The iteration result of the logistic function with the “ftj uksw” key is very different from the “fti uksw” as
shown in Figure 2(b) and Figure 4(b). These results indicate that the examination algorithm is designed to be able
to produce a unique initialization value of x0, so it is very sensitive towards small (1-bit) changes in the input. The
nature of the butterfly effect, which has a large effect on output even though it is a small change in input, also has a
significant effect on the encryption results. This can be seen in the differences of each ciphertext as shown in Table
1 and Table 2.
Table 2. Key test “ftj uksw”
i-th
Ciphertext-i
Value Category Mean
encryptions correlation correlation difference
1 D7C90D98956E3EDE1A3AA739C787E0FD20 −0.39498 Fair 96.38
2 CC7E0B1FD18F7E1DBAF874AB3EC9CC1F21 −0.06211 Very low 81.31
3 D6E9E93079E8EDD0EE9C26BCAB49093222 0.342195 Low 91.25
4 BF93CE1648789816ADBBB0B59A89119323 0.182852 Very low 75.44
5 34533007CBCC284C4A65C235F807F6DC24 −0.58524 Fair 89.81
Table 2 shows the test results using the key “ftj uksw”, which it indicates that a one-to-many scheme can
produce five different ciphertexts at each encryption process. Statistically, the resulting correlation value is close
to 0 with sufficient and very low categories. It indicates that the algorithm in making of plaintext and ciphertext
has no statistical relationship. The visualization in Figure 5 also shows that each ciphertext produced is different
although with the same input. Test of 1-bit key difference shows that one-to-many scheme can produce different
ciphertext, even though given key inputs with a small different. The visualization of the graphic shows a different
ciphertext which it produces a different sequence of numbers and does not form a pattern that facilitates direct
guessing, and as a result, cryptanalysts will need more work to solve this algorithm.
TELKOMNIKA Telecommun Comput El Control r 5
i xi set-1 set-2 set-3 set-4 set-5
1 0.100089355076193 100 089 355 076 193
2 0.360285904306499 360 285 904 306 499
3 0.921919885858189 921 919 885 858 189
4 0.287934439669651 287 934 439 669 651
5 0.820112792487100 820 112 792 487 100
6 0.590111200344443 590 111 200 344 443
7 0.967519886289935 967 519 886 289 935
(a)
0 100 200
0
0.5
1
i
x
i
(b)
Figure 5. Logistic function iteration with “ftj uksw” key. (a) Data of the first 7 iteration, (b) The first 200 Iteration
The second test using the key “ftj uksw”, and using the same index value B = 1, 2, · · · , 8, x0 =
0.0256818986893376 is obtained. The results of the first 7 iteration values are shown in Figure 5(a) and the
results of the first 200 iterations are shown in n Figure 5(b). Taking the value of t = 4, the iteration results are
obtained as shown in Figure 5(a). Five datasets A1 = set-1, A2 = set-2, · · · , A5 = set-5 are used as keys in the
encryption-decryption processes.
The iteration result of the logistic function with the “ftj uksw” key is very different from the “fti uksw” as
shown in Figure 3(b) and Figure 5(b). These results indicate that the examination algorithm is designed to be able
to produce a unique initialization value of x0, so it is very sensitive towards small (1-bit) changes in the input. The
nature of the butterfly effect, which has a large effect on output even though it is a small change in input, also has a
significant effect on the encryption results. This can be seen in the differences of each ciphertext as shown in Table
1 and Table 2.
Table 2. Key Test “ftj uksw”
i-th
Ciphertext-i
Value Category Mean
encryptions Correlation Correlation Difference
1 D7C90D98956E3EDE1A3AA739C787E0FD20 0.39498 Fair 96.38
2 CC7E0B1FD18F7E1DBAF874AB3EC9CC1F21 0.06211 Very Low 81.31
3 D6E9E93079E8EDD0EE9C26BCAB49093222 0.342195 Low 91.25
4 BF93CE1648789816ADBBB0B59A89119323 0.182852 Very Low 75.44
5 34533007CBCC284C4A65C235F807F6DC24 0.58524 Fair 89.81
Table 2 shows the test results using the key “ftj uksw”, which it indicates that a one-to-many scheme can
produce five different ciphertexts at each encryption process. Statistically, the resulting correlation value is close
to 0 with sufficient and very low categories. It indicates that the algorithm in making of plaintext and ciphertext
has no statistical relationship. The visualization in Figure 6 also shows that each ciphertext produced is different
although with the same input.
0 2 4 6 8 10 12 14 16
0
100
200
Index of character
ASCII
equivalent
value
plaintext
ciphertext 1
ciphertext 2
ciphertext 3
ciphertext 4
ciphertext 5
Figure 6. Comparison plot of plaintext and ciphertext for key “ftj uksw”
Test of 1-bit key difference shows that one-to-many scheme can produce different ciphertext, even though
given key inputs with a small different. The visualization of the graphic shows a different ciphertext which it
produces a different sequence of numbers and does not form a pattern that facilitates direct guessing, and as a
result, cryptanalysts will need more work to solve this algorithm.
3.2. Bit Differences On Each Ciphertext
Figure 5. Comparison plot of plaintext and ciphertext for key “ftj uksw”

3.2. Bit differences on each ciphertext
The next test is to see the bit difference on each ciphertext. The bigger the bit difference for each ci-
phertext, the better it is. The one-to-many scheme produces five ciphertexts, so that the number of comparisons
for each ciphertext is C5
2 = 10. Table 3 shows the comparison of ciphertexts for each key. With m, n to be the
selected ciphertext, m = n ∈ {1, 2, 3, 4, 5}. The symbol “><” denotes the comparison of each ciphertext.
Table 3. Bit differences on each ciphertext
No Ciphertext m >< Ciphertext n “fti uksw” “ftj uksw”
1 1 >< 2 63 72
2 1 >< 3 69 60
3 1 >< 4 72 66
4 1 >< 5 64 61
5 2 >< 3 66 62
6 2 >< 4 65 64
7 2 >< 5 65 67
8 3 >< 4 69 54
9 3 >< 5 65 73
10 4 >< 5 66 67
The average change of bit ciphertext for the key “fti uksw” is 66.4 bits or 51.87%. Meanwhile, for the key
“ftj uksw” an average of 64.6 bits (50.47%) is obtained. The minimum score for the overall data is 54(42.1875%),
and the maximum score is 73(57.03%), and overall, the mean ciphertext difference is 50%. These results indicate
that one-to-many scheme can produce different ciphertext and balanced ciphertext.
3.3. Differences test plaintext-ciphertext
This test is done to see whether the change in 1 bit can make a difference in the overall ciphertext pro-
duced. The greater the distance between plaintext and ciphertext indicates the algorithm that is designed is also
getting better. Conversely, the smaller the difference between the plaintext and the ciphertext produced the less
well the algorithm designed. The absolute difference between plaintext and ciphertext is shown in Table 4 using
key “fti uksw” and “ftj uksw”.
Table 4. Differences ciphertext-plaintext by key
Char-i
“fti uksw” “ftj uksw”
Cp 1 Cp 2 Cp 3 Cp 4 Cp 5 Cp 1 Cp 2 Cp 3 Cp 4 Cp 5
1 99 26 92 18 109 100 89 99 76 63
2 103 28 71 48 78 104 29 136 50 14
3 103 84 82 7 104 103 105 117 90 68
4 33 26 104 6 32 31 90 73 99 114
5 53 63 6 154 45 52 112 24 25 106
6 73 108 185 148 148 78 111 200 88 172
7 54 83 136 62 88 57 7 118 33 79
8 113 62 132 137 105 125 68 111 75 21
9 145 15 2 113 14 73 87 139 74 25
10 63 44 35 26 53 39 151 59 90 4
11 107 132 52 141 53 57 6 72 66 84
12 5 55 97 62 146 40 74 91 84 44
13 25 2 184 143 113 167 30 139 122 216
14 0 147 212 39 56 103 169 41 105 25
15 3 148 54 123 1 192 172 23 15 214
16 14 76 153 5 168 221 1 18 115 188
It can be seen from both tables that the plaintext-ciphertext distance produced by the algorithm design is
quite good. The difference of 1 bit in the key, does not make the difference in the distance in each ciphertext from
the two tables also differ by 1 bit. The results shown are quite varied and spread well. However, further testing
needs to be done to ensure that the spread of distance that occurs is already good. This becomes a note in the next
study to be able to test the plaintext-ciphertext difference with more key variations.
3.4. Avalanche effect test
Comparison of one-to-many scheme with several other cryptographs is done by observing the avalanche
effect value in carrying out the encryption process. In [29] has tested several classical and modern cryptographic

768 ❒ ISSN: 1693-6930
algorithms while this research continues its study with testing on AES and also one one-to-many scheme. Table 5
shows the result of the avalanche effect test for five encryption processes using the same key and plaintext. One-
to-many scheme can produce different avalanche effect value for each encryption, different from other algorithms
that always produce the same value.
Table 5. Result avalanche effect test
No Algorithm
The avalanche effect (%)
Test-1 Test-2 Test-3 Test-4 Test-5
1 Caesar cipher 01.56 01.56 01.56 01.56 01.56
2 Playfair cipher 06.25 06.25 06.25 06.25 06.25
3 Vigenere cipher 03.13 03.13 03.13 03.13 03.13
4 One-to-many 54.29 47.79 49.26 51.47 52.94
5 S-box fleksibel 49.75 49.75 49.75 49.75 49.75
6 DES 55.68 55.68 55.68 55.68 55.68
7 AES 45.58 45.58 45.58 45.58 45.58
The data encryption standard (DES) algorithm has the highest avalanche effect value and the lowest
Caesar cipher value. These results are reasonable because both DES and AES use all block cipher principles
in compiling their algorithms, while Caesar cipher only uses transposition for each text input. The one-to-many
scheme also does not use block cipher principle because the randomness of the generated keys use the logistical
function is a determining factor in the encryption-decryption processes. On average, one-to-many scheme can
produce an avalanche effect value of 52.20%. Although, it is still lower than DES, but it is still better than modern
Blowfish cryptography by 25.46% and 6% more than AES, both of which are still widely used as information
security algorithms.
3.5. Time requirement of cryptanalysis
Suppose an algorithm uses a key block length of 128 bits, the time needed to perform cryptanalysis is
2128
units. If a computer can crack 106
keys in one second, the time needed to crack the key is 5.4 × 1024
years.
As the first step, the one-to-many scheme also uses 128 bits which its algorithm can produce several ciphertexts.
In general, suppose that one-to-many scheme can produce different n-ciphertexts, the ciphertext search process
requires n × 5.4 × 1024
years. This result will certainly complicate the cryptanalysis. Although the possibility
of performing the cryptanalysis still remains but it will take n-times process longer time than cracking ordinary
cryptography.
4. CONCLUSION
The one-to-many encryption scheme produces five ciphertexts that have different and balanced visualiza-
tion. The plaintext-ciphertext correlation test is also close to zero. This condition illustrates that the algorithm can
make plaintext that is not statistically related to ciphertext. A plaintext input will produce several ciphertexts, and
each of them has a variety of more than 50%. An average of 52.20% is obtained from the avalanche effect test,
which is better than modern Blowfish cryptography by 25.46% and AES by 6%, both of which are still widely used
as information security algorithms. This condition will be preferable because it makes difficult for cryptanalysts to
find the relationship between plaintext and ciphertext directly by a process of trial and error. Many different forms
of ciphertext produced by one-to-many scheme will make it very difficult for performing a cryptanalysis. Although
the cryptanalysis process can still be done but it will require a longer process and time compared to algorithms that
are solved ordinarily. If there are n ciphertexts, it will take n-times for time and process to find the plaintext.
ACKNOWLEDGEMENT
The authors are grateful to Satya Wacana Christian University Research and Community Service Center
(BP3M) for the research funding support through the Internal Fundamental Research scheme in the 2017/2018
fiscal year.
REFERENCES
[1] A. Biryukov, J. Daemen, S. Lucks, and S. Vaudenay, “Topics and Research Directions for Symmetric Cryptography,” Early Symmetric
Crypto Workshop, 2017. [Online]. Available: https://orbilu.uni.lu/handle/10993/30953

[2] M. Yang, B. Xiao, and Q. Meng, “New AES Dual Ciphers Based on Rotation of Columns,” Wuhan University Journal of Natural
Sciences, vol. 24, pp. 93-97, 2019, doi: 10.1007/s11859-019-1373-y.
[3] A. Arab, M. J. Rostami, and B. Ghavami, “An image encryption method based on chaos system and AES Algorithm,” The Journal of
Supercomputing, vol. 75, pp. 6663-6682, 2019, doi: 10.1007/s11227-019-02878-7.
[4] A. A. Thinn and M. M. S. Thwin, “Modification of AES Algorithm by Using Second Key and Modified SubBytes Operation for Text
Encryption,” Computational Science and Technology part of Lecture Notes in Electrical Engineering, 2018, vol. 481, pp. 435-444, doi:
10.1007/978-981-13-2622-6 42.
[5] C. R. Dongarsane, D. Maheshkumar, and S. V. Sankpal , “Performance Analysis of AES Implementation on a Wireless Sensor Network,”
Techno-Societal , 2019, pp. 87-93, doi: 10.1007/978-3-030-16848-3 9.
[6] C. Ashokkumar, B. Roy, M. B. S. Venkatesh, and B. L. Menezes, “S-Box Implementation of AES Is Not Side Channel Resistant,”
Journal of Hardware and Systems Security, 2019. [Online]. Available: https://eprint.iacr.org/2018/1002.pdf
[7] T. Manojkumar, P. Karthigaikumar, and V. Ramachandran, “An Optimized S-Box Circuit for High Speed AES Design with Enhanced
PPRM Architecture to Secure Mammographic Images,” Journal of Medical Systems, vol. 43, no. 31, 2019, doi: 10.1007/s10916-018-
1145-9.
[8] S. D. Putra, M. Yudhiprawira, S. Sutikno, Y. Kurniawan, and A. S. Ahmad, “Power Analysis Attack Against Encryption Devices:
A Comprehensive Analysis of AES, DES, and BC3,” TELKOMNIKA, vol. 17, no. 3, pp. 2182-1289, 2019, doi: 10.12928/telkom-
nika.v17i3.9384.
[9] C. A. Sari, G. A. D. R. I. M. Setiadi, and E. H. Rachmawanto, “An improved security and message capacity using AES and Huffman
Coding on Image Steganography,” TELKOMNIKA, vol. 17, no. 5, pp. 2400-2409, 2019, doi: 10.12928/telkomnika.v17i5.9570.
[10] G. C. Prasetyadi, R. Refianti, and A. B. Mutiara, “File Encryption and Hiding Application Based on AES and Append Insertion Steganog-
raphy,” TELKOMNIKA, vol. 16, no. 1, pp. 361-367, 2018, doi: 10.12928/telkomnika.v16i1.6409.
[11] R. Srividyaand B. Ramesh, “Implementation of AES using Biometric,” International Journal of Electrical and Computer Engineering
(IJECE), vol. 9, no. 5, pp. 4266-4276, 2019, doi: 10.11591/ijece.v9i5.pp4266-4276.
[12] J. M. Espalmado and E. Arboleda, “DARE Algorithm: A New Security Protocol by Integration of Different Cryptographic
Techniques,” International Journal of Electrical and Computer Engineering (IJECE), vol. 7, no. 2, pp. 1032–1041, 2017, doi:
10.11591/ijece.v7i2.pp1032-1041.
[13] E. R. Arboleda, J. L. Balaba, and J. C. L. Espineli, “Chaotic Rivest-Shamir-Adlerman Algorithm with Data Encryption Standard Schedul-
ing,” Bulletin of Electrical Engineering and Informatics, vol. 6, no. 3, pp. 219–227, 2017, doi: 10.11591/eei.v6i3.627.
[14] S. D. Putra, A. S. Ahmad, S. Sutikno, Y. Kurniawan, and A. D. W. Sumari, “Revealing AES Encryption Device Key on 328P Micro-
controllers with Differential Power Analysis,” International Journal of Electrical and Computer Engineering (IJECE), vol. 8, no. 6, pp.
5144-5152, 2018, doi: 10.11591/ijece.v8i6.pp5144-5152.
[15] X. Bonnetain, M. Naya-Plasencia, and A. Schrottenloher, “Quantum Security Analysis of AES,” IACR Transactions on Symmetric
Cryptology, vol. 2019, pp 55-93, 2019, doi: 10.13154/tosc.v2019.i2.55-93.
[16] D. G´erault, P. Lafourcade, M. Minier, and C. Solnon, “Revisiting AES Related-Key Differential Attacks with Constraint Programming,”
Cryptology, vol. 139, 2017. [Online]. Available: https://eprint.iacr.org/2017/139.pdf
[17] A. Biryukov and D. Khovratovich , “Related-Key Cryptanalysis of the Full AES-192 and AES-256,” Advances in Cryptology - ASI-
ACRYPT, 2009, vol. 5912 , pp 1-18, doi: 10.1007/978-3-642-10366-7 1.
[18] J. Kim, S. Hong, and B. Preneel, “Related-key rectangle attacks on reduced AES-192 and AES-256,” International Conference on the
Theory and Application of Cryptology and Information Security, 2007, vol. 4593, pp 225–241, doo: 10.1007/978-3-540-74619-5 15.
[19] S. Lucks, “Ciphers secure against related-key attacks,” International Workshop on Fast Software Encryption, 2004, vol. 3017, pp.
359–370, doi: 10.1007/978-3-540-25937-4 23.
[20] D. Deutsch and R. Jozsa, “Rapid Solution of Problems by Quantum Computation,” Proceedings The Royal Society London A, Mahtem-
atical Physical & Engineering Sciences, 1992, vol. 439, pp. 553–558, doi: 10.1098/rspa.1992.0167.
[21] R, P. Feynman, “Simulating Physics with Computers,” International Journal of Theoretical Physics, vol. 21, pp. 467–488, 1982, doi:
10.1007/BF02650179.
[22] R. A. Perlner and D. A. Cooper, “Quantum resistant public key cryptography: a survey,” IDtrust ’09: Proceedings of the 8th Symposium
on Identity and Trust on the Internet, 2009, doi: 10.1145/1527017.1527028.
[23] M. S. Henriques and N. K. Vernekar, “Using Symmetric and Asymmetric Cryptography to Secure Communication Be-
tween Devices in IOT,” International Conference on IoT and Applications (ICIOT), Nagapattinam, India, 2017, pp. 1-4, doi:
10.1109/ICIOTA.2017.8073643.
[24] A. Hafsa, A. Sghaier, M. Zeghid, J. Malek, and M. Machhout, “An improved co-designed AES-ECC cryptosystem for se-
cure data transmission,” International Journal of Information and Computer Security, vol. 13, no. 1, pp. 118-140, 2020, doi:
10.1504/IJICS.2020.10028859.
[25] R. Riyaldhi, Rojali, and A. Kurniawan, “Improvement of Advanced Encryption Standard Algorithm with Shift Row and S.Box Modifica-
tion Mapping in Mix Coloumn,” Discovery and innovation of computer science technology in artificial intelligence era: The 2nd Interna-
tional Conference on Computer Science and Computational Intelligence, 2017, vol. 116, pp. 401-407, doi: 10.1016/j.procs.2017.10.079.
[26] G. Ribordy, J.-D. Gautier, N. Gisin, O. Guinnard, and H. Zbinden, “Fast and User-Friendly Quantum Key Distribution,” Journal of
Modern Optics, vol. 47, pp. 517-531, 2000, doi: 10.1080/09500340008244057.
[27] M. Aledhari, A. Marhoon, A. Hamad, and F. Saeed, “A New Cryptography Algorithm to Protect Cloud-based Healthcare Services,” Pro-
ceedings of the Second IEEE/ACM Interntional Confrence on Connected Health: Applications, Systems and Engeneering Technologies,
Philadelphia, PA, USA, 2017, pp. 37-43, doi: 10.1109/CHASE.2017.57.
[28] A. D. Wowor and V. B. Liwandouw, “Domain Examination of Chaos Logistics Function As A Key Generator in Cryptography,” Inter-
national Journal of Electrical and Computer Engineering (IJECE), vol. 8, no. 6, pp. 4577–4583, 2018, doi: 10.11591/ijece.v8i6.pp4577-
4583.
[29] S. Ramanujan and M. Karuppiah, “Designing an Algorithm with High Avalanche Effect”, International Journal of Computer Science
and Network Security, vol. 11, no. 1, pp. 106-111, 2011. [Online]. Available: http://paper.ijcsns.org/07 book/201101/20110116.pdf

770 ❒ ISSN: 1693-6930
BIOGRAPHIES OF AUTHORS
Alz Danny Wowor is currently a lecturer at the Faculty of Information Technology,
Satya Wacana Christian University in Salatiga, Indonesia. He received bachelor and master degree in
mathematics and informatics from Satya Wacana Christian University, in 2005 and 2011 respectively.
His researches are in fields of Primitive Cryptography, Symmetric Cryptography: Block Cipher and
Pseudorandom. He can be contacted at email: alzdanny.wowor@uksw.edu.
Bambang Susanto He is currently a lecturer at the Department of Mathematics, Faculty
of science and mathematics, Satya Wacana Christian University in Salatiga, Indonesia. His obtained
Bachelor Degree in Mathematics from Diponegoro University (Indonesia) in 1988. His received
master and doctor degree in mathematics from Bandung Institute of Technology, in 1992 and 2005
respectively. His researches are in fields of Data Mining, Big Data Analysis, Business Intelligence,
Risk Analysis, Primitive Cryptography, Symmetric Cryptography. He can be contacted at email:
bambang.susanto@uksw.edu.

One to many (new scheme for symmetric cryptography)

More Related Content

Similar to One to many (new scheme for symmetric cryptography) (20)

More from TELKOMNIKA JOURNAL (20)

Recently uploaded (20)

One to many (new scheme for symmetric cryptography)