International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 24 (2017) pp.

15363-15377
© Research India Publications. http://www.ripublication.com

Security Issues in DNA Based on Data Hiding: A Review

1
Dilovan Asaad Zebari, 2 Habibollah Haron and 3Subhi R. M. Zeebaree
Faculty of Computing, Universiti Teknologi Malaysia (UTM), 81310 Johor Bahru, Johor, Malaysia.
1,2
3
Department of Computer and Communications Engineering, Nawroz University, Tenahi, Duhok, Kurdistan Region, Iraq.

1
Orcid: 0000-0002-7643-6359

Abstract DATA HIDING REQUIREMENTS

Information security and confidentiality are a key concern, In literature, several data hiding schemes are discussed in this
particularly with the rampant growth and use of the internet. paper, along with their parameters. Different schemes use
Along with the growth comes the incidents of unauthorized various data embedding approaches, each of which have
information access which are countered by the use of varied common requirements which are used in measuring
secure communication techniques, namely; cryptography and performance and identifies their advantages and disadvantages.
data hiding. More recent trends are concerned with the One of the three common requirements is vulnerability to
application of DNA cryptography and data hiding by using it attacks by intruders. It is imperative that the embedded data is
as a carrier thereby making use of its bio-molecular kept undetectable perceptually as well as statistically so as not
computational properties. This paper provides a survey of to arouse any suspicion. In addition, care should be taken to
recently published DNA based data hiding algorithms which ensure that the carrier files’ attributes and properties are not
make use of DNA to safeguard critical data being transmitted tampered with in the course of embedding the message thereby
over an insecure communication channel. Several DNA-based providing a perfectly secure system where statistics of both
data hiding techniques will be discussed with particular carrier and stego files are identical [5, 6]. The second common
emphasis on strength and weaknesses of the algorithm in requirement is capacity, defined as the quantity of data hidden
question; algorithms are compared based on the cracking within the carrier. The design of a data hiding technique should
probability, double layer of security, single or double hiding be done in a manner that allows for more secret data to be
layer, blindness, and much more. This will be useful for future hidden within the carrier while retaining the properties of the
research in the design of more efficient and reliable of secure stego file [1, 5]. A good data hiding technique should have
DNA data hiding techniques. sufficient embedding capacity for holding the information [6].
The third common requirement is imperceptibility, defined as
Keywords: Data security, Cryptography, Data hiding, DNA.
the possession of a high embedding capacity as well as the
ability to resist intruders. The stego carrier should preferably be
free from visual artefacts; the higher the fidelity of the stego
INTRODUCTION
carrier, the better [2].
In general, the notion of security is the protection of
information from being accessed by unauthorized people. The
main purpose of security in the modern perspective of computer FUNDAMENTAL CONCEPTS OF DATA HIDING
science is to keep sensitive data from being changed, destroyed,
The concept of data hiding is usually modelled by a pair of
and stolen or being damaged by a third party [1]. The most
algorithms which are embedding and extracting as shown in
commonly used techniques in communication and computer
Figure 1. An embedding algorithm refers to the combination of
security fields are cryptography and data hiding. Usually,
two files together, the secret data and carrier with an optional
communications are kept secure through the use of
key in order to obtain a stego file which holds the secret data.
cryptography, and data hiding (which are related concepts) [2,
However, the extraction algorithm serves the purpose of
3]. Even though both methods have a similar goal, their
extracting the secret data from the stego file [7]. Pure data
implementation and usage differ greatly; cryptography changes
hiding is considered to be a category of data hiding type which
the meaning of secret writing while data hiding is a hidden form
does not utilize any key; its security is based on the privacy of
of writing which hides the existence of the embedded message.
the algorithm. Thus, it is considered to be a less secure method
Thus, data hiding is more secure, sufficient, and often preferred
[8, 9]. Secret data hiding is another data hiding type which uses
to cryptography in the transmission of data across an insecure,
a single key for both processes (embedding and extraction).
public channel [4, 5].
One of the greatest advantages of this type is the provision of a
fast process in during both procedures [10, 11]. Unlike previous
types, public data hiding uses a pair of keys in both processes:

15363
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 24 (2017) pp. 15363-15377
© Research India Publications. http://www.ripublication.com

one for embedding and another one for extracting. The main direct all chemical elements in the human body. For this reason,
advantage of this type is the robustness of the system; if one DNA has been put forward as a viable option for use in
key is known by a third party, it is quite hard to find the other computational applications [18].
key [10, 12]. However, this type is slower than private data
hiding by about 100-1000 times [13].
DNA Structure
DNA is defined as the genetic drawing of every living creature.
Each body cell has a unique complete set of DNAs; a polymer
comprised of monomers referred to as deoxyribose nucleotides
made up of three components as shown in Figure 2 [19].

Figure 1: General Principle of Data Hiding System

Several applications are used to represent a container for

sensitive information. These applications are used as cover
objects or carriers in the data hiding systems. Every carrier has
its own characteristics which serve the data hiding technology.
The amount of secret information required in order to conceal
data within each carrier depends on the availability of the Figure 2: The Structure of Deoxyribose Nucleic DNA [19]
region of the specific carrier. Therefore, carriers are an essential
ingredient of any data hiding system. Multimedia that serves
as data hiding carriers include text, audio, video, and images. The human body consists of trillions of cells each of which
Text can be hidden by modifying the layout of the text, using serves different functions. Each cell contains a nucleus that has
an nth character from text or by the alteration of some of the a number of chromosomes as illustrated in Figure 3. Most of
rules such as spaces, etc. Text can also be hidden by using a the DNA contents are found in a nucleus called nuclear DNA,
code made up of characters, lines and page numbers. This and the rest of it is contents are found in mitochondria which
approach, however, is not secure [2]. The main advantage of are called mitochondria DNA (mtDNA). The function of each
this carrier is that it does not require large volumes of memory cell is controlled by DNA. Each DNA's chromosome consists
and it is also easy to transfer. While it has very small amount of a DNA molecule which holds genes. The gene is the entire
redundant data compared with other carriers [10, 14]. Data can genetic makeup, which essentially contains information from
be hidden in audio files through the use of inaudible all the chromosomes [20].
frequencies and by slightly changing the binary sequence of an
audio file [2, 15]. Hiding data in video files is much more
effective and successful due to the large capacity available thus
making it possible to hide the data within different frames of a
video [16]. Generally, data hiding in video is classified into two
main types which are uncompressed and compressed video.
Digital images have also become popular carriers for hiding
secret information since they possess a high degree of
redundancy, high capacity in images, low impact on the
visibility and simplicity in their manipulation [15, 17]. DNA is
a recent carrier have been used in data hiding area. In this
paper, we focus only on data hiding in DNA. Figure 3: Cell and it is Genetic Material [20]

DNA The genetic material known as DNA structure was found out
In biology, a Deoxyribonucleic Acid (DNA) is the leading by Watson and Crick in (1953). DNA is a long molecule located
molecular structure for encoding the data required to create and in all living organism's body cells. The DNA is a germ form of

15364
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 24 (2017) pp. 15363-15377
© Research India Publications. http://www.ripublication.com

plasm that contains all lifestyle; it is formed by two backbone DNA Computing
strands twisted around each other, called a Double-helix as
For the being time, biology techniques diffuse in many
shown in Figure 4. Each DNA strand is composed of many tiny
different areas. DNA is the most recent biological technique
subunits called nucleotides. Adenine (A), thymine (T), guanine
being used in various applications [24]. This is because DNA
(G), and cytosine (C) are the four chemical bases held in the
computing can potentially be used to solve a wide range of
DNA sequence which are stuck onto sugar and phosphate on
computational problems whose computation time grows
the backbone to complete the nucleotide. Biologically, there are
exponentially; `NP-complete' or non-deterministic polynomial
two pairs of DNA bases: Purine (A and G) and Pyrimidine (T
time complete problems. A significant amount of research in
and C). Constantly (A) linked to (T) within two hydrogen bonds
this area was done, with great strides being made on the subject
and (C) linked to (G) within three hydrogen bonds [19, 21].
of DNA and the immune system [19]. In 1994 the first
experiment of DNA computing (bio-molecular computing) had
done by Leonard Adelman, in this method tools of molecular
biology had been used for solving a part of standard path of
Hamiltonian problem. At that time, computing with molecule
directly was discovered a field that considered as new in term
of science security [25]. The research study by Lipton in 1995
used DNA computing to solve the satisfaction problem (SAT),
which is a NP-complete problem. The solution provided made
Figure 4: Helical Structure of DNA [20] use of the parallelism property of DNA as well as it is
computing and storage capabilities [19]. Later, in 1997 Ogihara
and Ray discovered that DNA can simulate Boolean AND and
Every three adjacent nucleotides make up a codon. Given that OR gates [26]. The first experiment of data hiding technique in
each nucleotide could have any of the four chemical bases and DNA has been proposed successfully, using DNA microdots,
each codon is comprised of three nucleotides, then there is a for concealing secret data by Clelland [27].
sum of 43 = 64 different possible combinations. These
combinations determine the amino acids to be used by living
organisms, whose arrangement determines the structure and DNA Binary Coding
function of the resultant protein [10]. Transcription is the Each DNA strand has four chemical bases: A, C, G, and T.
process through which RNA, an intermediary copy of the Biologically, A is associated to T and C is associated to G. In
instructions contained in DNA, is created. The RNA is also
the binary computing area, the synthesis of DNA bases can be
made up of four bases: adenine (A), cytosine (C), uracil (U) and
changed by the input decisions, by assuming that T is
guanine (G). Figure 5, shows all of the 64 codons. Three of
associated with C or T is associated with G, and so on [28].
which, the ones labelled STOP, do not actually code for any
In order to store data in DNA molecules, researchers have to
amino acid but instead, they signal the end of the protein chain.
encode a secret message into DNA bases using a binary coding
The remaining 61 codons are responsible for specifying the 20
rule to combine with the DNA sequence. Researchers have the
amino acids. However, some of the amino acids are coded for
option of selecting any equivalent binary form for each base
by more than one codon through the feature referred to as
(A): the binary forms can be '00', '01', '10', or '11', and so on.
degeneracy [11]. It follows that this codon redundancy can be
This coding along with the randomness properties makes DNA
used to alter the genetic sequence while retaining its
a befitting application for both computing and cryptography.
functionality [11, 22, 23].
Therefore, the coding of DNA to binary form can give 4! = 24
different encoding ways [29]. Which in turn makes it possible
to carry out logical operations such as Addition, Subtraction,
XOR, AND, OR, and NOT over the DNA bases.
F(X): X → Y, where X = {A, C, G, T} and Y = {00, 01, 10,
11}. This can be express as in Table 1.

Table 1: DNA Binary Coding

DNA Base Binary code
A 00
C 01
G 10
T 11
Figure 5: The Amino Acid and Codons Table [11]

15365
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 24 (2017) pp. 15363-15377
© Research India Publications. http://www.ripublication.com

DNA Data Hiding Techniques a significantly alters the length of the DNA sequence which
rouses the suspicion of a hacker to the existence of an
Several carriers have been used in data hiding algorithms where
embedded message [30]. Legally there are six main
each carrier has it is own characteristics. Depending on the
complementary rules as it is shown in Figure 6.
specific carrier, different techniques have been used to hide
secret information. The process of hiding and the mechanism
of each technique is different from another. According to the
techniques, different changes will happen during the hiding
process. In DNA data hiding, three techniques were proposed
in 2010 by [23], all of which were considered to be the main
techniques of hiding data in DNA sequences. The three main
techniques can be defined as follow:

Insertion Technique
This technique depends on the merging among the reference of
DNA sequence (S) which use as a carrier and secret message.
Figure 6: The Six Complementary Rules
During the process of this technique, both of them are translated
into a binary system according to any binary coding rule. After
that, the DNA reference is separated into equal sized segments
Substitution Technique
in order to insert each bit of the secret message after each
segment of DNA reference and then converting back into a Regarding this technique there is no merge between reference
DNA sequence resulting in a stego DNA. Furthermore, less DNA sequence and the secret information. In this scheme,
modification rate is considered as a great feature of this specific positions in the DNA reference are selected randomly
technique because it depends on inserting secret data in the as determined by the algorithm. After that, at least one
DNA reference not replacing the contents of DNA reference complementary rule should be selected to replace each letter of
[30]. However, the main disadvantage of this technique is the the message with the DNA contents in particular locations.
increase in redundancy during the process, and the stego DNA Depending on the contents of the message, the process will be
length will be higher than that of the DNA reference. This carried out to obtain the stego DNA. Hence, the DNA length is
implies that the use of this technique will attract the attention maintained following the embedding of the message only the
of unauthorized users [30, 31, 32, 33, 34]. replacement has done between secret message and DNA
reference. This, in turn, means that in an effort to conceal the
secret data, the resulting stego DNA is highly modified [30].
Complementary Pair Rule Techniques As a result, this technique is considered as a more efficient
technique than the previous techniques because it provides
In this technique, the procedure begins with the selection of a
more complexity and better performance [36].
DNA sequence in which the longest existing complementary
pair is contained. This is followed by the random generation of
two complementary string pairs whose length is one more than
COMPARATIVE STUDY
that existing in the sequence, after which these pairs are padded
with a ‘T’ at the posterior and anterior. Afterwards, they are In this subsection some of security issues will be compared
inserted one at a time into S while ensuring that there is no among forty-one recent algorithms such as cracking
overlapping. The message is then divided into segments, each probability, double layer of security, single or double hiding
containing an even number of bits after which the data is coded layer, blindness, and more in Table II relative to strong and
back into nucleotides using the binary coding rule. For each weak points of each one. The objective of the comparison
pair of a complementary substring in the converted sequence, a provided in this study is to ensure that researchers are armed
message bit is inserted before TajT, where aj represents the pair with the knowledge of the disadvantages presented by existing
of longest complementary substrings. A resultant sequence data hiding schemes thereby providing motivation for future
containing message S' is then obtained. This scheme results in advances in this field.

15366
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 24 (2017) pp. 15363-15377
© Research India Publications. http://www.ripublication.com

Table 2: Comparison of Strong and Weak Points between Various DNA Based on Data Hiding Techniques
No Author Year Strong Points Weak Points
Insertion technique . Length of stego DNA is longer than DNA
. High capacity. reference.
. Easy to implement. . Payload not equal zero.
. Low modification rate. . Need multiple data in extraction process.
. Does not preserve the functionality of the amino
acid.
. Un-blind algorithm.
. Increase the redundancy.
. Pure data hiding method.
Complementary Pair rules technique . Payload not equal zero.
. Easy to implement. . High modification rate.
1 Ref [23] 2010 . Attackers need to know multiple data to crack . Does not preserve the functionality of the amino
the secret data. acid.
. Un-blind algorithm.
. Changing the length of DNA after embedding
process.
. Pure data hiding method.
Substitution technique . High modification rate.
. High capacity. . Does not preserve the functionality of the amino
. Payload is equal to zero. acid.
. Easy to implement. . Un-blind algorithm.
. More efficient, more complexity and better . Pure data hiding method.
performance than previous methods.
. Carrying out the result of hiding data in cloud . Un-blind algorithm.
to increase the level of confidentiality and . Increase the message size.
complexity. . Security depends on the DNA reference.
2 Ref [21] 2011 . Simple algorithm and high capacity. . Pure data hiding algorithm.
. Payload equal to zero.
. Preserve the functionality of the biological
DNA.
. Construct a reversible data hiding method. . Does not encrypt secret data before hiding.
. Preserve the functionality of DNA.
3 Ref [24] 2011
. Blind algorithm.
. Secret key is used.
. Mapped between DNA codons and amino to . Expand the length of stego DNA.
provide security. . Not preserve the functionality of the biological
. Encrypt secret message by playfair cipher DNA.
before hiding. . In order to extract the secret message from stego
. Improve playfair cipher by modifying to 5*5 to DNA, it needs to send multiple data to receiver.
avoid its drawbacks where after encryption the . Payload not equal to zero.
4 Ref [18] 2012 diagraphs and the structure of secret text still
exists.
. Providing double layer of security.
. Blind Algorithm.
. Hight capacity and better time performance.
. Provide high probability cracking.
. Secret key is used.
. Enhanced the effectiveness original . Not preserve the functionality of the biological
substitution method. DNA.
. Transmission efficiency can be increased for . To extract the secret message from stego DNA,
5 Ref [36] 2012
data hiding system in internet. it needs to send multiple data including DNA
. Providing more performance in term of reference, stego DNA, secret message location
capacity and security. set, and the table rule.

15367
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 24 (2017) pp. 15363-15377
© Research India Publications. http://www.ripublication.com

. TLSM is extended which can hide secret data . Un-Blind algorithm.

in any sequences of letters or symbols. . High modification rate.
. To improve the performance of the TLSM two . Pure data hiding algorithm.
approaches are presented which are the Base-t
TLSM and the Extended TLSM (ETLSM).
. Capacity increased.
. Propose a cryptographic-data hiding protocol . Un- blind algorithm.
reduce the using of public key as well as for . Does not preserve the functionality of the
best security. biological DNA.
. Payload is equal to zero.
6 Ref [13] 2012
. High capacity.
. Utilizing the innovative of DNA data hiding to
hide the secret key within DNA reference for
more security.
. Does not expand the length of stego DNA . Does not preserve the functionality of the
means payload equal to zero. biological DNA.
. Simple algorithm. . High modification rate if DNA reference contains
. Improved the hiding capacity. a lot of repeated nucleotides.
7 Ref [31] 2012 . Minimize the modification rate. . Un- blind algorithm also injective mapping and
. Used substitution method in hiding. complementary rule should be known by both
sender and receiver.
. Pure data hiding algorithm.
. Security depends on the DNA reference.
. Flexible algorithm. . Does not preserve the functionality of the
. Easy to implement. biological DNA.
. Encrypt secret message by modified playfair . Un- blind algorithm.
algorithm which using DNA and amino acid. . Security depends on the DNA reference.
8 Ref [37] 2012
. Does not expand the length of DNA after hiding
process.
. Secret key is used.
. Used substitution method in hiding.
. Using three keys in the algorithm. . Does not apply any encryption technique.
. Improved first and third techniques of Ref [23] . Hiding secret data only in nucleotides which their
9 Ref [38] 2012 in term of modification rate. marks are equal to zeros after converting to
. Does not expand the stego DNA. binary.
. Blind algorithm.
. Easy to implement. . Does not preserve the functionality of the
. Low modification rate. biological DNA.
. Does not expand the length of stego DNA. . Preserve the position of each DNA base which
. One of the powerful encryption technique hold the secret data then send to the receiver for
10 Ref [39] 2013
(RSA) is used to encrypt secret data before extraction.
hiding. . The size of secret data is increased.
. Public key is used. . Un- blind algorithm.
. Low probability cracking.
. Simple algorithm. . Does not preserve the functionality of the
. High capacity. biological DNA.
11 Ref [40] 2013 . Un- blind algorithm.
. Pure data hiding algorithm.
. Low probability cracking.
. Easy to implement. . Low capacity due of using LSB in hiding process.
. Preserve the functionality of the biological . No encryption on secret data before hiding.
DNA. . Low probability cracking.
12 Ref [11] 2013
. Low modification rate.
. Blind algorithm.
. Secret key is hiding in DNA reference which

15368
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 24 (2017) pp. 15363-15377
© Research India Publications. http://www.ripublication.com

provide more security.

. Does not expand the DNA reference after
hiding confidential data.
. Encrypt secret data by exhibiting DNA amino . Un- blind algorithm.
acids. . High modification rate.
. Encrypt the secret key by RSA algorithm before . Payload not equal to zero.
13 Ref [22] 2013 hiding within DNA reference. . Does not preserve the functionality of amino
. High capacity. acids.
. public key is used.
. High probability cracking.
. Preserve the translation of protein in protein . Complex calculation.
coding DNA (PcDNA). . Intruders can estimate the key for unconstrained
. Encoding the data in robust and near to optimal ncDNA hiding.
way.
14 Ref [41] 2013 . Preserve the codon statistics.
. Embedding data achieved near to optimum.
. Embedded data in DNA in an efficient and
robust way.
. Secret key is used.
. Encrypt secret data before hiding by playfair . Multiple data need to extract the secret data.
algorithm. . Un- blind algorithm.
. High security.
. Since hiding at last step in an audio will not
attract attackers.
15 Ref [42] 2013
. Hide secret data and interpret into audio file
so it is not easy to prove both data are existing
inside audio.
. Provide double hiding layers.
. Secret key is used.
. The key space is big enough to resist brute force . Complex calculation.
passive intruders.
. Encrypt the secret data before hiding in host
document.
. Blind algorithm.
16 Ref [43] 2013
. The ratio of embedding capacity is 100%.
. Provide double hiding layers.
. DNA reference is construct by Chebyshev
maps.
. Substitution method is used in hiding.
. Modified the secret key of Ref [41] algorithm . Complex calculation.
to incorporate the secret key. . Pure data hiding algorithm.
17 Ref [44] 2013
. As well as retaining all strong points of Ref
[41].
. Increased the capacity and security of the . Does not preserve the functionality of the
original substitution technique. biological DNA.
. Blind algorithm. . Increased the size of secret binary in case of
18 Ref [35] 2014 . Improved substitution method. multiplying by 6 if not equal zero will add extra
zeros.
. Expanding the length of stego DNA.
. Pure data hiding algorithm.
. High capacity. . Un- blind algorithm.
. Simple algorithm. . Random DNA sequence and selected
19 Ref [45] 2014 . Sending secret data as (ABCD) format. complementary pair rule should send to the
receiver.
. No encryption on data before embedding.

15369
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 24 (2017) pp. 15363-15377
© Research India Publications. http://www.ripublication.com

. Low probability cracking.

. Pure data hiding algorithm.
. Blind algorithm. . Increased the redundancy.
. Sending only the integer value of stego DNA to . Increased the message size.
receiver. . Does not preserve the DNA functionality.
. High security. . Expand the length of stego DNA.
. It is hard for intruders to know the generated
random number seeds.
20 Ref [46] 2014 . It is hard for intruders to know how many
packets are divided, also how many message
bits and DNA binary in a packet.
. Randomly merged secret message bits and
DNA reference bits.
. Secret key is used.
. High probability cracking.
. Encrypt secret message before hiding using . The algorithm needs multiple keys during
RC4. extraction process.
. High capacity.
21 Ref [47] 2014 . Providing good security.
. Provide double hiding layers.
. Construct DNA from image.
. Secret key is used.
. Security of secret data is improved in better
extent.
. Unequal size of extra grids can be used to store
additional data.
22 Ref [48] 2014 . Encrypt secret data based upon BASE64 --------------------------------------------
encoding.
. Provide double hiding layers.
. DNA is construct from secret text.
. Secret key is used.
. High security. . The algorithm has two parts of extraction header
. High capacity. and data extractions.
. The possible prime range is between 420- 440
because key of prime length is between 20-40.
. Increased payload amount and decreased the
23 Ref [49] 2014 image distortion.
. RC4 encryption is used to encrypt confidential
data before hiding.
. Provide double hiding layers.
. Construct DNA from cover image.
. Secret key is used.
. Provide double layers of security. . Multiple data needs during extraction.
. Encrypt secret data by AES-128. . Does not preserve the DNA functionality.
. Good amount of security has provided by AES.
. Applied different operations on confidential
24 Ref [50] 2014 data before and after encrypting such as XOR
and HASH-512 operations.
. To enhance the security DNA is embedded
within microdot.
. Secret key is used.
. Low modification rate. . Replacement rules should be send to receiver.
. No expand of DNA reference after embedding . Plain text must be containing only (capital,
25 Ref [33] 2015
confidential data. letters, small letters, 0, …, 9, period, and dot). It
. Uses double DNA references. cannot contain other punctuation marks.

15370
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 24 (2017) pp. 15363-15377
© Research India Publications. http://www.ripublication.com

. Preserved the functionality of the original DNA . Pure data hiding algorithm.
reference.
. Blind algorithm.
. Does not change in the non-labelled
nucleotides.
. High capacity.
. Encrypt plain text before embedding.
. High probability cracking.
. Uses three DNA reference in the proposed . High modification rate.
algorithm. . Does not preserve the functionality of the
. Blind algorithm. biological DNA.
26 Ref [34] 2015
. Encrypt the plain text before hiding.
. Secret key is used.
. High probability cracking.
. Easy to implement using any programming . No encryption.
language. . Does not Preserving the information of
. Create random codon table to convert secret organism’s life.
27 Ref [51] 2015 message to DNA format. . Expand the length of DNA reference after
. High redundancy because of using insertion embedding.
technique. . Un- blind algorithm.
. Pure data hiding algorithm.
. Low modification rate. . If the DNA reference contains many repeated
. Preserved the functionality of the original DNA bases the modification rate will be high.
reference. . Low probability cracking.
. Blind algorithm.
. Using playfire algorithm to encrypt the secret
28 Ref [32] 2015
message.
. No expansion in the DNA reference after hiding
secret data.
. Substitution method is used in hiding.
. Secret key is used.
. Encrypt secret message by modified playfair . Does not preserve the functionality of the
algorithm. biological DNA.
. Does not expand the length of DNA after hiding . Un- blind algorithm.
process. . Probability cracking is very low.
. High capacity.
29 Ref [52] 2015
. Simple, low execution time, and better
performance than Ref [37].
. Enhanced the hiding process of Ref [37].
. Substitution method is used in hiding.
. Secret key is used.
. Imperceptible technique. . Un- blind algorithm.
. Encrypt secret message before hiding. . Hiding DNA message in only one component of
30 Ref [53] 2015
. Provide double hiding layers. cover image.
. Pure data hiding algorithm.
. Hiding two secret images within image without . No encryption technique applied on secret data.
distortion. . Un- blind algorithm.
31 Ref [54] 2015
. Provide double hiding layers.
. Secret key is used.
. Improved better security. . Require multiple data during the whole
. Improved double carrier by reducing noise bits embedding and extraction Processes.
in image.
32 Ref [55] 2015
. Provide a reasonable capacity.
. Using several parameters of two dimensional
2D logistic map.

15371
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 24 (2017) pp. 15363-15377
© Research India Publications. http://www.ripublication.com

. RC4 used to encrypt secret data.

. Provide double hiding layers.
. DNA is constructed from image.
. Secret key is used.
. Substitution method is used in hiding.
. Imperceptible technique. . Un- blind technique.
. An efficient technique. . Multiple data need in extraction process.
33 Ref [56] 2015 . Providing security by hiding in random video . Pure data hiding algorithm.
frame.
. Provide double hiding layers.
. Highly secure and efficient method. . Does not preserve the functionality of the
. Encrypt secret data using RSA algorithm before biological DNA.
embedding.
34 Ref [57] 2015
. Blind algorithm.
. Provide double hiding layers.
. Public key is used.
. Preserving the information of organism’s life. . Not easy to implement.
. Blind algorithm. . High modification rate.
. High capacity.
. Does not expand the length of stego DNA.
. XOR and PRBG is used to encrypt the secret
35 Ref [58] 2016
data.
. Errors derived and corrected by Reed-Solomon
(RS) code.
. Secret key is used.
. High probability cracking.
. Any type and any size of the secret data and the . Un- blind algorithm.
key can be used. . Does not preserve the amino acid functionality.
. Applying different encryption techniques and . High modification rate.
analysing them to select the good one before . Low probability cracking.
hiding.
. Normal key is used to select English letters to
36 Ref [59] 2016 generate the playfair cipher grid which can be
more secure.
. High hiding capacity.
. No redundancy in the process.
. More simplicity.
. Good performance with low time execution.
. Substitution method is used in hiding.
. Using the vigenere notion and a complementary . Un- blind algorithm.
pair rule to construct a table for hiding secret . Does not preserve the amino acid functionality.
37 Ref [60] 2016 data.
. Simple algorithm.
. Secret key is used.
. Blind algorithm. . Does not preserve the amino acid functionality.
. Encrypting secret message before hiding using . Payload not equal zero.
playfair. . High modification rate.
38 Ref [61] 2016 . High embedding capacity. . High redundancy.
. High robustness method.
. Secret key is used.
. High probability cracking.
. High payload capacity.
. It is a fragile reversible technique.
39 Ref [62] 2016 ---------------------------------------------
. Transparency is very high.
. The reliability of secret information is

15372
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 24 (2017) pp. 15363-15377
© Research India Publications. http://www.ripublication.com

provided.
. High quality.
. DNA is construct from secret text.
. Provide double hiding layers.
. Secret key is used.
. Encrypt secret data by vigener or playfair . Un- blind algorithm.
cipher. . Sending multiple data to the receiver for
. Double amount of data hiding. extraction process.
. High security. . High modification rate in non-coding region.
. After hiding DNA reference will be send in a
microdot in a Paper before sending to the
40 Ref [63] 2017
receiver.
. Regenerate another key and DNA sequence
then the process of hiding will happen again if
the paper is contaminated.
. Preserve the functionality of DNA sequence
and avoid any mutations.
. High security. . Un- blind algorithm.
. Double random key generator. . Does not preserve the functionality of DNA.
41 Ref [64] 2017
. Secret key is used. . Payload not equal to zero.
. Probability cracking is very high.

The conclusion behind the derived comparison in Table II is to for, as achieved in [11, 21, 24, 32, 33, 41, 52, 58, 63].
explain the strength and weakness points of the proposed DNA
In most data hiding algorithms, the carrier will face some
based on data hiding algorithms. The strength is fostered by
distortion after hiding process. Techniques of hiding data are
encrypting the confidential data into cipher data before
concerned with embedding the existence of the embedded data
embedding instead of embedding the original data format thus
that’s why it can be unobtrusively communicated. Thus, carrier
providing more security as achieved by [13,18, 22, 32, 33, 34,
distortion is very important to minimize. In the DNA sequence,
35, 37, 39, 42, 47, 48, 50, 52, 55, 57, 58, 59, 60, 61, 63]. The
the length of the DNA sequence and the extent of modification
most promising encryption technique to combine with DNA
are considered as the measures of the quality of a stego DNA
based data hiding technique is playfair technique as it is
once data has been embedded. A low modification rate and no
approved in [59]. In their work a comprehensive comparative
expansion rate result in a better quality stego DNA which draws
of some encryption techniques has been done which are
less attention from potential attackers. A low modification rate
vigenere and playfair, AES, and RSA ciphers. Each of them
is achieved in [11, 31, 32, 33, 38, 39]. Furthermore, no
combined with a data hiding in DNA based on substitution
expansion rate attribute of stego DNA which means payload is
method. The results showed that the playfair cipher is not only
equal to zero is achieved in [11, 13, 21, 31, 32, 33, 37, 38, 39,
fast and simple, but it can also provide more security and large
58].
capacity.
Two phase data hiding approach is proposed to hide secret data
The main feature which supported by DNA based data hiding
in more depth than the original data hiding approaches. Using
technique is the blindness property which does not need to send
two different carriers in the same approach will provide more
the original DNA reference to the receiver. The main objective
security and more difficulties to understand or retrieve the
behind the blindness property is to maximize the security
secret data by attackers. Several approaches used DNA
degree as possible and to prevent any way can be discovered by
reference with another multimedia carrier to hide secret data as
an attacker to be obtained through minimizing the required data
achieved in [42, 43, 47, 48, 49, 54, 55, 56, 57, 62]. Some of
that’s sent to the receiver as much as possible as it is achieved
them constructed the DNA from a cover image or from secret
in [11, 18, 24, 32, 33, 34, 35, 38, 43, 46, 57, 58, 61].
data as in [43, 47, 48, 49, 55], while some others either carried
Biologically conserving the original functionality of the DNA out a random selection or selected from the online database as
reference after embedding process is considered as one of the in [42, 54, 56, 57, 62].
strength points while maintaining a reasonable data payload.
One of the most essential elements in data hiding approaches is
DNA reference could be used to hide secret data in such a way
the key. Data hiding systems can be divided into three types
that it is functionality of producing proteins is not affected.
based on the key used. The first type is called pure data hiding
Exploiting some of DNA properties such as silent mutation and
which is less secure because it does not use any type of key as
codon redundancy can be used to hide data and change the
in [21, 23, 31, 33, 35, 36, 40, 45, 51, 53, 56]. Thus, to provide
genetic sequence without changing the protein chain it codes
more security for the system it is important to use a key which

15373
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 24 (2017) pp. 15363-15377
© Research India Publications. http://www.ripublication.com

makes an attack on the data hiding system much more difficult CONCLUSION
because even if the attackers identify the data hiding scheme
A surge in the demand for storage has created a tremendous
used, they still cannot extract the secret information from the
demand for the development of new and emerging techniques
carrier because they do not have the key. Only the sender and
for powerful data storage. Presently, DNA has been identified
the receiver have the key. So, in data hiding, it is preferable to
as an effective data carrier which also has the added advantage
use a strong key thereby ensuring a more secure system. The
of reliable data storage. The bio-molecular computational
second type is called secret key as achieved in [11, 13, 18, 24,
abilities of DNA are being applied in cryptography and data
32, 34, 37, 38, 41, 42, 43, 44, 46, 47, 48, 49, 50, 52, 54, 55, 58,
hiding. This paper lays out a comparison of several recent DNA
59, 60, 61, 62, 63, 64]. The third type is called public key as
based data hiding algorithms highlighting their security issues.
achieved in [22, 39, 57]. In general, the public key is more
The pros and cons of each algorithm are also laid out in. Some
secure but slower than secret key.
critical issues are discussed in term of probability cracking,
The algorithm’s cracking probability is the probability of double layer of security, single and double hiding layers,
breaking the algorithm and obtaining the hidden confidential blindness, biologically preserved DNA, modification rate,
data. The reason for analyzing the probability of the scheme expansion of DNA reference, not pure algorithm, applying
being cracked is so as to distinguish the factors which ensure a substitution method, and capacity also. The objective of the
minimum cracking probability. The cracking probability is not comparison provided in this study is to equip the researchers
based on the number of attempts made before an intruder gains with the knowledge to conduct future research on improved
access to the secret hidden data, but is based on the inclusion of secure DNA data hiding techniques which are both efficient
every unknown variable applied in the algorithm so as to and reliable.
conceal the confidential data. Therefore, high probability
cracking leads for high security for the data hiding approach as
achieved in [18, 22, 33, 34, 46, 58, 61, 64]. ACKNOWLEDGEMENT
The substitution method is considered as a more efficient The authors greatly acknowledge the Research Management
method for hiding data in DNA. This method can preserve the Centre (RMC) Universiti Teknologi Malaysia and Ministry of
length of the DNA sequence, always keeping the payload at Higher Education (MOHE) for the financial support through
zero. Also, it has a more efficient capacity because it substitutes the Fundamental Research Grant Scheme (FRGS) No.
some of DNA nucleotides with secret data bits or other R.J130000.7828.4F860.
nucleotides based on the secret data as achieved in [31, 32, 35,
36, 37, 43, 52, 55, 59].
REFERENCES
In any data hiding technique, capacity is a critical part which
considered as a one of the main requirements of data hiding [1] Singh, G., A study of encryption algorithms (RSA,
techniques. It is imperative that a data hiding technique DES, 3DES and AES) for information security.
possesses a sizable hiding capacity. This capacity can be International Journal of Computer Applications, 2013.
described in terms of absolute measures such as the size of the 67(19).
secret message or using a relative value relative (for instance, [2] Subhedar, M.S. and V.H. Mankar, Current status and
data embedding rate, bits per pixel, bits per non-zero discrete key issues in image steganography: A survey.
cosine transform coefficient, or the ratio of the secret message Computer science review, 2014. 13: p. 95-113.
to the cover-medium, etc.). The DNA capacity can be measured
in terms of bit per nucleotide (pbn). Thus, improving the [3] Hamed, G., et al. Comparative study for various DNA
capacity of the hidden data is one of the major concern for based steganography techniques with the essential
researchers in this field, which has previously been achieved in conclusions about the future research. in Computer
[13, 18, 21, 22, 31, 33, 35, 36, 40, 45, 47, 49, 52, 55, 58, 59, 62, Engineering & Systems (ICCES), 2016 11th
63]. International Conference on. 2016. IEEE.

Thus, it can be concluded that the main objective for proposed [4] Amin, M.M., et al. Information hiding using
DNA based double layer hiding algorithms, double layer steganography. in Telecommunication Technology,
security which are encrypt confidential data then to hide high 2003. NCTT 2003 Proceedings. 4th National
capacity, blind, biologically preserved, low modification rate, Conference on. 2003. IEEE.
a zero-payload algorithm, not a pure system, and with high [5] Al-Mohammad, A., Steganography-based secret and
probability cracking. In [32, 33, and 58], a low modification reliable communications: Improving steganographic
rate, conserving the expansion length of DNA reference, capacity and imperceptibility. 2010, Brunel
blindness, conserving the DNA functionality, double layer of University, School of Information Systems,
security, high capacity, and the not a pure algorithm are Computing and Mathematics Theses.
proposed.
[6] Santoso, K.N., et al., Information Hiding in

15374
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 24 (2017) pp. 15363-15377
© Research India Publications. http://www.ripublication.com

Noncoding DNA for DNA Steganography. IEICE Computer Engineering, 2012. 2(3): p. 44.
Transactions on Fundamentals of Electronics,
[19] AL-WATTAR, A.H.S., MAHMOD, R.,
Communications and Computer Sciences, 2015.
ZUKARNAIN, Z. A., UDZIR, N.,, REVIEW OF DNA
98(7): p. 1529-1536.
AND PSEUDO DNA CRYPTOGRAPHY.
[7] Kumari, P. and R. Kapoor, Image Steganography for International Journal of Computer Science and
Data Embedding & Extraction using LSB Technique. Engineering (IJCSE), 2015. 4(4): p. 65-76.
International Journal of Computer Applications &
[20] Tornea, O., Contributions to DNA cryptography:
Information Technology, 2016. 9(2): p. 192.
applications to text and image secure transmission.
[8] Ashok, J., et al., Steganography: an overview. 2013, Université Nice Sophia Antipolis.
International Journal of Engineering Science and
[21] Abbasy, M.R., et al., DNA base data hiding algorithm.
Technology, 2010. 2(10): p. 5985-5992.
International Journal of New Computer Architectures
[9] Nickfarjam, A.M. and Z. Azimifar. Image and their Applications (IJNCAA), 2012. 2(1): p. 183-
steganography based on pixel ranking and Particle 192.
Swarm Optimization. in Artificial Intelligence and
[22] Skariya, M. and M. Varghese, Enhanced double layer
Signal Processing (AISP), 2012 16th CSI
security using RSA over DNA based data encryption
International Symposium on. 2012. IEEE.
system. International Journal of Computer Science&
[10] Sheelu, A., B., An Overview of Steganography. IOSR Engineering Technology (IJCSET), 2013. 4(06): p.
Journal of Computer Engineering (IOSR-JCE), 2013. 746-750.
11(1): p. 15-19.
[23] Shiu, H., et al., Data hiding methods based upon DNA
[11] Khalifa, A. LSBase: A key encapsulation scheme to sequences. Information Sciences, 2010. 180(11): p.
improve hybrid crypto-systems using DNA 2196-2208.
steganography. in Computer Engineering & Systems
[24] Mousa, H., et al., Data hiding based on contrast
(ICCES), 2013 8th International Conference on. 2013.
mapping using DNA medium. Int. Arab J. Inf.
IEEE.
Technol., 2011. 8(2): p. 147-154.
[12] Jain, S. and V. Bhatnagar. Analogy of various DNA
[25] Adleman, L.M., Molecular computation of solutions
based security algorithms using cryptography and
to combinatorial problems. Nature, 1994. 369: p. 40.
steganography. in Issues and Challenges in Intelligent
Computing Techniques (ICICT), 2014 International [26] OGIWARA, M. Simulating Boolean Circuits on DNA
Conference on. 2014. IEEE. Computers. in Proceedings of the 1st International
Conference on Computational Molecular Biology,
[13] Torkaman, M.R.N., N.S. Kazazi, and A. Rouddini,
1997. 1997. ACM Press.
Innovative approach to improve hybrid cryptography
by using DNA steganography. International Journal of [27] Clelland, C.T., V. Risca, and C. Bancroft, Hiding
New Computer Architectures and their Applications messages in DNA microdots. Nature, 1999.
(IJNCAA), 2012. 2(1): p. 224-235. 399(6736): p. 533-534.
[14] Bansod, S. and G. Bhure, Data encryption by image [28] Sureshraj, D. and V.M. Bhaskaran, Automatic DNA
steganography. Int. J. Inform. Comput. Technol. Int. sequence generation for secured cost-effective multi-
Res. Publ. House, 2014. 4: p. 453-458. cloud storage. 2012.
[15] Singh, K.U., Video steganography: text hiding in [29] Singh, A. and R. Singh. Information hiding techniques
video by LSB substitution. International Journal of based on DNA inconsistency: An overview. in
Engineering Research and Applications, 2014. 4(5): p. Computing for Sustainable Global Development
105-108. (INDIACom), 2015 2nd International Conference on.
2015. IEEE.
[16] Chandel, B., Jain, S., Video Steganography: A Survey.
IOSR Journal of Computer Engineering (IOSR-JCE) [30] Bhateja, A. and K. Mittal, DNA Steganography:
2016. 18(1): p. 11-17. Literature Survey on its Viability as a Novel
Cryptosystem. Journal of Computer Science and
[17] Yang, Y., Information analysis for steganography and
Engineering, 2015. 2(1): p. 8-14.
steganalysis in 3D polygonal meshes. 2013, Durham
University. [31] Guo, C., C.-C. Chang, and Z.-H. Wang, A new data
hiding scheme based on DNA sequence. Int. J. Innov.
[18] Atito, A., A. Khalifa, and S. Rida, DNA-based data
Comput. Inf. Control, 2012. 8(1): p. 139-149.
encryption and hiding using playfair and insertion
techniques. Journal of Communications and [32] Hamed, G., et al. Hybrid technique for

15375
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 24 (2017) pp. 15363-15377
© Research India Publications. http://www.ripublication.com

steganography-based on DNA with n-bits binary 2013. IEEE.

coding rule. in Soft Computing and Pattern
[45] Menaka, K. Message encryption using DNA
Recognition (SoCPaR), 2015 7th International
sequences. in Computing and Communication
Conference of. 2015. IEEE.
Technologies (WCCCT), 2014 World Congress on.
[33] 33 Ibrahim, F.E., H. Abdalkader, and M. Moussa. 2014. IEEE.
Enhancing the Security of Data Hiding Using Double
[46] Manna, S., et al. Modified technique of insertion
DNA Sequences. in Industry Academia Collaboration
methods for data hiding using DNA sequences. in
Conference (IAC).
Automation, Control, Energy and Systems (ACES),
[34] El-Latif, E.I.A. and M.I. Moussa, Chaotic 2014 First International Conference on. 2014. IEEE.
Information-hiding Algorithm based on DNA.
[47] Das, P. and N. Kar. A DNA based image
International Journal of Computer Applications
steganography using 2d chaotic map. in Electronics
(0975–8887) Volume, 2015. 122(10).
and Communication Systems (ICECS), 2014
[35] Agrawal, R., M. Srivastava, and A. Sharma. Data International Conference on. 2014. IEEE.
hiding using dictionary based substitution method in
[48] Majumdar, A., M. Sharma, and N. Kar, An Improved
DNA sequences. in Industrial and Information
Approach to Steganography using DNA
Systems (ICIIS), 2014 9th International Conference
Characteristics. 2014: p. 138-145.
on. 2014. IEEE.
[49] Das, P. and N. Kar. A highly secure DNA based image
[36] Taur, J.-S., et al., Data hiding in DNA sequences
steganography. in Green Computing Communication
based on table lookup substitution. International
and Electrical Engineering (ICGCCEE), 2014
Journal of Innovative Computing, Information and
International Conference on. 2014. IEEE.
Control, 2012. 8(10): p. 6585-6598.
[50] Chaudhary, H. and V. Bhatnagar. Hybrid approach
[37] Khalifa, A. and A. Atito. High-capacity DNA-based
for secure communication of data using chemical
steganography. in Informatics and Systems (INFOS),
DNA. in Confluence the Next Generation Information
2012 8th International Conference on. 2012. IEEE.
Technology Summit (Confluence), 2014 5th
[38] Huang, Y.-H., C.-C. Chang, and C.-Y. Wu, A DNA- International Conference-. 2014. IEEE.
based data hiding technique with low modification
[51] Yamuna, M., Elakkiya, A., Codons in Data Safe
rates. Multimedia tools and applications, 2014. 70(3):
Transfer. International Journal of Engineering Issues,
p. 1439-1451.
2015(2): p. 85-90.
[39] Mitras, B.A. and A. Abo, Proposed steganography
[52] Marwan, S., A. Shawish, and K. Nagaty. An Enhanced
approach using DNA properties. International Journal
DNA-based Steganography Technique with a Higher
of Information Technology and Business
Hiding Capacity. in BIOINFORMATICS. 2015.
Management, 2013. 14(1): p. 96-102.
[53] Manisha, B., P., Mohit, DOUBLE LAYERED DNA
[40] Bhattacharyya, D. and S.K. Bandyopadhyay, Hiding
BASED CRYPTOGRAPHY. IJRET: International
secret data in dna sequence. International Journal of
Journal of Research in Engineering and Technology,
Scientific &Engineering Research, 2013. 4(2).
2015. 4(4): p. 2321-7308.
[41] Haughton, D. and F. Balado, Biocode: Two
[54] Chakraborty, S., Bandyopadhyay, K. S., Data Hiding
biologically compatible algorithms for embedding
by Image Steganography Appling DNA Sequence
data in non-coding and coding regions of dna. BMC
Arithmetic. International Journal of Advanced
bioinformatics, 2013. 14(1): p. 121.
Information Science and Technology (IJAIST) 2015.
[42] Shyamasree, C. and S. Anees. Highly secure DNA- 44(44).
based audio steganography. in Recent Trends in
[55] Das, P., et al., An Improved DNA based dual cover
Information Technology (ICRTIT), 2013 International
steganography. Procedia Computer Science, 2015.
Conference on. 2013. IEEE.
46: p. 604-611.
[43] Liu, H., D. Lin, and A. Kadir, A novel data hiding
[56] Indora, S. Cascaded DNA cryptography and
method based on deoxyribonucleic acid coding.
steganography. in Green Computing and Internet of
Computers & Electrical Engineering, 2013. 39(4): p.
Things (ICGCIoT), 2015 International Conference
1164-1173.
on. 2015. IEEE.
[44] Haughton, D. and F. Balado. Security study of keyed
[57] Tank, R.M., H.D. Vasava, and V. Agrawal, DNA-
DNA data embedding. in Global Conference on Signal
Based Audio Steganography. Oriental journal of
and Information Processing (GlobalSIP), 2013 IEEE.
Computer Science and Technology, 2015. 8: p. 43-48.

15376
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 12, Number 24 (2017) pp. 15363-15377
© Research India Publications. http://www.ripublication.com

[58] Santoso, K., et al., Sector‐ based DNA information

hiding method. Security and Communication
Networks, 2016. 9(17): p. 4210-4226.
[59] Marwan, S., A. Shawish, and K. Nagaty, DNA-based
cryptographic methods for data hiding in DNA media.
Biosystems, 2016. 150: p. 110-118.
[60] Meena, K., Menaka, K., Subramanian, R. K., Secure
Message Transfer Using DNA Sequences. Journal of
Computational Intelligence in Bioinformatics, 2016.
9(1): p. 1-6.
[61] Khalifa, A., A. Elhadad, and S. Hamad, Secure Blind
Data Hiding into Pseudo DNA Sequences Using
Playfair Ciphering and Generic Complementary
Substitution. Appl. Math, 2016. 10(4): p. 1483-1492.
[62] Tuncer, T. and E. Avci, A reversible data hiding
algorithm based on probabilistic DNA-XOR secret
sharing scheme for color images. Displays, 2016. 41:
p. 1-8.
[63] Marwan, S., A. Shawish, and K. Nagaty, Utilizing
DNA Strands for Secured Data-Hiding with High
Capacity. International Journal of Interactive Mobile
Technologies, 2017. 11(2).
[64] Malathi, P., et al., Highly Improved DNA Based
Steganography. Procedia Computer Science, 2017.
115: p. 651-659.

15377