Conference PaperPDF Available

Video steganography using encrypted payload for satellite communication

Authors:

Figures

Content may be subject to copyright.
Video Steganography Using Encrypted Payload for
Satellite Communication
Swadhin Thakkar1, Kaustubh Shivdikar2, Chirag Warty3
Research Fellow1, Solutions Architect2, Chief Solutions Architect3
Quanical Innovation Lab, Mumbai, India123
thakkar.swadhin.a@ieee.org 1, kaustubh.c.shivdikar@ieee.org 2, chiragwarty@ieee.org 3
Abstract—In November 2014, a massive cyber-attack on a major
company leaked sensitive data including thousands of personal
records. Despite the fact that a lot of research and development
has gone in the past decade to develop robust encryption algo-
rithms, hackers have managed to break them. Effective infor-
mation security is an important aspect and cannot be under-
mined in today’s times. When an adversary intercepts encrypted
data, it can be known that a secret message was sent. This can
raise alarms and invite attacks. Moreover, many times it is im-
portant to communicate secretly without drawing any attention
and not raise suspicion even if the message is intercepted. We
feel that the key to heightened security is to not only encrypt the
data but to hide the fact that secret message is being sent. This is
implemented by steganography. Steganography is the science
of concealing information (secret message) in an apparently
innocuous media file (image, audio, video). We use an algorithm
to hide the message inside the carrier (cover) media. Although
steganography facilitates covert communication, it suffers from
attacks like detection, modification, extraction and destruction.
In this paper, we present a powerful combination of steganog-
raphy and cryptography - a novel method to realize a highly
secured level of communication. The proposed system encrypts
the secret message via secure encryption algorithms and spreads
it out over a broad bandwidth using spread spectrum technique.
Then, we embed it into the cover media without affecting the
perceptual fidelity as the amount of encoded information is
below the threshold of perception and is regarded as noise. The
original video is available at the receiver for successful retrieval
of the secret message. This paper achieves additional security
layer of data authentication and integrity check by adding
a Hash-based Message Authentication Code (HMAC) module
to the message. Further, we analyze the performance of the
proposed technique towards different detection, modification
and destruction attacks. The proposed blend of techniques
showcases a higher immunity to these attacks while maintaining
private, confidential communication. The proposed system is
briefly discussed in the context of satellite communication.
Keywords - Video Steganography, RSA Encryption, Spread Spec-
trum, HMAC, Bit Substitution, SSIM, Satellite Communication,
Inter-Satellite Link (ISL)
TABLE OF CONTENTS
1. INTRODUCTION......................................1
2. PREVIOUS WORK ...................................2
3. SYSTEM PARAMETERS ..............................3
4. SYSTEM MODEL .....................................4
5. SECURED INTER-SATELLITE COMMUNICATION ...8
6. CONCLUSION ........................................9
ACKNOWLEDGMENTS ..................................9
REFERENCES ...........................................9
BIOGRAPHY .......................................... 11
978-1-5090-1613-6/17/31.00 c
2017 IEEE
1. INTRODUCTION
Cryptography is the practice and study of techniques for
secure communication in the presence of third parties called
adversaries [1]. More generally, cryptography is about con-
structing and analyzing protocols that prevent third parties or
the public from reading private messages [2]. The growth
of cryptographic technology has raised some legal issues in
the information age. Cryptography’s potential for use as a
tool for espionage and sedition has led many governments
to classify it as a weapon and to limit or even prohibit
its utilization and export [3]. In some jurisdictions where
the use of cryptography is legal, laws permit investigators
to compel the disclosure of encryption keys for documents
relevant to an investigation [4] [5]. Therefore, it becomes
compelling to make use of other techniques to carry out secret
communication without inviting attention.
Present day requirements of security systems are confiden-
tiality, authenticity, integrity, and non-repudiation. The need
to have total secrecy in an open-systems environment is the
main idea behind steganography. Steganography is derived
from the Greek words ’stegos’ meaning ’cover’ and ’grafia’
meaning ’writing.’ Throughout history, steganography has
been used to secretly communicate information between
people. Few examples of steganography from the past are
discussed from [6].In ancient Greece, messengers used to
shave their head. Then a message would be written on their
heads and hair was allowed to grow back. He was sent
to deliver the message only after the hair grew back. The
recipient would shave off the messenger’s hair and receive
the message. Another method used in Greece was by peeling
the wax off a tablet which was covered in wax. A message
would be written underneath the wax, and then wax was re-
applied. The recipient of the message would simply remove
the wax from the tablet to view the message. During World
War 2, an invisible ink was used to write messages on pieces
of paper. Liquids such as milk and vinegar were used. It
appeared to the average person as just being blank pieces
of paper. However, when these substances are heated, they
darken and become visible to the human eye.
While cryptography keeps the content of the message secret,
steganography aims at hiding the existence of messages. Dig-
ital steganography is currently an active research area and has
applications in hidden communications, image authentication
and copyright protection. Information concealment can be
used for secure communication. Hiding messages inside
other seemingly harmless messages in a way that does not
allow any enemy even to detect that there is a second message
present is the essence of steganography. This feature enables
it to have true confidentiality and secrecy. Steganographic
techniques are also crucial to safety and privacy on the
internet, as most of the data is routed via public.
1
Steganography replaces redundant data in the cover media
with the secret information to be sent. The key idea is
inserting the information inside a cover media as noise. This
noise is similar to the one native to the communication of
data across noisy channels. This noise is regarded as one
generated as a consequence of the channel and hence is not
perceptible to humans or computer analysis, if it is kept at
low levels [7]. Without real awareness of the presence of
secret information or access to the original media, it is tough
for a party to know whether the information is embedded.
Moreover, even if they have doubt that secret information is
being sent, it is tough to find out what it is, when powerful
steganographic techniques are used. The requirement is
that the carrier media must contain a sufficient amount of
redundant data or noise to accommodate a secret message.
This requirement limits the types of data that we can use
in steganography [8]. The cover media can be text, image,
audio or video. Steganography can be used in different types
of data formats in today’s world of digital communication.
The popular data formats for cover media are .bmp, .doc, .gif,
.jpeg, .mp3, .txt and .wav. The secret message is a stream of
information which is in the form of bits.
Most of the current steganographic methods rely on two
parameters - the secret key and the strength of the stegano-
graphic algorithm. However, they either do not address
the issue of encryption of the payload before embedding or
just use one or more of the conventional cipher algorithms.
Hence, Westfeld et. al. concluded their CRYSTAL project
with an important observation that ”Crypto-Stego interaction
is not very well researched yet” [9]. Some authors have
discussed encryption of the payload before embedding [10].
Steganalysis is the study of detecting messages hidden using
steganography; this is analogous to cryptanalysis applied to
cryptography. Steganalysis involves the practice of attacking
steganographic methods by detection, destruction, extraction
or modification of embedded data [8]. The strength of embed-
ding algorithms depends on how immune is it to the various
attacks carried out. Cryptanalysis is considered successful
when the attacker retrieves information successfully from
the message. Whereas steganography involves an additional
criterion the attacker should not know about the presence of
a hidden message. Knowing the fact that there is information
hidden in the cover destroys the purpose of steganography.
To be effective, the embedded data should be [11]:
(1) Unobtrusive / Perceptually invisible - Its presence should
not interfere significantly with characteristics of the media.
(2) Robust - It should be difficult to remove and retrieve it. If
only partial knowledge is available (for example known cover
or known stego), then attempts to delete the message, should
result in severe degradation of the cover media. The stego file
should be immune to:
a. Common signal processing techniques like digital-to-
analog and analog-to-digital conversion, resampling, requan-
tization, and common signal enhancements like image con-
trast and color, or audio bass and treble change. It should
be possible to retrieve the message even if some level of
compression is applied.
b. Common geometric distortions like rotation, translation,
cropping and scaling.
(3) Universal - The data should be compatible with all kinds
of multimedia cover messages.
(4) Unambiguous Retrieval - The secret data should be accu-
rately retrieved at the receiver.
(5) Authentication - The receiver should be able to confirm
the validity and integrity of the message. The message should
be distinctly degraded in the event of an attack.
Neither Steganography nor Cryptography can be thought of
as the ultimate answer to open systems privacy, but using both
technologies together can provide very high levels of privacy
for anyone. Both Steganography and Cryptography are excel-
lent means for secure communication, but neither technology
alone is perfect, and both can be broken. Therefore, experts
would suggest using both to add multiple layers of security
[8].
Contribution - In this paper, we use a combo of cryptography
and steganography. A strong encryption algorithm combined
with clever steganography techniques can help achieve a
highly secured level of communication. This fusion helps
overcome disadvantages of both practices - disguises the
otherwise obviously encrypted data and encrypts the hidden
information to render extraction futile.
The paper is organized into following sections. Section
2 documents previous work. Section 3 enlists the various
parameters that govern a steganography system. Section
4 discusses the embedding and extraction process in the
proposed system. It also explains and evaluates the various
algorithms used. Section 5 describes the application in Inter-
Satellite Links (ISL). Conclusion is described in section 6.
2. PREVIOUS WORK
There have been various implementations of steganography
and steganalysis. Some of these also employ cryptography.
[12] Embeds the cipher in the image in an encrypted form
using a reference database instead of direct bit variations.
Asymmetric key cryptography is implemented using Data
Encryption Standard (DES) algorithm. The cipher sequence
can be decoded without the original image, and only the
edited image will be transmitted to the receiver.
This technique [13] uses least significant bit (LSB) steganog-
raphy as the basis and randomly disperses the secret message
over the full picture to make sure that the secret message
cannot be recovered easily from the picture. Detailed visual
and statistical analysis of the algorithm reveals that it yields
satisfactory results. The LSB substitution also uses plane
cycling to distribute bits in the red, blue and green (RGB)
planes randomly. A steganalysis method [14] is introduced
to detect the existence of hidden message that is randomly
embedded in LSB and the second significant bits (second-
LSBs) of image pixels. It is proposed based on investigating
the statistical characters of image data in LSB and the sec-
ond significant bits with hidden message. The application
discussed in this paper [15] ranks images in a users library
based on their suitability as cover objects for some data. By
matching data to a picture, there is less chance of an attacker
being able to use steganalysis to recover the data. Before
hiding the data in a picture, the application first encrypts
it. [16] Discusses the concept of random pixel embedding
instead of sequential embedding. This technique is a security
feature added to the usual LSB manipulation of the pixel. [17]
mentions about the upper bound on the number of bits that
can be hidden in an image without causing major statistical
changes. The importance of using encrypted messages is also
mentioned.
2
The technique [6] embeds the hidden information in the
spatial domain of the cover image and can verify if the
attacker has tried to edit, erase or forge the secret information
in the stego-image. [18] have presented a method of embed-
ding information within digital image via a combination of
techniques like spread spectrum, error control coding, and
image processing. The secret information is embedded within
the noise, which is then added to the digital image. [19]
have proposed to identify the existence of hidden message
and locate the position of the hidden information in the cover
image. [20] have suggested a steganalysis of a block Dis-
crete Cosine Transform (DCT) image steganography. Here
information is hidden in each of the 8 x 8 DCT blocks.
Because of the block structure of this technique, pairs of
neighboring pixels within a block have different statistics
from those of two 8 x 8 blocks. [21] have mentioned about the
use of gray level modification of an image to embed binary
data stream into the cover image. First, specific pixels are
transformed, and then the binary stream is integrated into
the picture. This algorithm allows secret communication.
The information is hidden and recovered within the spatial
domain of the image. This technique has low computational
complexity and high information hiding capacity. [22] Uses a
technique to encrypt the secret message and then embeds into
an animation. It explains the use of error control coding and
ASCII text embedding into frames of animation to propose a
robust steganographic algorithm.
Steganalysis attacks can be classified in different ways.
Wayner [23] divides common attack methods by functional
properties; attacks fall into visual or aural, structural and sta-
tistical. Figure 1 shows six categories of detection techniques
available for steganalysis [8]. It demonstrates that different
attacks are possible when the attacker has access to various
types of information.
Figure 1.Six types of detection attacks
[11] Proposes a spread spectrum technique where data is
embedded into the perceptually insignificant portions of the
image. The DCT of the cover image is taken, and the
information is hidden in the entire frequency spectrum. It
has been noted that the frequency domain methods are more
robust than the spatial domain techniques. The original image
is present at the receiver for successful retrieval of the secret
message. [7] Uses Spread Spectrum Image Steganography
(SSIS) by spreading the classified information with a wide-
band signal generated with a pseudorandom sequence. The
information is encrypted before being spread out. It uses
interleaving to not only correct for burst errors but provides
an additional level of security. It also mentions that LSB
methods are vulnerable to extraction by unauthorized parties.
[24] Mentions about watermarking of video via a combi-
nation of spread spectrum and modulation. Redundancy is
added to the secret message which is then modulated with a
pseudorandom number to embed the watermark. Because of
the properties of the pseudo-noise signal, it becomes difficult
for attacks to detect, locate and modify data.
3. SYSTEM PARAMETERS
There are few terms which need to be understood in the
context of steganography. These elements form the building
blocks of a steganography system. Figure 2 shows a typical
model of steganography.
Figure 2.Steganography Model
Definitions:
(1) Hidden message - This is the secret data which needs to
be communicated over a public channel. It can be a text,
image, audio or video.
(2) Cover Media - The carrier into which the secret infor-
mation is hidden. The message should be embedded in the
redundant regions of the cover media. It should be done in
such a way that there are no major alterations in the statistical
properties of the cover.
(3) Embedding Algorithm - It is defined as the technique
with which the message is embedded into the cover media.
The aim of the algorithm is to not only embed the information
efficiently but also not affect the quality of the cover media.
For example, the visual quality of an image.
(4) Stego Media - It is the synthesized image obtained by
the combination of the payload and cover media after the
application of the algorithm. The stego media should be
statistically similar to the cover media.
(5) Adversary - The attacker or hacker who is interested in
and intends to extract the hidden information from the stego
media in the communication channel.
(6) Perceptibility - It describes the ability to detect the pres-
3
ence of hidden information in a stego media. For example,
visual perception in an image. It is dependent on the nature
and complexity of embedding algorithm.
(7) Robustness - It defines how immune the payload is in the
event of attacks like modification and destruction on the stego
media. A robust algorithm makes it difficult for an attacker
even to detect the hidden message.
(8) Security - It the level of secrecy in the process. Factors
like confidentiality, authenticity and data integrity of the
hidden message are taken into account. The ability of the
receiver to verify error-free delivery of message corresponds
to a high level of security.
Error Analysis:
(1) Mean Square Error (MSE) - It is defined as the square
of the error between the cover and the stego. The distortion
in the stego can be measured using MSE. The cover image
is represented as ’cov’ and stego image as ’steg’ in the given
equation.
MSE =ΣM
i=1ΣN
j=1(cov(i, j)steg(i, j))2
MN(1)
whereM*Nistheimage size.
(2) Peak Signal to Noise Ratio (PSNR) - It is the ratio of the
maximum signal to noise in the stego image.
PSNR =20log10
255
MSE (2)
(3) Bit Error Rate (BER) - The error while retrieving hidden
information from a communication channel.
BER =1
|imagecov |Σallpixels |imagecov imagesteg |
(3)
(4) % BER - It is BER expressed in percentage.
%BER =BER 100 (4)
(5) Structural Symmerty (SSIM) - It is the measurement or
prediction of image quality based on an initial uncompressed
or distortion-free image as reference. It is a full reference
metric. SSIM is used for measuring the similarity between
two images [25].
SSIM(x, y )= (2μxμy+c1)(2σxy +c2)
(μ2
x+μ2
y+c1)(σ2
x+σ2
y+c2)(5)
where μxis average of x, μyis average of y, σ2
xis variance of
x, σ2
yis variance of y, σxy is covariance of x and y, k1=0.01,
k2=0.03, L is dynamic range i.e. 2bitsper pixel-1, c1=(k
1L)2,
c2=(k
2L)2.
4. SYSTEM MODEL
Every steganographic system discussed consists of two tech-
niques - the embedding and extraction process. The embed-
ding process accepts three inputs - the secret message, the
algorithm that defines the processes and the cover object. The
output of the embedding process is called the stego object.
When the stego object is presented as an input to the receiver,
it produces the secret message with the help of extraction
algorithm [26].
The Prisoner’s problem is a common example used to explain
the scenario for secure communication. The ’Warden’ in
this example is the communication channel between ’Alice’
and ’Bob’. Warden is the entity which can compromise the
communication between Alice and Bob. There are three
types of Wardens - passive, active and malicious. Passive
Warden only checks for the presence of secret messages and
is not allowed to modify it. Active Warden may intentionally
alter the message to destroy it when hidden information is
detected. A Malicious Warden may impersonate and try
various methods to extract the embedded information [27].
The proposed method suggests ways to counter all the three
types of wardens - passive, active and malicious.
There are two elements involved in making a robust stego
media - the message structure and the insertion strategy. For
the stego media to be secure, these two elements must be
designed correctly [11]. The proposed method describes the
various algorithms for both of these elements. A flow chart
of the method is shown in Figure 3.
Figure 3.Proposed Method
(A) Message Structure
RSA (Rivest Shamir Adleman) Encryption:
Encrypting the data before embedding provides in-depth se-
curity. This makes attacker’s task more difficult if the goal is
to extract and read the secret data. The statistical properties
of meaningful data can be known. However, the statistical
properties of the data post-encryption may not possess similar
patterns.
The payload in the proposed system is encrypted via the
RSA algorithm [28]. There are two advantages in doing so.
First, by using a public key the need for a shared private
key between the sender and recipient is discarded. Shared
keys are not practical because they require a secure way of
distributing the key to parties who may want to communicate.
This means sharing the key via a communication channel
which is subject to detection, modification, destruction and
4
extraction. A public key can be distributed easily by giving
it to the intended parties. There is no risk even if a public
key is leaked. Second, the RSA algorithm is widely known
and considerably secure if large prime numbers are used to
generate the keys. Using an algorithm like the RSA which
is public knowledge also means following Kerkchoffs second
principle. The principle states that ”A cryptographic system
should be secure even if everything about the system, except
the key, is public knowledge” [29]. Adhering to this principle
means adhering to open design of secure software systems.
Hence we use RSA and not any custom made encryption
algorithm [15].
Two large prime numbers are randomly picked. The public
and private key are dependent on this selection. The public
key is used for encryption. The private key is known to
the intended recipient, and decryption is possible only by
the intended party. Usually both the keys are generated at
the receiver. Only the public key is communicated to the
transmitter. The active warden will not be able to understand
what is embedded inside the stego media because of the use
of encryption.
Spread Spectrum Modulation:
The encrypted data is made further secure by using Direct
Sequence Spread Spectrum (DSSS) Technique. The data is
converted into a stream of bits and is spread over a broad
bandwidth with the help of a spreading signal. This spreading
over a large bandwidth makes the resulting wideband signal
appear as a noise signal which helps create greater resistance
to intentional and unintentional interference for the transmit-
ted signal [30].
The database is a Pseudo Random Number Generator
(PRNG) which generates long bit sequences known as Pseudo
Noise (PN) sequence - a code with shorter duration (larger
bandwidth) as compared to the data. The data is modulated
with this PN code which is equivalent to chopping data into
smaller pieces and spreading it over the large bandwidth.
The smaller the chip duration of the PN code, the larger
the bandwidth and therefore higher immunity to interference.
The resulting signal is similar to white noise and has a band-
width almost equal to the PN code. Attacks of information
extraction become computationally complex as the message
is now spread out.
Hash-based Message Authentication Code Generation:
A malicious warden might try to fake an identity in a com-
munication channel. It is imperative for the communicating
parties to verify the authenticity and integrity of the message.
The proposed method accomplishes this by generating a
Hash-based Message Authentication Code (HMAC) with the
help of a cryptographic hash function called Secure Hash
Algorithm (SHA). SHA is a mathematical operation which
generates a hash value for a given input data. Following are
the important reasons for using cryptographic hash functions
for data authentication:
(1) Variable size input.
(2) Fixed size output
(3) Collision resistance - Each input has a unique out-
put. There are no two x and y such that: H(x) = H(y).
where H is the cryptographic hash function.
(4) Pseudorandom - The output sequence is entirely random.
(5) One way - It is not possible to find x given y that:
H(x) = y. where H is the cryptographic hash function.
The cryptographic hash function used here is SHA-256. The
output is a 64-character hexadecimal number which corre-
sponds to 256 bits. HMAC is unique to the given input
message. HMAC is essentially a unique digital signature of
the message. Examples of HMAC using SHA-256 [31]:
(1) Input message: Hello
Output HMAC: 66a045b452102c59d840ec097d59d9467e1
3a3f34f6494e539ffd32c1bb35f18
(2) Input message: ’IhaveaDream.txt’ [32])
Output HMAC: 047488747b53209475291a30638400e73a0
038377f4e1aaab9edcb2d27270aaf
We can get to know of steganalysis attacks like modification
and destruction with HMAC. A new HMAC will be generated
for the message retrieved at the end. The generated HMAC
must match with transmitted HMAC for authentication and
data integrity. Any change in either the message sequence or
the HMAC sequence will result in an unsuccessful verifica-
tion.
(B) Insertion Strategy
Data Scrambler:
A data scrambler is a digital device that jumbles and encodes
the message at the sender’s end to make it unintelligible at
a receiver not equipped with a corresponding de-scrambling
device. The HMAC and the wideband data stream are scram-
bled together through random permutation. This permutation
further secures the system from detection attacks which aim
to find meaningful portions of data inside a stego media. A
passive warden will find it tough to conclude if meaningful
data is present. It is essential that insertion strategy should
offer a balance of 0s and 1s as well as be random enough
so that it can mimic the LSBs of the cover image [10]. The
modification attacks which involve destroying the message
without damaging other aspects of the stego media become
tough.
Video Steganography:
Figure 4 shows the data embedding algorithm used to hide the
data in the stego media. The input media is a video stream.
A video is a combination of two types of data - audio and
image. Three types of steganography are possible for a video
- only image, only audio, and combined audio-image. These
options correspond to a larger room for error for the warden
to conclusively state the presence of hidden information. In
the event of a suspicion that steganography is taking place, a
video proves to provide a higher level of secrecy.
Image Extraction—To minimize vulnerability, information in
the proposed model is hidden only in the images. The audio
stream is not used for steganography. Hence image frames are
extracted from the video stream. This selection strategy helps
prevent major changes in the statistical properties of the video
thereby reducing the possibility of attracting the attention of
an attacker. Additionally, a video being streamed at 30 frames
per second and with 1 million pixels a frame sums up to a
total of 30 million pixels per second (90 million pixels for
a color image). This gives the algorithm huge space to hide
the information. Videos with higher resolution streamed at
higher frame rate prove to be even better.
Color Plane Selection—Another layer of security is proposed
in the system by selecting only 1 out of 3 color planes (R
G B or Y Cb Cr) in a given color model. This aids in the
concealment of information as well as helps maintain the
statistical properties compared to embedding information in
5
Figure 4.Steganography - Data Embedding Algorithm
all the three color planes. The strength of the algorithm lies
in decreasing the probability of a change in the palette color
values of each pixel and in minimizing the visual distortion
that is introduced [33].
Algorithm—This algorithm forms the core of the proposed
system. It specifies the type and manner of substitution of
bits into the specified image color plane. The main idea is
to add noise-like signal to the video pixels. This signal is
below the threshold of perception [24]. Because we have
used a Pseudo Noise (PN) signal for DSSS modulation,
the embedded information is also noise-like and difficult to
locate and extract. The algorithm can be made to follow
particular functions which will aid in the process of seamless
embedding of data.
Image Bit Substitution—In the LSB substitution technique,
the information is usually hidden in a sequential fashion.
Hence the risk of information being uncovered is high as it
is susceptible to sequential scanning based techniques of the
active warden. The Random Pixel Manipulation Technique
attempts at overcoming this problem, where pixels are chosen
in a random fashion instead of a sequential one [16]. Such
techniques make it difficult for the active warden to detect
information [17] [34]. It would be a smarter move to insert
the message only into a subset of the available pixels. Figure
5 and Figure 6 show the random pixel allocation on a subset
of pixels available in an image frame.
Each letter has N number of bits, where N depends on the
length of the PN code. Each letter of a word is embedded as
payload into a single color plane of each frame, with N pixels
being modified in it.
Combination—After the image bit substitution, the untouched
color planes are recombined with the stego plane. The audio
is also combined with the stego image to produce the stego
video which is then transmitted over the communication
channel for the intended the recipient.
Figure 5.Random Pixel Manipulation Technique in Blue
plane with 350 pixels
Figure 6.Random Pixel Manipulation Technique in Red
plane with 2100 pixels
(C) Message Retrieval and Extraction:
Figure 7 lists the steps at the receiver end. It is essentially
an inversion of the operations involved in the encoding pro-
cess. The data extraction algorithm is separately illustrated in
Figure 8. The original cover video is required for successful
extraction of the secret message. The RSA Private Key,
PN Code, Data De-scrambler, Color Plane Selector and Bit
Extraction Algorithm are also needed for fruitful and error-
free message extraction. The SHA-256 HMAC module is
further used to ensure integrity and confidentiality of the
secret message that is extracted.
(D) Performance Analysis:
Methods like PSNR and MSE calculate the absolute error.
They assess perceptual image quality to quantify the visibility
of error between the reference and modified image and are
based on a variety of known properties of the human visual
system. Structural Similarity Index was developed with an
assumption that human visual system can detect structural
changes in an image. It is based on an assessment of
perceived change in the structural information of a scene,
along with including luminance and contrast masking [35].
This index, therefore, gives an idea of structural similarity
6
Figure 7.Message Retrieval and Authentication
Figure 8.Data Extraction Algorithm
between cover and stego image.
All results presented were performed on a 0.3 MP image
flowers.jpg. It is a Standard Definition (SD) color image with
a resolution of 640 by 480 pixels. The original image is
shown in Figure 9. A letter was embedded in each frame
as per the proposed method. For illustration, letter I was
embedded in the cover image. Figure 9 shows the stegoed
image obtained after embedding 7000 characters into the
cover image. The results indicate that there were no visual
changes.
For a video of SD resolution running at 30 frames per second,
Table 1 states the length of video required to embed each of
Figure 9.Cover vs Stego with 7000 pixels modified
the secret messages.
Table 1.Length of Stego Video
No. Secret Message File Video
Size Length
1. IhaveaDream.txt [32] 10 KB 320 secs
2. Address.txt [36] 2KB 58 secs
Various types of statistical analyses were performed on the
stego image to give an idea of the efficiency of embedding
algorithm. Table 2 depicts the change in parameters for
letters ’I’ and ’P’ when no. of modified pixels are increased.
Each letter is analyzed with different RSA keys. Figures 10
through Figure 13 summarize these results. The SSIM values
are illustrated in Table 2. These values are very close to
1.0 even when 4000 pixels are modified. It means that the
cover and stego image are structurally identical which is an
essential requirement in steganography.
Table 2.Image Parametes with change in length of PN
L MSE PSNR SSIM % BER
5 5.2083e-05 95.7350 0.999999966 0.0114
10 1.2695e-04 91.8655 0.999999960 0.0227
50 5.7942e-04 85.2720 0.999999849 0.1139
100 0.00111328 82.4359 0.999999393 0.2278
200 0.00236328 79.1668 0.999998715 0.4557
500 0.00414388 76.7279 0.999998677 0.8138
7
Figure 10.Mean Square Error with change in pixels
Figure 11.Peak Signal to Noise Ratio with change in
pixels
5. SECURED INTER-SATELLITE
COMMUNICATION
In satellite communication, signals are transmitted between
sender and receiver with the help of satellites in space. It is a
unique type of communication. The distance between sender
and receiver is not a limiting factor. With the use of active
satellites, a signal can be sent from one location on the globe
to another. This makes satellite communication strategic from
the viewpoint of national security. For example, a comman-
der sitting at the armed forces headquarters can directly com-
municate in real-time to a remotely located mobile soldier on
a mission. There are missions where signal connectivity is
not present, or it is not desirable to route the communication
via public networks. Satellite communication provides confi-
dential end to end communication in such scenarios. Signals
are relayed between a ground station (GWL) and satellite
multiple times before it reaches the intended user (UML).
This has a distinct disadvantage the number of uplinks
and downlinks required increase drastically. Inter-Satellite
link (ISL) can be used to overcome this bandwidth wastage.
Figure 12.Structural Similarity with change in pixels
Figure 13.Bit Error Rate with change in pixels
ISL is a direct connection between two satellites. Signals
originating at a ground station are uplinked to a satellite in
view. Then it is relayed via multiple ISLs to the satellite
connected to the intended user. A network of satellites is
used around the globe for the same, thus reducing the number
of links required. Each satellite is provided with three links
- the User Mobile Link (UML) for communication with a
mobile station, the Gateway Link (GWL) for communication
with an earth station and the Inter-satellite Link (ISL) for
communication between two satellites, which are close to
each other.Low Earth Orbit (LEO) satellite constellations can
provide high bit rate interconnections, which enable them to
provide low delay connections between two distant points on
the earth surface [37].
There are thousands of satellites hovering around the earth at
any moment. A signal can be intercepted during ISL, UML or
GWL. Figure 14 shows an example of an adversary trying to
intercept an ISL. Though tough and complex it is possible. It
is important for countries to safeguard the information being
sent via these links. Using steganography can help secure
8
this communication link. An adversary who may want to
intercept the channel finds nothing but innocuous media. It
does not raise suspicion and not invite attacks as in the case
when heavily encrypted data. The proposed hybrid system
can be used to realize private and secure communication via
ISL along with maintaining data integrity and confidentiality.
Figure 14.Adversary Satellite in ISL
6. CONCLUSION
The work presented in this paper is an effort to develop a
robust Steganographic System. The proposed combination
of techniques algorithm incorporates features from various
existing techniques to form an efficient steganographic algo-
rithm which can help in cover communication.
The science of Steganography requires that the cover image
must be carefully selected. A familiar image should not be
used for steganography. It is better for steganographers to
create their cover media [10]. The aim of any steganographic
algorithm to pass unnoticed [13]. The resulting stego image
obtained contains minimal visually perceptible changes. This
perceptibility is crucial because as discussed earlier, visual
perceptibility (visual quality) is one of the critical parameters.
The size of the payload is determined by the resolution and
frame rate of the video. The RSA-encrypted message is
modulated by a PN code. This information is scrambled ran-
domly along with the HMAC of the secret message, adding
randomness to the process. Then, this data is embedded into
the LSBs of pixels of a particular color plane in each of
the image frames of the cover video. Since information is
present in all the three color planes of the video, no single
plane carries meaningful information. This combination of
techniques makes retrieving the hidden message from a video
a tough task. The RSA keys, PN code, Plane Selector
and Data Scrambling algorithms are needed, failing which
extracting the information is useless. Further, the amount
of payload inside each frame is kept strategically low not to
raise suspicion and not invite detection attacks. As shown
by the results, the statistical properties of low-density stego
images are comparable to the cover image, thus giving no
information on the presence of a secret message.
The secret message to be communicated can be compressed
to maximize the net payload in a stego media. Compression
algorithms work on the redundant data present inside a mes-
sage by finding some patterns. An encryption algorithm will
randomize the data, leaving no patterns. Hence, the payload
should be compressed even before it is encrypted. Many types
of lossless compression techniques can be used. Correlated
Steganography can also be used to reduce the bits needed
to encode a message [10]. The proposed method uses only
image steganography. However, advancement in Voice over
Internet Protocol (VoIP) and various Peer-to-Peer (P2P) audio
services offer numerous opportunities for secret communica-
tion using audio steganography. Minor change in the binary
sequence of audio samples with existing steganography tools
can easily facilitate secret communication. Moreover, audio
signals have an inherent redundancy and unpredictable nature
which makes them ideal to be used as a cover media to
hide secret messages in secret communications [33]. Similar
techniques can be used to hide data inside the audio of a video
stream. Interleaving is the reordering of data to be transmitted
so that consecutive bytes of data are distributed over a larger
sequence of data. Interleaving is done to reduce the number
of burst errors. The use of interleaving significantly improves
the ability of error protection codes to rectify burst errors.
Usually, error protection coding processes can correct small
numbers of errors and not the ones which occur in groups
[38]. Forward secrecy feature can be applied to the public
key encryption. A unique random public key without using
a deterministic algorithm will be utilized for each session.
Therefore, compromise of a given message does not mean a
compromise of other messages. Moreover, there is no single
way to compromise multiple messages together.
Steganography can be used for a wide variety of applications
relating to national security. It can help aid the circula-
tion of vital and classified information. The applications
are numerous. For example, applying steganography in
intelligent restricted Content Based Image Retrieval (CBIR)
[39]. Steganography also has major advantages in inter-
satellite communication where secret information needs to be
shared but the risk of interception looms. However, there
are some applications which can misuse this science [10].
Examples include the use by criminals to communicate with
each other. The question whether child pornography exists
inside seemingly innocent media files and whether robust
stego files can bypass an anti-virus system to communicate
hidden information are few among many disturbing questions
which need to be addressed. However, it is clear without a
doubt that steganography can be used for useful applications.
Moreover, along with encryption, it becomes a formidable
tool for private and secured communication [10].
ACKNOWLEDGMENTS
The authors would like to thank the entire team at Quanical
Innovation Lab for their constant support, valuable inputs and
motivation given throughout the research project. We would
also like to take this opportunity to thank all the anonymous
reviewers for their time and valuable suggestions.
REFERENCES
[1] Rivest, Ronald L. (1990), ”Cryptography”, In J. Van
Leeuwen. Handbook of Theoretical Computer Science 1,
Elsevier.
[2] Bellare, Mihir; Rogaway, Phillip (21 September 2005),
”Introduction”, Introduction to Modern Cryptography, p.
10.
[3] ”Overview per country”, Crypto Law Survey, February
2013.
9
[4] ”UK Data Encryption Disclosure Law Takes Effect”, PC
World, 1 October 2007.
[5] Ranger, Steve (24 March 2015), ”The undercover war on
your internet secrets: How online surveillance cracked
our trust in the web”.
[6] K. S. Babu, K. B. Raja, K. K. Kiran, T. H. Manjula Devi,
K. R. Venugopal and L. M. Patnaik, ”Authentication of
secret information in image Steganography,” TENCON
2008 - 2008 IEEE Region 10 Conference, Hyderabad,
2008, pp. 1-6.
[7] L. M. Marvel, C. G. Boncelet and C. T. Retter, ”Spread
spectrum image steganography,” in IEEE Transactions
on Image Processing, vol. 8, no. 8, pp. 1075-1083, Aug
1999.
[8] S. M Thampi. Information Hiding Techniques: A Tuto-
rial Review.
[9] The CRYSTAL project. Available from:
http://bit.ly/2eiONGq
[10] Cheddad, J. Condell, K. Curran and P. Mc Kevitt,
”Digital Image Steganography: Survey and Analyses of
Current Methods”. Signal Processing, Volume 90, Issue
3, March 2010, Pages 727-752.
[11] J. Cox, J. Kilian, F. T. Leighton and T. Shamoon,
”Secure spread spectrum watermarking for multimedia,
in IEEE Transactions on Image Processing, vol. 6, no. 12,
pp. 1673-1687, Dec 1997.
[12] P. Marwaha and P. Marwaha, ”Visual cryptographic
steganography in images,” Computing Communication
and Networking Technologies (ICCCNT), 2010 Interna-
tional Conference on, Karur, 2010, pp. 1-6.
[13] Mishra, A. Gupta and D. K. Vishwakarma, ”Proposal
of a New Steganographic Approach,” Advances in Com-
puting, Control, and Telecommunication Technologies,
2009. ACT ’09. International Conference on, Trivan-
drum, Kerala, 2009, pp. 175-178.
[14] K. Zhang, H. y. Gao and W. s. Bao, ”Steganalysis
Method of Two Least-Significant Bits Steganography,”
Information Technology and Computer Science, 2009.
ITCS 2009. International Conference on, Kiev, 2009, pp.
350-353.
[15] M. Juneja and P. S. Sandhu, ”Designing of Robust
Image Steganography Technique Based on LSB Insertion
and Encryption,” Advances in Recent Technologies in
Communication and Computing, 2009. ARTCom ’09.
International Conference on, Kottayam, Kerala, 2009, pp.
302-305.
[16] S. Venkatraman, Ajith Abraham, M. Paprzycki, ”Signif-
icance of steganography on data security”, Information
Technology: Coding and Computing, 2004.
[17] R. Chandramouli and N. Memon, ”Analysis of LSB
based image steganography techniques,” Image Process-
ing, 2001. Proceedings. 2001 International Conference
on, Thessaloniki, 2001, pp. 1019-1022 vol.3.
[18] L. M. Marvel, C. T. Retter and C. G. Boncelet, ”Hid-
ing information in images,” Image Processing, 1998.
ICIP 98. Proceedings. 1998 International Conference on,
Chicago, IL, 1998, pp. 396-398 vol.2.
[19] N. F. Johnson and S. Jajodia, ”Steganalysis: The inves-
tigation of Hidden information,” in International Confer-
ence on Information Technology, pp. 113116, Jun 1998.
[20] Y. Wang and P. Moulin, ”Steganalysis of Block DCT
Image Steganography,” in International Conference on
Image Processing, pp. 339342, Mar 2003.
[21] V. M. Potdar and E. Chang, ”Grey level modification
steganography for secret communication,” Industrial In-
formatics, 2004. INDIN ’04. 2004 2nd IEEE Interna-
tional Conference on, Berlin, 2004, pp. 223-228.
[22] H. Sung, G. R. Tadiparthi and S. Mukkamala, ”Defeat-
ing the current steganalysis techniques (robust steganog-
raphy),” Information Technology: Coding and Comput-
ing, 2004. Proceedings. ITCC 2004. International Con-
ference on, 2004, pp. 440-444 Vol.1.
[23] P. Wayner, Disappearing Cryptography: Information
Hiding: Steganography and Watermarking.
[24] F. Hartung B. Girod, ”Copyright Protection in Video
Delivery Networks by Watermarking of Pre-Compressed
Video”, Multimedia Applications, Services and Tech-
niques ECMAST ’97 Volume 1242 of the series Lecture
Notes in Computer Science pp 423-436
[25] Available from https://en.wikipedia.org/wiki/
Structural similarity
[26] J. Fridrich, ”Steganography in Digital Media: Princi-
ples, Algorithms, and Applications”.
[27] I. Cox, M. Miller, J. Bloom, ”Digital Watermarking and
Steganography.”
[28] R. Rivest, A. Shamir and L. Adleman, ”Cryptographic
communications system and method”, U.S. Patent: 4 405
829 A, issued September 20, 1983.
[29] Kerckhoffs, ”La cryptographie militaire”, Journal des
sciences militaires, vol. IX, pp. 583, January 1883, pp.
161191, February 1883.
[30] S. Haykin, ”Communication systems (4th edition)”,
John Wiley & Sons, 2008 pp. 48899.
[31] Available from: http://hash.online-convert.com/sha256-
generator
[32] Speech of Martin Luther King, Jr. , Available
from: http://www.americanrhetoric.com/
speeches/mlkihaveadream.htm
[33] N. Meghanathan and L. Nayak, ”Steganalysis algo-
rithms for detecting the hidden information in image,
audio and video cover media”, International Journal of
Network Security and Its Application (IJNSA), Vol.2,
No.1, January 2010.
[34] N.F. Johnson, Z. Duric, and S. Jajodia, ”Information
Hiding: Steganography and Watermarking - attacks and
countermeasures,” Kluwer Academic Publishers, 2000
[35] Z. Wang, A. C. Bovik, H. R. Sheikh and E. P. Simon-
celli, ”Image quality assessment: from error visibility
to structural similarity,” in IEEE Transactions on Image
Processing, vol. 13, no. 4, pp. 600-612, April 2004.
[36] Speech by President Abraham Lincoln.
Available from https://en.wikipedia.org/wiki/
Gettysburg Address
[37] E. Papapetrou, I. Gragopoulos and F.-N. Pavlidou, Per-
formance evaluation of LEO satellite constellations with
inter-satellite links under self-similar and Poisson traffic,
International Journal of Satellite Communications 17,51-
64 (1999)
[38] Available from: http://www.wirelessdictionary.com/
Wireless-Dictionary-Interleaving-Definition.html
[39] X. Li, ”Watermarking in secure image retrieval”, Pattern
Recognition Letters, vol. 24, no. 14, 2003 pp. 2431-2434.
10
BIOGRAPHY[
Swadhin Thakkar is currently a re-
search fellow at Quanical Innovation
Lab at Mumbai, India. His research ar-
eas span Cyber-Security, Networks, Em-
bedded Systems and Machine Learning.
He received his Bachelor of Technology
from Veermata Jijabai Technological In-
stitute (VJTI) at University of Mumbai.
He majored in the field of Electron-
ics and Telecommunications. He is an
incoming graduate student at Carnegie Mellon University
(CMU). Swadhin has worked on health technology projects
with MIT Media Labs and Camera Culture Group. He has
also worked at National Centre for Excellence in Technology
of Internal Security (NCETIS) - an Initiative by Indian Insti-
tute of Technology, Bombay and Department of Electronics
and Information Technology (Government of India) under
Digital India Program. His work included research and
development of a Remotely Operated All-Terrain Vehicle for
security and surveillance. He has also worked on projects in
the field of embedded systems, circuit design, and computer
vision. He is a recipient of the ’Sir Ratan Tata Trust Schol-
arship’ and ’R. D. Sethna Scholarship Fund’ for excellent
overall performance.
Kaustubh Shivdikar is currently a Re-
search Fellow at the Quanical Inno-
vation Lab and Solutions Architect at
Quanical Technologies Pvt. Ltd. Kaus-
tubh has been in the field of Network
Security for a long time. His latest
work engulfs risk assessment of cryptog-
raphy algorithms, especially DES, RSA,
and HASH. He is the youngest student
from his college to represent it at three
international conferences including TED, Indian Science
Congress and AeroConf 2016. Exploring the depths of
artificial intelligence, he is currently working on Network
Intrusion Detection using Artificial Immune System. He was
an integral part of the MIT Media Lab India Initiative where
his work revolved around Augmented Reality based projects
like the Magic Pillow and ARtifact. Bagging a scholarship at
the national level, he has received the National Talent Search
Examination (NTSE) Award. He has previously worked in
close associations with MIT Media Lab, Harvard Medical
School and the University of Massachusetts Lowell as a
research intern. An entrepreneur with a startup of his own,
Kaustubh has received the Innovation Award at Maker Fest
India.
Chirag Warty is currently the CEO of
Quanical. He is also the Chief Scientist
at Quanical Innovation Lab. He man-
ages an extensive network of research
teams in the US and India. He re-
ceived his Bachelor of Science in Elec-
trical Engineering from University of
Mississippi, USA, Masters of Engineer-
ing from University of Illinois Chicago,
USA. His other alma maters include
Stanford, Cornell, Univ. of Illinois Urbana-Champagne
(UIUC), UCLA, UC Berkeley, UC San Diego, USC. He
also has several academic accolades in communications and
signal processing from several IEEE societies. He is a
visiting faculty for University of Mumbai, VJTI, and SNDT
University for management, IT, and engineering programs.
He currently manages several projects and is responsible for
technology related global strategic decisions. His research
interest includes Software Defined Networking (SDN), 4G/5G
wireless technologies and Cyber-Physical Systems, Big data
in Mobile Cloud Computing, Physical Layer Security. He
has organized several panels and hosted forums on various
technical subjects. He is a well-published author with several
international publications and is a well-renowned speaker for
future technologies. He holds several key and executive posi-
tions on IEEE societies. He is currently the guest editor for
Elsevier’s Special Issue on Big Data Inspired Data Sensing,
Processing and Networking Technologies. He serves on the
editorial boards of IEEE and non-IEEE ventures. He has also
served as chair, co-chair (general or TPC) and international
advisor in various IEEE conferences around the globe.
11
... Encryption provides security as data is transmitted over the internet. However, classical encryption schemes require that data be decrypted prior to performing computation, exposing sensitive data to untrusted cloud providers [3], [4]. Using Homomorphic Encryption (HE) allows computations to be run directly on encrypted operands, offering ideal security in the cloud-computing era ( Figure 1). ...
Preprint
Full-text available
Homomorphic Encryption (HE) enables users to securely outsource both the storage and computation of sensitive data to untrusted servers. Not only does HE offer an attractive solution for security in cloud systems, but lattice-based HE systems are also believed to be resistant to attacks by quantum computers. However, current HE implementations suffer from prohibitively high latency. For lattice-based HE to become viable for real-world systems, it is necessary for the key bottlenecks - particularly polynomial multiplication - to be highly efficient. In this paper, we present a characterization of GPU-based implementations of polynomial multiplication. We begin with a survey of modular reduction techniques and analyze several variants of the widely-used Barrett modular reduction algorithm. We then propose a modular reduction variant optimized for 64-bit integer words on the GPU, obtaining a 1.8x speedup over the existing comparable implementations. Next, we explore the following GPU-specific improvements for polynomial multiplication targeted at optimizing latency and throughput: 1) We present a 2D mixed-radix, multi-block implementation of NTT that results in a 1.85x average speedup over the previous state-of-the-art. 2) We explore shared memory optimizations aimed at reducing redundant memory accesses, further improving speedups by 1.2x. 3) Finally, we fuse the Hadamard product with neighboring stages of the NTT, reducing the twiddle factor memory footprint by 50%. By combining our NTT optimizations, we achieve an overall speedup of 123.13x and 2.37x over the previous state-of-the-art CPU and GPU implementations of NTT kernels, respectively.
... Encryption provides security as data is transmitted over the internet. However, classical encryption schemes require that data be decrypted prior to performing computation, exposing sensitive data to untrusted cloud providers [3], [4]. Using Homomorphic Encryption (HE) allows computations to be run directly on encrypted operands, offering ideal security in the cloud-computing era ( Figure 1). ...
Conference Paper
Full-text available
Homomorphic Encryption (HE) enables users to securely outsource both the storage and computation of sensitive data to untrusted servers. Not only does HE offer an attractive solution for security in cloud systems, but lattice-based HE systems are also believed to be resistant to attacks by quantum computers. However, current HE implementations suffer from prohibitively high latency. For lattice-based HE to become viable for real-world systems, it is necessary for the key bottlenecks-particularly polynomial multiplication-to be highly efficient. In this paper, we present a characterization of GPU-based implementations of polynomial multiplication. We begin with a survey of modular reduction techniques and analyze several variants of the widely-used Barrett modular reduction algorithm. We then propose a modular reduction variant optimized for 64-bit integer words on the GPU, obtaining a 1.8× speedup over the existing comparable implementations. Next, we explore the following GPU-specific improvements for polynomial multiplication targeted at optimizing latency and throughput: 1) We present a 2D mixed-radix, multi-block implementation of NTT that results in a 1.85× average speedup over the previous state-of-the-art. 2) We explore shared memory optimizations aimed at reducing redundant memory accesses, further improving speedups by 1.2×. 3) Finally, we fuse the Hadamard product with neighboring stages of the NTT, reducing the twiddle factor memory footprint by 50%. By combining our NTT optimizations, we achieve an overall speedup of 123.13× and 2.37× over the previous state-of-the-art CPU and GPU implementations of NTT kernels, respectively.
... 1. Data encryption: AES, SHA1, SHA2, Twofish [16,46,24,67,25,101,58,98]; 2. Data compression and Decompression: zip files, JPEG and PNG image compression [80,59]; ...
Preprint
Full-text available
Sparse matrices, more specifically SpGEMM kernels, are commonly found in a wide range of applications, spanning graph-based path-finding to machine learning algorithms (e.g., neural networks). A particular challenge in implementing SpGEMM kernels has been the pressure placed on DRAM memory. One approach to tackle this problem is to use an inner product method for the SpGEMM kernel implementation. While the inner product produces fewer intermediate results, it can end up saturating the memory bandwidth, given the high number of redundant fetches of the input matrix elements. Using an outer product-based SpGEMM kernel can reduce redundant fetches, but at the cost of increased overhead due to extra computation and memory accesses for producing/managing partial products. In this thesis, we introduce a novel SpGEMM kernel implementation based on the row-wise product approach. We leverage atomic instructions to merge intermediate partial products as they are generated. The use of atomic instructions eliminates the need to create partial product matrices. To evaluate our row-wise product approach, we map an optimized SpGEMM kernel to a custom accelerator designed to accelerate graph-based applications. The targeted accelerator is an experimental system named PIUMA, being developed by Intel. PIUMA provides several attractive features, including fast context switching, user-configurable caches, globally addressable memory, non-coherent caches, and asynchronous pipelines. We tailor our SpGEMM kernel to exploit many of the features of the PIUMA fabric. This thesis compares our SpGEMM implementation against prior solutions, all mapped to the PIUMA framework. We briefly describe some of the PIUMA architecture features and then delve into the details of our optimized SpGEMM kernel. Our SpGEMM kernel can achieve 9.4x speedup as compared to competing approaches.
... 1. Data encryption: AES, SHA1, SHA2, Twofish [16,46,24,67,25,101,58,98]; 2. Data compression and Decompression: zip files, JPEG and PNG image compression [80,59]; ...
Thesis
Full-text available
Sparse matrices, more specifically Sparse Matrix-Matrix Multiply (SpGEMM) kernels, are commonly found in a wide range of applications, spanning graph-based path-finding to machine learning algorithms (e.g., neural networks). A particular challenge in implementing SpGEMM kernels has been the pressure placed on DRAM memory. One approach to tackle this problem is to use an inner product method for the SpGEMM kernel implementation. While the inner product produces fewer intermediate results, it can end up saturating the memory bandwidth, given the high number of redundant fetches of the input matrix elements. Using an outer product-based SpGEMM kernel can reduce redundant fetches, but at the cost of increased overhead due to extra computation and memory accesses for producing/managing partial products. In this thesis, we introduce a novel SpGEMM kernel implementation based on the row-wise product approach. We leverage atomic instructions to merge intermediate partial products as they are generated. The use of atomic instructions eliminates the need to create partial product matrices, thus eliminating redundant DRAM fetches. To evaluate our row-wise product approach, we map an optimized SpGEMM kernel to a custom accelerator designed to accelerate graph-based applications. The targeted accelerator is an experimental system named PIUMA, being developed by Intel. PIUMA provides several attractive features, including fast context switching, user-configurable caches, globally addressable memory, non-coherent caches, and asynchronous pipelines. We tailor our SpGEMM kernel to exploit many of the features of the PIUMA fabric. This thesis compares our SpGEMM implementation against prior solutions, all mapped to the PIUMA framework. We briefly describe some of the PIUMA architecture features and then delve into the details of our optimized SpGEMM kernel. Our SpGEMM kernel can achieve 9.4x speedup as compared to competing approaches.
... Main limitation of this work is stego analysis part has missed. It extends to that some other methods also using to enhance the capacity of payload embedding into images like LSB method as well as attacks on stegaanalysis [12][13]. ...
Article
In current days, there is a constant evolution in modern technology. The most predominant usage of technology by society is the internet. There are many ways and means on the internet through which data is transmitted. Having such rapid and fast growth of communicating media also increases the exposure to security threats, causing unintellectual information ingress. Steganography is the main aspect of communicating in an aspect that hides the extent of communication. Steganalysis is another essential concern in data concealing, which is the art of identifying the existence of steganography. A framework has been designed to identify the concealed data in the multimedia file in the proposed system. This work's main strength is analyzing concealed data images without embedding and extracting the image's payloads. A quantitative steganalysis approach was considered to accomplish the proposed objective. By using this approach, the results were achieved with 98% accuracy.
Article
Full-text available
A real-time simulation study for the evaluation of traffic flow in low earth orbit (LEO) satellite constellations, used for the interconnection of high-speed networks, is presented in this paper. The proposed model simulates the traffic process end to end at the packet level, supporting successfully the implementation of self-similar traffic sources, a modelling approach that has been considered more realistic than the well-known Poisson, for real-time communications. An in-depth study for the establishment of inter-satellite links (ISLs) and the design of the terrestrial and space segments is presented and the performance of the integrated system is evaluated in terms of delay and throughput parameters. Copyright
Article
Objective methods for assessing perceptual image quality have traditionally attempted to quantify the visibility of errors between a distorted image and a reference image using a variety of known properties of the human visual system. Under the assumption that human visual perception is highly adapted for extracting structural information from a scene, we introduce an alternative framework for quality assessment based on the degradation of structural information. As a specific example of this concept, we develop a Structural Similarity Index and demonstrate its promise through a set of intuitive examples, as well as comparison to both subjective ratings and state-of-the-art objective methods on a database of images compressed with JPEG and JPEG2000. A MatLab implementation of the proposed algorithm is available online at http://www.cns.nyu.edu/~lcv/ssim/.
Conference Paper
Digital video data can be copied repeatedly without loss of quality. Therefore, copyright protection of video data is a more important issue in digital video delivery networks than it was with analog TV broadcast. One method of copyright protection is the addition of a "watermark" to the video signal which carries information about sender and receiver of the delivered video. Therefore, watermarking enables identification and tracing of different copies of video data. Applications are video distribution over the World Wide Web (WWW), pay-per-view video broadcast, or labeling of video discs and video tapes. In the mentioned applications, the video data is usually stored in compressed format. Thus, the watermark must be embedded in the compressed domain. We present a scheme for robust watermarking of MPEG-2 encoded video. The scheme is of much lower complexity than a complete decoding process followed by watermarking in the pixel domain and re-encoding. Although an existing MPEG-2 bitstream is partly altered, the scheme avoids drift by adding a drift compensation signal. The scheme has been implemented and the results confirm that a robust watermark can be embedded into MPEG encoded video which can be used to securely transmit arbitrary binary information at a data rate of several bytes/second. The scheme is easily applicable to other video coding schemes like MPEG-1, H.261, and H.263.
Chapter
This chapter discusses the theory of cryptography. Cryptography is about communication in the presence of adversaries. Cryptology provides methods that enable a communicating party to develop trust that his communications have the desired properties, despite of the best efforts of an untrusted party. The desired properties might include: (1) privacy: an adversary learns nothing useful about the message sent; (2) authentication: the recipient of a message can convince himself that the message as received originated with the alleged sender; (3) signatures: the recipient of a message can convince a third party that the message as received originated with the alleged signer; (4) minimality: nothing is communicated to other parties except that which is specifically desired to be communicated; (5) simultaneous exchange: something of value is not released until something else of value is received; and (6) coordination: in a multi-party communication, the parties are able to coordinate their activities toward a common goal even in the presence of adversaries.
Article
List of Figures. List of Tables. Preface. 1. Introduction. 2. Exploring Steganography. 3. Steganalysis: Attacks Against Hidden Data. 4. Counter Measures to Attacks. Appendix A: Hiding Data in Network Traffic. Appendix B: Glossary of Methods to Distort Stego-Images. References. Index.
Book
Digital audio, video, images, and documents are flying through cyberspace to their respective owners. Unfortunately, along the way, individuals may choose to intervene and take this content for themselves. Digital watermarking and steganography technology greatly reduces the instances of this by limiting or eliminating the ability of third parties to decipher the content that he has taken. The many techiniques of digital watermarking (embedding a code) and steganography (hiding information) continue to evolve as applications that necessitate them do the same. The authors of this second edition provide an update on the framework for applying these techniques that they provided researchers and professionals in the first well-received edition. Steganography and steganalysis (the art of detecting hidden information) have been added to a robust treatment of digital watermarking, as many in each field research and deal with the other. New material includes watermarking with side information, QIM, and dirty-paper codes. The revision and inclusion of new material by these influential authors has created a must-own book for anyone in this profession. *This new edition now contains essential information on steganalysis and steganography *New concepts and new applications including QIM introduced *Digital watermark embedding is given a complete update with new processes and applications.
Article
pp. 161–191, Feb. 1883.
Article
Steganography, the art of hiding of information in apparently innocuous objects or images, is a field with a rich heritage, and an area of rapid current development. This clear, self-contained guide shows you how to understand the building blocks of covert communication in digital media files and how to apply the techniques in practice, including those of steganalysis, the detection of steganography. Assuming only a basic knowledge in calculus and statistics, the book blends the various strands of steganography, including information theory, coding, signal estimation and detection, and statistical signal processing. Experiments on real media files demonstrate the performance of the techniques in real life, and most techniques are supplied with pseudo-code, making it easy to implement the algorithms. The book is ideal for students taking courses on steganography and information hiding, and is also a useful reference for engineers and practitioners working in media security and information assurance.