Master’s Project Report
Matrix Encryption Algorithm
Sandeep M Chandrashekaregowda
Bachelor of Engineering VTU, India
A Project
Submitted to the Graduate School Faculty of the
University of Colorado at Colorado Springs
In Partial Fulfillment of the Requirements
For the Degree of Master of Science
Department of Computer Science
Spring 2014
This project for the Master of Science degree by
Sandeep M Chandrashekaregowda
Has been approved for the
Department of Computer Science
By
________________________________________________________________
Dr. C. Edward Chow
Date
________________________________________________________________
Dr. Rory Lewis
Date
________________________________________________________________
Dr. Jia Rao
Date
1. INTRODUCTION
With the advancement of the ages, man has greatly found the need to communicate through
distances. Initially this being accomplished through snail-mail was not real enough. He wanted to
communicate vital moments of his file, his thoughts through the usage of more realistic means by
the usage of multimedia, which is nothing but audio and video, which helped to share interesting
thoughts, interesting audio/video files among people. Sharing of such files often requires
communicating through networks of computers, which is not always secure enough. It is often a
requirement that the file being shared is only visible or usable by the intended recipient, sometimes it
is also may be essential to disguise the intruder of the file being different than what it really is. And
in some commercial purposes it also may be required that only parts of the communicated
audio/video files are playable. This arouses the need to device a methodology to securely
communicate these multimedia files and hence protect the intellectual property of multimedia from
attacks arising out of a hostile network environment.
1.1. Aim and Objective of the project:
Aim:
To device a methodology by which the video and audio files are secured in time and space efficient
manner.
Objective:
To device an encryption methodology that utilizes various available encryption techniques and helps
secure multimedia data files in such a manner that securing only information in the frame data
provides an effect of securing the file as a whole. This securing process is to be carried out in a
manner so as to reduce the amount of time used up in securing the file.
1.2. Need of Secure communication
With the advent and consequent vast growth of the Internet. Intellectual property has become
vulnerable to a number of threats that range from information retrieval to destruction of the
intellectual property. Hence one has found the extensive need to secure such intellectual property.
Intellectual property in the form of multimedia data files has been under constant threat over the
years. Given the fact that often files (including multimedia) would need to be communicated
through possibly insecure channels where an imposter or an intruder may cause extensive damage to
such intellectual property. It has become the need of the hour that methods are devised to ensure
secure communication of such file.
2. Background and Prior Work
2.1. SECMPEG by Meyer and Gadegast, 1995
In 1995 Meyer and Gadegast introduced the encryption method called Secure WPEG, or
shortly SECMPEG, designed for the MPEG-1 video standard. The SECMPEG I contains four
different levels of security. At the first level, SECMPEG encrypts the [headers from the sequence
layer to the slice layer, while the motion vectors and DCT [blocks are unencrypted. At the second
level, most relevant parts of the I-blocks are I additionally encrypted (upper left corner of the block).
At the third level, SECMPEG encrypts all I-frames and all I-blocks. Finally, at the fourth level,
SECMPEG encrypts the whole MPEG-1 sequence. The authors chose Data Encryption Standard
(DES) symmetric key cryptosystem, which was the natural choice, given that this cryptosystem had
been around since 1976 and was the official symmetric encryption algorithm standardized by
National Institute of Standard and Technology (NIST) and adopted by the US Government. Since
DES is a symmetric key cryptosystem, it could only be used to achieve confidentiality. Meyer and
Gadegast targeted solving the problem of data integrity as well. For that reason, the Cyclic Redundancy-Check (CRC) was incorporated as a low-level solution to the integrity. The real data
integrity mechanisms that included public key cryptography and cryptographically good hash
functions such as MD4, MD5, or SHA were left for further research. [4]
The encryption in SECMPEG (levels 1, 2, and 3) has some weaknesses. It is own that even
though single P- or B-frame on its own carries almost no information tout the corresponding Iframe, a series of P- or B-frames can tell a lot if their base I¬-frames are correlated. Since
SECMPEG introduces changes to the MPEG-1 format, a special encoder and decoder is needed to
handle SECMPEG streams. Nevertheless, the SECMPEG paper and implementation my Meyer and
Gadegast was one of the first important research initiatives for selective encryption of multimedia
streams.[4]
2.2 Video Encryption Algorithm by Qiao and Nahrstedt, 1997
The Video Encryption Algorithm (VEA) by Qiao and Nahrstedt is constructed with the goal to exploit
the statistical properties of the MPEG video standard. The algorithm consists of the following four
steps:
Step 1: Let the 2n byte sequence, denoted by ala2...a2n, represent the chunk of an I-frame
Step 2: Create two lists, one with odd indexed bytes ala3...a2n-l, and the other with even indexed
bytes a2a4...a2n.
Step 3: XOR the two lists into an n-byte sequence denoted with clc2...en
Step 4: Apply the chosen symmetric cryptosystem E (for example DBS or AES) with the secret key
KeyE on either odd list or even list, and thus create the cipher text sequence clc2...cn
EKeyE(ala3...a2n-l) or clc2...cnEKeyE(a2a4...a2ri) respectively.[1]
2.3. Video Encryption Methods by Alattar, Al-Regib and Al-Semari, 1999
In 1999, Alattar, Al-Regib and Al-Semari presented the three methods for selective video encryption
based on DES cryptosystem. These methods, called simply Method /, Method II and Method III, were
computationally improved versions of the previous work from two of the co-authors, which is
referred to as Method 0. The first algorithm (Method 0), proposed by Alattar and Al-Regib in,
essentially encrypts all macro blocks from I-frames and the headers of all prediction macro blocks
using DES cryptosystem. This method performs relatively poorly because encryption is carried out
on 40%-79% of the MPEG video stream. [20]
In Method I, the data of every nth macro block from the I-frame of MPEG video stream is
encrypted using DES cryptosystem, while the information from the all other I-frame macro blocks
is left unencrypted. The value of n was not specified, and it can be chosen depending on the
application needs. If the value of n is 2 then the encryption is performed on approximately a half of
all I-frame macro blocks, but the security level is higher. On the other hand, if the value of n is
higher, the computational savings are bigger, yet the security level is lower. An important
observation is that even though the certain number of I-macro blocks is left unencrypted, they are
not expected to reveal any information about the encrypted ones. [20]
To improve the security of Method I, Alattar, Al-Regib and Al-Semari suggested Method II, which
additionally encrypts the headers of all predicted macro blocks using DES. Since DES is a block
cipher that operates on 64-bit blocks, a 64-bit segment starting from the header of a predicted macro
block is processed in the beginning. This segment may include exactly the whole header (which is
the rare case when header size is equal to 64 bits), a part of the header (when header size is larger
than 64 bits), or the whole header along with the part of the macro block data (when the header size
is smaller than 64 bits). In the case when the encrypted segment contains a part of the header, an
adversary would have serious problems with synchronization, which adds to the security regarding
motion vectors. [2] The security is further increased if the encrypted segment also contains a part of
the macro block data. The computation performed using Method II is clearly faster than that of the
Method 0, but slower than that of Method I. Finally, Alattar, Al-Regib and Al-Semari proposed
Method III to reduce the amount of computation from Method II. Namely, instead of encrypting all
predicted macro blocks, the encryption in Method III is performed on every nth predicted macro
block, along with encrypting every nth I-macro block. [20]
2.4 Partial Encryption Algorithms for Videos by Cheng and Li, 2000
The partial encryption schemes for still images introduced by Cheng and Li are also further extended
to the videos. The approaches proposed by Cheng and Li are not suitable for JPEG image
compression, and thus naturally also not suitable for the MPEG video compression standard.
Instead, the partial encryption algorithms are designed for the video compression methods, which
use either quadtree compression or wavelet compression based on zero trees for the video sequence
intraframes, motion compensation, and residual error coding. For example, the partial encryption is
applicable to the videos that are based on the Set Partitioning In Hierarchical Trees (SPIHT) image
compression algorithm, which is an application of zerotree wavelet compression. Cheng and Li's
partial encryption algorithms are designed to disguise the intraframes (I-frames), the motion vectors,
and the residual error code of the given video sequences. In both quadtree compression and wavelet
compression based videos, all I-frames are encrypted using the previously discussed methods for
partial encryption of still images by Cheng and Li. In addition, it is also important to encrypt the
motion vectors. If the motion vector information is unencrypted, the adversary may be able to use
an image frame to obtain approximations to the successive frames. Almost all motion estimation
algorithms divide the frame into blocks and try to predict their movement (the position in the next
frame) by constructing the estimated motion vector for each block. The blocks that belong to the
same large object often have identical motion vectors and it is efficient to encode these vectors
together. The authors restrict to those video encryption algorithms that use a quadtree for merging
these blocks. Then, quadtree partial encryption is used to encrypt the motion vectors. Finally, for the
security purposes it is important to encrypt the residual error as well. Unencrypted residual error
may reveal the outline of a moving object. The residual error is often treated as an image frame and
then compressed using some standard image compression algorithm. Again, we restrict ourselves to
video compression algorithms that use either quadtree or wavelet based image compression
algorithm to compass the residual error frames. Thus, Partial encryption schemes for both quadtree
and wavelet compression can be applied to the residual error encryption. [21]
3. Introduction to Multimedia
Multimedia is media and content that uses a combination of different content forms. The term can
be used as a noun (a medium with multiple content forms) or as an adjective describing a medium as
having multiple content forms. The term is used in contrast to media which only use traditional
forms of printed or hand-produced material. Multimedia includes a combination of text, audio, still
images, animation, video, and interactivity content forms. Multimedia has become an inevitable part
of any presentation. It has found a variety of applications right from entertainment to education.
The evolution of internet has also increased the demand for multimedia content. Multimedia is the
media that uses multiple forms of information content and information processing (e.g. text, audio,
graphics, animation, video, interactivity) to inform or entertain the user. Multimedia also refers to the
use of electronic media to store and experience multimedia content. Multimedia is similar to
traditional mixed media in fine art, but with a broader scope. The term "rich media" is synonymous
for interactive multimedia.
Multimedia may be broadly divided into linear and non-linear categories. Linear active content
progresses without any navigation control for the viewer such as a cinema presentation. Non-linear
content offers user interactivity to control progress as used with a computer game or used in selfpaced computer based training. Non-linear content is also known as hypermedia content.
Multimedia presentations may be viewed in person on stage, projected, transmitted, or played locally
with a media player. A broadcast may be a live or recorded multimedia presentation. Broadcasts and
recordings can be either analog or digital electronic media technology. Digital online multimedia may
be downloaded or streamed. Streaming multimedia may be live or on-demand. Multimedia games
and simulations may be used in a physical environment with special effects, with multiple users in an
online network, or locally with an offline computer, game system, or simulator.
Multimedia Building Blocks
Any multimedia application consists any or all of the following components:
1. Text: Text and symbols are very important for communication in any medium. With the recent
explosion of the Internet and World Wide Web, text has become more the important than ever.
Web is HTML (Hypertext Markup language) originally designed to display simple text
documents on computer screens, with occasional graphic images.
2.
Audio: Sound is perhaps the most element of multimedia. It can provide the listening pleasure
of music, the startling accent of special effects or the ambience of a mood-setting background.
3.
Images: Images whether represented analog or digital plays a vital role in a multimedia. It is
expressed in the form of still picture, painting or a photograph taken through a digital camera.
4.
Video: Digital video has supplanted analog video as the method of choice for making video for
multimedia use. Video in multimedia are used to portray real time moving pictures in a
multimedia project.
3.1. Audio:
Sound is perhaps the most important element of multimedia. It is meaningful "speech" in any
language, from a whisper to a scream. It can provide the listening pleasure of music, the startling
accent of special effects or the ambience of a mood setting background. Sound is the terminology
used in the analog form, and the digitized form of sound is called as audio.
An audio file format is a file format for storing audio data on a computer system. It can be a raw bit
stream, but it is usually a container format or an audio data format with defined storage layer. The
general approach towards storing digital audio is to sample the audio voltage which, on playback,
would correspond to a certain level of signal in an individual channel with a certain resolution—the
number of bits per sample—in regular intervals (forming the sample rate). This data can then be
stored uncompressed, or compressed to reduce the file size. [15]
It is important to distinguish between a file format and a codec. A codec performs the encoding and
decoding of the raw audio data while the data itself is stored in a file with a specific audio file
format. Most of the publicly documented audio file formats can be created with one of two or more
encoders or codecs. Although most audio file formats support only one type of audio data (created
with an audio coder), a multimedia container format (as MKV or AVI) may support multiple types
of audio and video data.
3.2. Video
Video can be basically understood as a process of displaying still images at a rapid rate giving a
notion of a moving image, which is coupled with perfectly synchronized audio stream. Each such
still image is referred to as a frame.
Modern video file formats interleave audio and video to allow for playing even partially loaded video
stream on the network. And they also employ video compression methodologies to conserve space.
This compression is made possible by using relative references that is to say if two consecutive
frames have almost the same content except for partial changes, it is preferable to record only the
changes and use the reference frame to generate the current frame.
Thus we normally find the frames distinguished as,
I-frame (intra-coded frame) is an infra-coded pictures in effect a fully specified picture, like a
conventional static image file. I-frames are pictures coded without reference to any pictures except
themselves. They may be generated by an encoder to create a random access point (to allow a
decoder to start decoding properly from scratch at that picture location). They may also be
generated when differentiating image details prohibit generation of effective P or B frames. I-frames
typically require more bits to encode than other picture types. Often, I-frames are used for random
access and are used as references for the decoding of other pictures. Intra refresh periods of a halfsecond are common on such applications as digital television broadcast and DVD storage. Longer
refresh periods may be used in some environments. [18]
P-frame (Predicted frames) holds the changes in the image from the previous frame (Ex: Moving a car
across a stationary background, only car’s movement needs to be recorded).
P-frames require the prior decoding of some other picture(s) in order to be decoded. They
may contain both image data and motion vector displacements and combinations of the two. They
can reference previous pictures in decoding order.
In the older standard designs (such as MPEG-2), use only one previously-decoded picture as a
reference during decoding, and require that picture to also precede the P picture in display order. In
H.264, it can use multiple previously-decoded pictures as references during decoding, and can have
any arbitrary display-order relationship relative to the picture(s) used for its prediction. Typically, Pframes require fewer bits for encoding than I-frames do. [17]
B-frame (bi-directional predicted frame) helps specify the content by using differences between the
current and both preceding and following frames. [17]
B-frames require the prior decoding of some other picture(s) in order to be decoded. It may
contain both image data and motion vector displacements and combinations of the two. They
include some prediction modes that form a prediction of a motion region by averaging the
predictions obtained using two different previously-decoded reference regions.
In older standard designs (such as MPEG-2), B pictures are never used as references for the
prediction of other pictures. As a result, a lower quality encoding (resulting in the use of fewer bits
than would otherwise be the case) can be used for such B pictures because the loss of detail will not
harm the prediction quality for subsequent pictures. In H.264, they may or may not be used as
references for the decoding of other pictures.
In older standard designs (such as MPEG-2), use exactly two previously-decoded ^pictures as
references during decoding, and require one of those pictures to precede the B j picture in display
order and the other one to follow it. In H.264, can use one, two, or more than two previously-
decoded pictures as references during decoding, and can have any arbitrary display-order
relationship relative to the picture(s) used for its prediction. Typically, B-frames require fewer bits
for encoding than either I or P frames do. The following figure shows the relationship among the
various frame types mentioned above. [17]
Fig: 1 Group of frames [17]
3.3. Images
An image (from Latin imago) is an artifact, for example a two-dimensional picture that has a similar
appearance to some subject, usually a physical object or a person. Image file formats are
standardized means of organizing and storing digital images. Image files are composed of either
pixel or vector (geometric) data that are rasterized to pixels when displayed (with few exceptions) in
a vector graphic display. The pixels that constitute an image are ordered as a grid (columns and
rows); each pixel consists of numbers representing magnitudes of brightness and color.
4. Introduction to Secure Communication
4.1 Need of secure communication
The requirements of information security within an organization have undergone two major changes
in the last several decades. Before the wide spread use of data processing equipment, the security of
information felt of be valuable to an organization was provided primarily by physical and
administrative means.
With the advent of the computer, the need for automated tools for protecting files and other
information stored on the computer became evident. This is especially the case for a shared system.
4.2 Secure communication
When two entities are communicating with each other, and they do not want a third party to listen
to their communication, then they want to pass on their message in such a way that nobody else can
understand their message. This is known as communicating in a secure manner or secure
communication. Secure communication includes means by which people can share information with
varying degrees of certainty that third parties cannot know what was said. Other than
communication spoken face to face out of possibility of listening, it is probably safe to say that no
communication is guaranteed secure in this sense, although practical limitations such as legislation,
resources, technical issues (interception and encryption), and the sheer volume of communication
are limiting factors to surveillance.
4.3. How security is provided?
There are two major encryption methodologies used to provide secure communication:
Method I: Symmetric Encryption
Encryption methodologies that require the same secret key to encipher and decipher the message are
using what is called private key encryption or symmetric key encryption. Symmetric encryption
methods use mathematical operations that can be programmed into extremely fast computing
algorithms so that the encryption and decryption processes are executed quickly by even small
computers. The challenges that this type of encryption methodology faces is that if either copy of
the key falls into the wrong hands, massages can be decrypted by others and the sender and intended
receiver may not know the message was intercepted. The primary challenge of symmetric key
encryption is getting the key to the receiver, a process that must be conducted out of band to avoid
interception.
There are number of popular symmetric encryption cryptosystems. One of the most widely known
is the Data Encryption Standard (DBS). DBS uses a 64-bit block size and 56-bit key. But over the
years DBS has been proven to be easily compromised using a dedicated attack supported by proper
hardware. [22]
Method II: Asymmetric Encryption
Another category of encryption technique is asymmetric encryption. Whereas the symmetric
encryption system use a single key both encrypted and decrypted a message, asymmetric encryption
uses two different but related keys, and either key can be used to encrypt or decrypt the message. If,
however, key A is used to encrypt the message, only key B can decrypt it, and if key B is used to
encrypt the message, only key A can decrypt it. Asymmetric encryption can be used to provide
elegant solution to problem of security and verification. This technique has its highest value when
one key used as a private key, which means that it is kept secret (much like the key of the symmetric
encryption), known only to the owner of the key pair, and other key serves as a public key, which
means that it is stored in a public location where anyone can use it. This is why the more common
name for asymmetric encryption is public-key encryption.
One of the most popular asymmetric encryption techniques is the RSA algorithm. There are many
such algorithms that are widely in use in the present day. [22]
5. Structure of MP4
MP4 follows the ISO base format. The ISO Base Media File Format is designed to contain timed
media information for a presentation in a flexible, extensible format that facilitates interchange,
management, editing, and presentation of the media. This presentation may be 'local' to the system
containing the presentation, or may be via a network or other stream delivery mechanism. The file
structure is object-oriented; a file can be decomposed into constituent objects very simply, and the
structure of the objects inferred directly from their type. The file format is independent of any
particular network protocol while enabling efficient support for them in general. The ISO Base
Media File Format is a base format for media file formats. One such being MPEG-4 is technically
described by the ISO/IEC 14496-12, in accordance with its compliance to the base format. [6]
The responsibility of maintaining the ISO Base Media File Format rests on WG1 and WG11.
5 .1 Terms and definitions [17]
The following terms and definitions are useful in understanding the file format
Box: Object-oriented building block defined by a unique type identifier and length. It is also
popularly known as 'atom' in some specifications, including the first definition of MP4.
Chunk: Contiguous set of samples for one track
Container Box: whose sole purpose is to contain and group a set of related boxes.
Hint track: Special track which does not contain media data, but instead contains instructions for
packaging one or more tracks into a streaming channel
Hinter: tool that is run on a file containing only media, to add one or more hint tracks to the file
and so facilitate streaming
Movie box: Container box whose sub-boxes define the metadata for a presentation ('moov') media
data box, box which can hold the actual media data for a presentation ('mdat')
Sample: All the data associated with a single timestamp. No two samples within a track can share
the same time-stamp. In non-hint tracks, a sample is, for example, an individual frame of video, a
series of video frames in decoding order, or a compressed section of audio in decoding order; in hint
tracks, a sample defines the formation of one or more streaming packets.
Sample description: Structure which defines and describes the format of some number of samples
in a track
5.2 Object-structured File Organization
Files are formed as a series of objects, called boxes. All data is contained in boxes; there is no other
data within the file. This includes any initial signature required by the specific file format. All files
conformant to the ISO base format are required to contain the File Type Box.
Object Structure
An object in this terminology is a box. Boxes start with a header which gives both size and type. The
header permits compact or extended size (32or 64 bits) and compact or extended types (32 bits or
full Universal Unique Identifiers, i.e. UUIDs). The standard boxes all use compact types (32-bit) and
most boxes will use the compact (32-bit) size. Typically only the Media Data Box (es) need the 64bit size. The size is the entire size of the box, including the size and type header, fields, and all
contained boxes. This facilitates general parsing of the file. The definitions of boxes are given in the
syntax description language (SDL). The fields in the objects are stored with the most significant byte
first, commonly known as network byte order or big-endian format. When fields smaller than a byte
are defined, or fields span a byte boundary, the bits are assigned from the most significant bits in
each byte to the least significant. For example, a field of two bits followed by a field of six bits has
the two bits in the high order bits of the byte. [17]
aligned(S) class Box (unsigned int(32) boxtype,
optional unsigned int(8)[16] extended_type) {
unsigned int(32) size;
unsigned int(32) type = boxtype; if(size==l) {
unsigned int(64) largesize; } else if (size-- =0) { //box extends to end of file
} if (boxtype -='uuid') { unsigned int(8) [16] usertype = extended _type;
The semantics of these two fields are
Size is an integer that specifies the number of bytes in this box, including all its fields and contained
boxes; if size is 1 then the actual size is in the field largesize; if size is 0, then this box is the last one
in the file, and its contents extend to the end of the file (normally only used for a Media Data Box)
Type: identifies the box type; standard boxes use a compact type, which is normally four printable
characters (commonly known as the 4cc or 4 character code), to permit ease of identification, and is
shown so in the boxes below. User extensions use an extended type; in this case, the type field is set
to 'uid'.
Version: is an integer that specifies the version of this format of the box.
Flags: is a map of flags.
The semantics of these two fields are:
aligned(S) class FullBox(unsigned int(32) boxtype, unsigned int(8) v, bit(24)f) extends Box(boxtype)
{ unsigned int(8) version = v; bit(24) flags =f;
Many objects also contain a version number and flags field:
It suggested to ignore any boxes with an unrecognized type and is not to be considered for the
normal processing of the video.
Box Order
In order to improve interoperability and utility of the files, the ISO base format specifications
requires following of the following rules and guidelines for the order of boxes: [17]
1. The file type box 'ftyp' shall occur before any variable-length box (e.g. movie, free space,
mediadata). Only a fixed-size box such as a file signature, if required, may precede it.
2. It is strongly recommended that all header boxes be placed first in their container: these boxes
are the Movie Header, Track Header, Media Header, and the specific media headers inside the
Media Information Box (e.g. the Video Media Header).
3. Any Movie Fragment Boxes shall be in sequence order
4. It is recommended that the boxes within the Sample Table Box be in the following order: Sample
Description, Time to Sample, Sample to Chunk, Sample Size, Chunk Offset.
5. It is strongly recommended that the Track Reference Box and Edit List (if any) should precede
the Media Box, and the Handler Reference Box should precede the Media Information Box, and the
Data Information Box should precede the Sample Table Box.
6. It is recommended that user Data Boxes be placed last in their container, which is either the
Movie Box or Track Box.
7. It is recommended that the Movie Fragment Random Access Box, if present, be last in the file.
8. It is recommended that the progressive download information box be placed as early as possible
in files, for maximum utility.
The table shows those boxes that may occur at the top-level in the left-most column; indentation is
used to show possible containment. Thus, for example, a Track Header Box (tkhd) is found in a
Track Box (trak), which is found in a Movie Box (moov). Not all boxes are required to be present in
all the files; the mandatory boxes are marked with an asterisk (*), these mandatory boxes provide the
minimal information necessary for the normal processing and rendering of the file.
User data objects shall be placed only in Movie or Track Boxes, and objects using an extended type
may be placed in a wide variety of containers, not just the top level. [17]
6. Matrix Encryption Algorithm
This project will encrypt only the selected frames in multimedia, render the file playable.
Each selected frames are encrypted separately.
Analyze to find out
frames to be encrypted
Multimedia
Selected
Encrypt frames
frames
Remaining frames
Encrypted
multimedia
Transmit
Network
Transmit
Multimedia
Unencrypted
Analyze as to which frames
to be decrypted
frames
Frames to be decrypted
Decrypt frame
Decrypted frames
Fig 2: Figure 1 illustrates a simplified implementation of the algorithm
6.1. Matrix Encryption Algorithm
Figure 3 – Matrix Encryption Algorithm
1. Initial Round:
a. Divide the stream in frames to 16 bytes of sub-frames; consider it as a 4x4 matrix.
Each sub-frame is encrypted separately.
2. Rounds:
a. Operate on the part of matrix: Divide the matrix as 2x2 matrix (each cell will be
containing 2x2 matrix in it). Perform Xor operation on 1st part of the matrix with its
next one.
b. Operate with Key: Perform Xor operation on the 4x4 matrix with key of 16 bytes
(considering the key as 4x4 matrix for the operation)
c. Shift cells: shift the divided matrix cell values clockwise.
3. Final Round:
Append the encrypted sub-frames
The Algorithm is:
Step 1. Create matrix
Step 2. xor X00(00,01, 10, 11) with X01(02, 03, 12, 13) respectively which updates only 1/4th of
matrix.
Step 3. Rotate X00->X01->X11->X10->X01
Step 4. Add Key.
Step 5. Repeat step 2, 3, 4 for 3 more times. (Means all the 4 parts of matrix are updated).
Since the first step operates on itself (in all the 4 rounds, the same above operation works), we need
not to keep the s-Box like AES algorithm.
The second step is just like AES algorithm, it uses 16 bytes key, which will be XORed with the
resultant of the above step. (AES algorithm also uses 128 bit key, which mean 16 bytes).
The third step will actually shuffles the bytes in the matrix. Unlike AES, which will shifts values
within row. This algorithm will rotate the value position. The values changes row-wise as well as
column-wise.
6.2 Schema:
This project will provide an encryption methodology with the bit-wise encryption. This helps secure
multimedia data files in such a manner that securing only information in the frame data provides an
effect of securing the file as a whole. The Matrix encryption algorithm is based on Rijndael cipher.
This has 128 bit (16 bytes) of block size and 128 bit (16 bytes) of size. This uses 4 rounds that
convert the original block to encrypted block. This uses 4 rounds that convert the original block to
encrypted block. Each round consists of several processing steps, each containing three similar but
different stages, including one that depends on the encryption key itself The Matrix Encryption
Algorithm consists of a set of processing steps repeated for four rounds with 128 bit (16 bytes) of
key. The algorithm works on 16 bytes of data. We can divide the frame data in set of 16 bytes. Since,
each set is encrypted individually and independently. We can process each steps parallel, which will
increase the speed of encryption. The output of all blocks creates the encrypted video. A set of
reverse rounds are applied to transform ciphertext back into the original plaintext using the same
encryption key. Matrix encryption algorithm does not increase the size of the files, like other
encryption algorithms.
6.3 Encryption:
The frame divided into blocks before it is processed with Matrix Encryption. Each block of
data is encrypted separately and appended again to construct the encrypted frame.
{
public int[] EncryptFrame(int[] frame)
int length = frame.Count();
int[] encryptedFrame = new int[length];
int i = 0;
while (i < length)
{
int[] tempBlock = CeateBlock(frame, i);
if (tempBlock != null)
{
tempBlock = MatrixEncryption(tempBlock);
}
//Append the blocks to create encrypted frame
AppendArray(ref encryptedFrame, i, tempBlock);
i += 16;
}
}
return encryptedFrame;
The Matrix encryption is applied on block of code. The block of code will be in the form of an
array. Which will be first arranged as matrix for the matrix encryption.
The matrix is processed with 4 rounds of 3 operations.
After completion of operations, the encrypted matrix will be available. This encrypted matrix will be
converted back to array before appending with other blocks to construct encrypted frame.
private int[] MatrixEncryption(int[] block)
{
int[,] matrix = new int[4, 4];
//Create matrix of 4x4 with 16 bytes of subFrame
for (int row = 0, count = 0; row < 4; row++)
{
for (int col = 0; col < 4; col++)
{
matrix[row, col] = block[count];
count++;
}
}
for (int executionCount = 0; executionCount < 4; executionCount++)
{
matrix = InitialXorOperation(matrix);
matrix = KeyOperation(matrix, Key);
matrix = ShiftCells(matrix);
}
for (int row = 0, count = 0; row < 4; row++)
{
for (int col = 0; col < 4; col++)
{
block[count] = matrix[row, col];
count++;
}
}
}
return block;
The set of operations performed in each round in the Matrix encryption algorithm are as follows.
1. Initial XOR operation:
Quarter of matrix (cells: 00, 01, 10, 11) is performed XOR operation with next Quarter of
matrix (cells: 02, 03, 12, 13).
private int[,] InitialXorOperation(int[,] matrix)
{
for (int row = 0; row < 2; row++)
{
for (int col = 0; col < 2; col++)
{
matrix[row, col] ^= matrix[row + 2, col + 2];
}
}
return matrix;
}
2. Operation with Key:
The first quarter of the matrix is XORed with the key.
private int[,] KeyOperation(int[,] matrix, int[] key)
{
for (int row = 0, count = 0; row < 4; row++)
{
for (int col = 0; col < 4; col++)
{
matrix[row, col] ^= key[count];
count++;
}
}
return matrix;
}
3. Shift cells as to shuffle the data:
This step will actually shuffles the bytes in the matrix. Unlike AES, which will shifts values
within rows. This algorithm will rotate the value position. The values changes row-wise as well as
column-wise.
private int[,] ShiftCells(int[,] matrix)
{
int temp;
for (int row = 0; row < 2; row++)
{
for (int col = 0; col < 2; col++)
{
temp = matrix[row, col];
matrix[row, col] = matrix[row, col + 2];
matrix[row, col + 2] = matrix[row + 2, col + 2];
matrix[row + 2, col + 2] = matrix[row + 2, col];
matrix[row + 2, col] = temp;
}
}
return matrix;
}
6.4 Decryption
The decryption process follows the same process of dividing the frame in to blocks as to
process each block to convert it to original block. Each block are processed for decryption.
public int[] DecryptFrame(int[] frame)
{
int length = frame.Count();
int[] decryptedFrame = new int[length];
int i = 0;
while (i < length)
{
int[] tempSubFrame = CeateBlock(frame, i);
if (tempSubFrame != null)
{
tempSubFrame = MatrixDecryption(tempSubFrame);
}
AppendArray(ref decryptedFrame, i, tempSubFrame);
i += 16;
}
return decryptedFrame;
}
Similar to encryption process, The Matrix decryption is applied on block of code. The block of
code will be in the form of an array. Which will be first arranged as matrix for the matrix decryption.
The matrix is processed with 4 rounds of 3 operations.
After completion of operations, the encrypted matrix will be available. This decrypted matrix will be
converted back to array before appending with other blocks to construct decrypted frame.
private int[] MatrixDecryption(int[] subFrame)
{
int[,] matrix = new int[4, 4];
//Create matrix of 4x4 with 16 bytes of subFrame
for (int row = 0, count = 0; row < 4; row++)
{
for (int col = 0; col < 4; col++)
{
matrix[row, col] = subFrame[count];
count++;
}
}
for (int executionCount = 0; executionCount < 4; executionCount++)
{
matrix = ShiftCellsBack(matrix);
matrix = KeyOperation(matrix, Key);
matrix = InitialXorOperation(matrix);
}
}
for (int row = 0, count = 0; row < 4; row++)
{
for (int col = 0; col < 4; col++)
{
subFrame[count] = matrix[row, col];
count++;
}
}
return subFrame;
The decryption performs the same 3 operations as encryption but in reverse order.
7. Security Analysis
Secure Key: The main challenge was to secure the stored key. The storage also depends on the
hardware platform what we use for encryption and decryption. Use the ProtectedMemory class to
encrypt an array of in-memory bytes. This functionality is available in Microsoft Windows XP and
later operating systems. You can specify that memory encrypted by the current process can be
decrypted by the current process only, by all processes, or from the same user context. [8]
[HostProtectionAttribute(SecurityAction.LinkDemand, MayLeakOnAbort = true)]
public static class
ProtectedMemory
Use the ProtectedData class to encrypt a copy of an array of bytes. This functionality is available in
Microsoft Windows 2000 and later operating systems. You can specify that data encrypted by the
current user account can be decrypted only by the same user account, or you can specify that data
encrypted by the current user account can be decrypted by any account on the computer. [8]
[HostProtectionAttribute(SecurityAction.LinkDemand, MayLeakOnAbort = true)]
public static class ProtectedData
Other recent papers conducting analysis of standards and implementations of high-profile secure
protocols include [11], [12], [13], [14], and [17]. These papers, like ours, highlight the problems that
arise in protocol specifications with respect to implementation details having the potential to
undermine security. They also show that, in order to evaluate security, it is not enough to look at the
specification alone – rather, one must look at how the specifications have been implemented in
order to gauge whether an attack idea will work against a real system [18]
8. Implementation
The Matrix Encryption Algorithm was implemented for this project to validate the designed
algorithm, evaluate performance, and provide a tool that could be used for future applications. The
implementation is done with Windows application used for debugging, program validation and
testing. . The implementation was developed and complied on Windows 8 using the Microsoft
Visual Studio 2010 Express Edition compiler. The program is implemented in C# language.
Windows forms are used for User Interface for testing the algorithm, evaluating the performance.
The implementation includes reading a MP4 file and breaking the file into metadata, video frames
and audio frame. Encryption will be performed only on the frames, which is decided based on the
selection of encryption mode, video only, audio only or both. The implementation includes creation
of key file with random key, generated in the program. The implementation of encryption includes
breaking the frame into parts of 64 bits. Each part is then converted to matrix to apply Matrix
Encryption Algorithm on it. The encrypted matrix is again serialized to 64 bits to create the
encrypted MP4 file. The frames which are not required to encrypt are passed from original file to
encrypted file as it is. The person encrypts the video with the Matrix Encryption Algorithm and user
the server to upload along with the key. The recipient person could download with the password
and decrypts with the same Matrix Encryption Algorithm and it could play in any computer and on
any video players available.
9. Challenges Encountered
The implementation effort encountered several challenges in reading the MP4 file. Identifying the
frames type, size and data was a challenge. Encrypting the other data, which are not actually video or
audio data, would make the d output encrypted file non-playable. So it was very challenging to study
the file format and catch the exact frame data to perform the encryption operation.
Encryption of the media files not only mean just to make the video non-playable or to jumble the
data in frame to confuse the unauthorized viewers. But encryption also includes giving the security
from the hackers. Here with the Matrix Encryption Algorithm, we can achieve both jumbling the
data inside the frame and also secure the data with strong key. This will increase the strength of the
security as to provide high secured transmission over un-secured network. Encryption algorithm
should not increase the size of the file, which will add an overhead to the transmission. To avoid
size growth, Matrix Encryption Algorithm will perform rounds of operations within the matrix
along with key in one operation.
Chapter 10: Testing
Test results Diagrams
Chapter 11: Performance Analysis---> Fig
12. Conclusion
13. Future work:
There is broad perspective of future work related to this paper. As the technology grows, the recent
developments shows that implementation of the various encryption algorithms lead the attacks on
the algorithms like RSA. There should complex encryption techniques to make sure the public key
encryption is reliable. The paper could be extended to mobile devices which in turn based on the
cloud that requires dedicated servers and encryption key managers. The dedicated servers and
hardware’s can be used to make the encryption more specific to the hardware and the key
management can be easy.
The paper proposes the Matrix Encryption Algorithm; future work can present a proof of the
security for the algorithm and comply the algorithm on cloud and mobile devices like an application,
which should focus on the reliability of the algorithm.
14. References
Preliminary references relating to research done for this paper are listed below.
[1]
Qiao, Lintian, and Klara Nahrstedt. "A new algorithm for MPEG video encryption." Proc. of First
International Conference on Imaging Science System and Technology. 1997.
[2]
Liu, Fuwen, and Hartmut Koenig. "A survey of video encryption algorithms."computers &
security 29.1 (2010): 3-15.
[3]
Lui, Oi-Yan, Ching-Hung Yuen, and Kwok-Wo Wong. "Chaos-Based selective encryption for
AVS video coding standard." Advances in Multimedia Information Processing–PCM 2012.
Springer Berlin Heidelberg, 2012. 501-512.
[4]
Lim, Youngkwon, et al. "MMT: An Emerging MPEG Standard for Multimedia Delivery over
the Internet." MultiMedia, IEEE 20.1 (2013): 80-85.
[5]
QuickTime API Reference, 2005-08-11, Apple Computer, Inc. © 2005 Apple Computer, Inc.
[6]
http://en.wikipedia.org/wiki/Mp4
[7]
http://en.wikipedia.org/wiki/I-frame
[8]
http://wolfcrow.com/blog/what-is-video-encryption-and-should-you-care/
[9]
http://msdn.microsoft.com/en-us/library/ms229741%28v=vs.100%29.aspx
[10]
http://cs.uccs.edu/~gsc/pub/master/bbaker/doc/bbaker_project_slides.pdf
[11]
Bleichenbacher, Daniel. "Chosen ciphertext attacks against protocols based on the RSA
encryption standard PKCS# 1." Advances in Cryptology—CRYPTO'98. Springer Berlin Heidelberg,
1998.
[12]
Brumley, David, and Dan Boneh. "Remote timing attacks are practical."Computer Networks 48.5
(2005): 701-716.
[13]
Canvel, Brice, et al. "Password interception in a SSL/TLS channel." Advances in CryptologyCRYPTO 2003. Springer Berlin Heidelberg, 2003. 583-599.
[14]
Albrecht, Martin R., Kenneth G. Paterson, and G. Watson. "Plaintext recovery attacks against
SSH." Security and Privacy, 2009 30th IEEE Symposium on. IEEE, 2009.
[15]
Ravishankar, K. C., and M. G. Venkateshmurthy. "Region based selective image
encryption." Computing & Informatics, 2006. ICOCI'06. International Conference on. IEEE, 2006.
[16]
Multimedia Security Handbook by Borko Furht and Darko Kirovski, published by CRC Press
LLC in December 2004
[17]
http://cs.uccs.edu/~cs525/
[18]
Albrecht, Martin R., Kenneth G. Paterson, and G. Watson. "Plaintext recovery attacks against
SSH." Security and Privacy, 2009 30th IEEE Symposium on. IEEE, 2009.
[19]
Meyer, Jürgen, and Frank Gadegast. "Security mechanisms for multimedia data with the example
MPEG-1 video." Project Description of SECMPEG, Technical University of Berlin, Germany (1995).
[20]
Alattar, Adnan M., Ghassan I. Al-Regib, and Saud A. Al-Semari. "Improved selective encryption
techniques for secure transmission of MPEG video bit-streams." Image Processing, 1999. ICIP 99.
Proceedings. 1999 International Conference on. Vol. 4. IEEE, 1999.
[21]
Cheng, Howard, and Xiaobo Li. "Partial encryption of compressed images and videos." Signal
Processing, IEEE Transactions on 48.8 (2000): 2439-2451.
[22]
Simmons, Gustavus J. "Symmetric and asymmetric encryption." ACM Computing Surveys
(CSUR) 11.4 (1979): 305-330.