design and development of audio encryption algorithm with ... · everyday huge amount of personal,...

21
Design and Development of Audio Encryption Algorithm with High Energy Efficiency for Secure Communication Ph.D Synopsis Submitted to Gujarat Technological University For the Degree of Doctor of Philosophy In Computer Science By Rashmi A. Gandhi Enrollment No: 139997431002 (Computer Science) Supervisor: Co-Supervisor: Dr. Atul M. Gonsai Prof. Subramaniam Ganesan Professor Electrical & Computer Eng. Computer Science Department Director Saurashtra University Real Time Embedded DSP Lab Rajkot-Gujarat Oakland University India Rochester, USA

Upload: others

Post on 13-Oct-2020

13 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Design and Development of Audio Encryption Algorithm with ... · Everyday huge amount of personal, secret, and official information is exchanged over the network. Security of voice

Design and Development of Audio Encryption Algorithm

with High Energy Efficiency for Secure Communication

Ph.D Synopsis

Submitted to

Gujarat Technological University

For the Degree

of

Doctor of Philosophy

In

Computer Science

By

Rashmi A. Gandhi

Enrollment No: 139997431002

(Computer Science)

Supervisor: Co-Supervisor:

Dr. Atul M. Gonsai Prof. Subramaniam Ganesan

Professor Electrical & Computer Eng.

Computer Science Department Director

Saurashtra University Real Time Embedded DSP Lab

Rajkot-Gujarat Oakland University

India Rochester, USA

Page 2: Design and Development of Audio Encryption Algorithm with ... · Everyday huge amount of personal, secret, and official information is exchanged over the network. Security of voice

2

INDEX

Sr. No Topic Page

Abstract

1. State of the Art ………………………………………………………. 4

2. Problem Definition ………………………………………………….. 5

3. Objectives …………………………………………………………… 5

4. Scope of Work ………………………………………………………. 5

5. Original Contribution by the Thesis ………………………………… 6

6.

Experiments and Results …………………………………………….

6.1: Training Sets ………………………………………………………

6.2: Software Used ……………………………………………………..

6.3: Results ……………………………………………………………..

12

13

7. Contribution from Research Work …………………………………... 19

8. Conclusion …………………………………………………………… 19

9. List of Paper Published/Presented ……………………………….. 19

10. References …………………………………………………………… 20

Page 3: Design and Development of Audio Encryption Algorithm with ... · Everyday huge amount of personal, secret, and official information is exchanged over the network. Security of voice

3

ABSRTACT

Cryptography is the study of securing information. It is the physical process that scrambles the

information by rearrangement and substitution of content, so that it becomes difficult for anyone to

understand.

For cellular communication two competing network technologies are GSM and CDMA. The most

popularly and widely accepted technology is GSM. User Authentication and Security of data

communicated over the network is of prime concern. Currently communication is secured from the

Mobile Terminal (MT) to Base Station (BS), after that it is open on the network. GSM employs many

cryptographic algorithms like A5/1, A5/2 and A5/3 but all the algorithms are reverse engineered.

Different algorithms are developed to secure data but they are not fast as well as they are claimed to be

cracked. So there is need for an algorithm that improves speed, provides good security as well as easy to

implement.

Literature study reveals two approaches: one is to develop an add-on module outside the mobile device

that will encrypt/decrypt voice before it enters mobile device and the second one is to replace the

cryptographic algorithm A5. My research work is based on the second approach.

For security of data a large number of cryptographic algorithms are available with their respective

advantages and disadvantages. Cryptographic algorithms can be divided into Symmetric and Asymmetric

based on the number of key used, and as Block Cipher and Stream Cipher depending on whether the data

is processed in blocks or in a continuous fashion as a stream.

From the literature study and practical test bed implementation of symmetric block cipher algorithms,

Blowfish came out as the best performing algorithms. A new improved algorithm is developed based on

the Blowfish algorithm to improve speed and security.

Two different algorithms are proposed, one is for securing data from Mobile Station to Base Station and

other is to provide end to end security that is from sender side mobile station to receiver side mobile

station. In both the algorithms speed and security is of prime concern.

For MS to BS algorithm, the sender is the MS and receiver is the BS. The secure key Kc is already shared

between both of them. The proposed algorithm does not have to take care of key exchange. The key Kc is

generated using the A8 algorithm that is implemented in the MS. The same key Kc is also generated at

the BS. Before encryption a random number Salt is generated based on the system clock of the MS. The

salt with Kc will be input to the proposed algorithm’s key expansion part to create a number of sub-keys.

All the sub-keys are used during the 16 rounds of the data encryption. At the receiving end Salt is

communicated and the same process is reversed to get the original data back. The objective of the

proposed algorithm is to not only increase the speed of encryption/decryption but also improves security.

For end to end algorithm the sender is the sender side mobile station and receiver is the receiver side

mobile station. Communication will be encrypted from MS to MS. The intermediate BS of both the sides

has nothing to do. The exchange of secure key is done by Diffie_Hellman key exchange algorithm.

Encryption of data is done by the proposed modified Blowfish Algorithm.

Page 4: Design and Development of Audio Encryption Algorithm with ... · Everyday huge amount of personal, secret, and official information is exchanged over the network. Security of voice

4

1. State of Art of the Research Topic

In the present day digital world importance of networks, their effect and their presence can’t be ignored.

The widespread use of digital data in real life applications and their importance have craved the need of

new and effective ways to ensure their security. The quick development in computer technologies and

internet had made the security of information as most important factor in information technology and

communication.

Now a days Mobile phones and communication over mobile becomes an integral part of everybody’s life.

Global System for Mobile Communications (GSM) is the most popular and widely accepted digital

mobile communication system. GSM calls are exposed to be cracked by any third party on the network.

Everyday huge amount of personal, secret, and official information is exchanged over the network.

Security of voice over the network is of utmost demand.

To secure information numbers of techniques are available, like Cryptography, Steganography and

Digital Watermarking. Just a decade or two before peoples are used to only text data. So emphasis was

there only on how to encrypt/decrypt text data. But with the growing use of internet and multimedia data,

now there emerges the need for multimedia data security.

Audio encryption is more required to propagate encrypted voice communication between parties for real

time application like voice talk between intelligence bureau officials, CBI officials, defences etc. for top

secret communication. With the help of audio encryption other hackers or persons with malicious

intention will not be able to decrypt such communication for national security

Since in the current research work, security voice over GSM network is of concern we are discussing

different Audio Encryption Techniques. Audio encryption techniques can be divided into: Complete

encryption, Selective encryption and combined compression-encryption approach.

The Complete Encryption Approach encrypts the whole file with ciphers like DES, AES, 3DES, RC4, or

RSA. It leads to high processing and computational complexity. The Selective Encryption Approach

encrypts the parts of a multimedia file to reduce the computational requirements. Combined Compression

Encryption Approach combines the compression process and the encryption process in a single step.

With the improved technologies in terms of hardware and software selective encryption is not considered.

The work presented here demonstrates to design a new algorithm that will provide complete encryption,

less complexity, high energy efficiency in terms of CPU time and high encryption/decryption speed. It is

also considering reversible effective original audio generation based on speed and time with higher

security in original encrypted audio. So my work is focused on Design and implementation of a new

algorithm for audio encryption based on above parameters.

Page 5: Design and Development of Audio Encryption Algorithm with ... · Everyday huge amount of personal, secret, and official information is exchanged over the network. Security of voice

5

2. Problem Definition:

In the GSM network currently data is secured from the Mobile Terminal (MT) to Base Station (BS). After

that it is open on the network. A5/1, A5/2 and A5/3 are security algorithms for GSM. All the algorithms

are reverse engineered. Different algorithms are developed to secure data but they are not fast as well as

they are claimed to be cracked. Two implementations are done: 1. Algorithm that will secure

communication between MS and BS 2. Algorithm that will secure communication from Sender Side MS

to Receiver side MS.

3. Objectives of Research

1. To Study the security requirement of current GSM system and the need for improvement.

2. To find the security flaw in the A5 encryption algorithm.

3. To study and analyze different encryption algorithm already available AES, DES, 3DES, Blowfish

and RSA.

4. To study and improve encryption CPU time, memory utilization and total encryption time.

5. To design and develop an Audio encryption algorithm with reduced encryption/decryption time,

CPU load, CPU time and Memory utilization From Mobile Station to Base Station.

6. To design and develop an Audio encryption algorithm with reduced encryption/decryption time,

CPU load, CPU time and Memory utilization for end to end security.

7. To test performance analysis of newly developed algorithm for audio encryption.

8. To test the newly developed algorithm for Cryptographic Attack Testing.

4. Scope of Work:

1. Cryptographic algorithms help to secure confidential information. But less work is there for Audio file

encryption.

2. A5 algorithm working for GSM communication is reverse engineered and also it does not support for

End to end communication.

3. DES, AES and Blowfish algorithms are tested for securing Audio file.

4. Modified Blowfish algorithm is developed and tested for Speed and Security.

Page 6: Design and Development of Audio Encryption Algorithm with ... · Everyday huge amount of personal, secret, and official information is exchanged over the network. Security of voice

6

5. Original contribution By the Thesis:

For cellular communication two competing network technologies are GSM and CDMA. The most

popularly and widely accepted technology is GSM. User Authentication and Security of data

communicated over the network is compromised [1,2,3,4]. The current research work is targeted for

security of data. Security of the voice communicated is not assured, particularly over the core network.

Keeping the volume and importance of data communicated over the cell phone, end to end security is of

prime concern. Currently data is secured from the Mobile Terminal (MT) to Base Station (BS), after that

it is open on the network. GSM employs many cryptographic algorithms like A5/1, A5/2 and A5/3 but all

the algorithms are reverse engineered [8,9].

In GSM network the secret key is generated by A8 algorithm. The secret key Kc is input to the A5

encryption algorithm. A5 is a stream cipher. It operates on 228-bit blocks called “frames” sent and

received over the air every 4.6 milliseconds. GSM transmission is organized as a sequence of bursts. In a

typical channel and in one direction, one burst is sent every 4.615 ms and contains 114 bits available for

info. 114 bits represent data sent from the MSE and the other 114 bits are data received by the MSE, both

mainly containing digitized audio signals (after error correction). Taking the 64 bit session key Kc

produced by A8 and a frame counter Fn, A5 generates 228 pseudo random bits (PRAND) which are

XOR’ed with the plaintext frame resulting in 228 bits of ciphertext.

Page 7: Design and Development of Audio Encryption Algorithm with ... · Everyday huge amount of personal, secret, and official information is exchanged over the network. Security of voice

7

A5/1 and A5/2 use just a few LFSR; they are remarkably power-efficient. An A5/1 core can use less than

600 transistors, while a 3DES inefficient core would need 5 to 10 times as much; and AES even more

(AES was invented ten years later anyway). A5/1 still offers an extremely high security-to-power ratio.

Unfortunately, A5/1 was designed with too small an inner state, reflecting the traditions of the industry at

that time. The A5/1 structure offers, for a total internal state of n bits, a security level of about 22n/3

. This

means that the 64-bit space of A5/1 brings security to be about the same as a 42-bit block cipher, i.e. not a

lot... and quite a few researchers have describes "breaks" which apply various precomputed table methods

to this 42-bit strength problem.

Besides it speed researchers have confirmed security flaws in GSM network like Short encryption keys,

Flaws in the A5 structure as well as one way authentication. The said algorithm is reverse engineered. So

the focus is developing an algorithm which is secure and fast.

The current working of the GSM is demonstrated as shown below:

Figure 3: working of GSM Algorithms

For Securing GSM communication two approaches are suggested. One is to develop an add on module

[11-17] and the second one is replacing A5 algorithm [5,6,7]. The research work is targeted to develop

two algorithms: 1. Replacing A5 algorithm for MS to BTS communication. 2. Replacing A5 algorithm for

MS to MS, i.e. to provide end to end communication.

Page 8: Design and Development of Audio Encryption Algorithm with ... · Everyday huge amount of personal, secret, and official information is exchanged over the network. Security of voice

8

Cryptographic algorithm for Audio Encryption

Cryptographic algorithms can be divided into three independent dimensions

1. The type of operations used for transforming plaintext to ciphertext:

a) Substitution: Each element in the plaintext is mapped into another element.

b) Transposition: Elements in the plaintext are rearranged.

In both the schemes, the operation must be reversible that is no information should be lost.

2. The number of key used:

a) Symmetric: Single key, secret key, private or conventional encryption- Sender and receiver

use the same key.

b) Asymmetric: Two keys or public key encryption. Sender and receiver use the different key.

3. The way in which plaintext is processed:

a) Block Ciphers: Processes each input block one at a time, producing an output block.

b) Stream Ciphers: Processes input elements continuously, one element at a time.

To concentrate on our work of Audio Encryption, the literature work [20-27] is started on all types of

cryptographic algorithms. In our work media type as well as wired and wireless media is of great concern.

While analysing all the algorithms, factors like throughput, speed, and security are important but the need

of computing resources like CPU time, memory and battery power is of utmost concern. So while

considering symmetric or asymmetric encryption algorithm it can be observed that public key encryption

is based on mathematical functions, computationally intensive and is not very efficient for small mobile

devices. Asymmetric encryption techniques are almost 1000 times slower than Symmetric techniques due

to the high amount of computations.

Again with Symmetric ciphers, block ciphers and stream ciphers play important role. A block cipher

processes one block of data at a time while a stream cipher processes input elements continuously one

element at a time. While encrypting an offline file and sending it over networks block cipher will give

good result whereas encrypting real time data on a network in a continuous basis stream cipher will be a

better solution. But with present day computing resources Block Cipher can also be used for real time

communication. Kausumi is a block cipher employed in 3G GSM network. Hence an extensive literature

survey is done on Symmetric Block ciphers. The most common symmetric Block ciphers are: DES,

3DES, AES, Blowfish, RC4, Twofish and ThreeFish.

Table 1: Settings of Symmetric Encryption Algorithm

Algorithm Block Size Key Size Rounds Structure

DES 64 64 16 FIESTEL

3DES 192 64 16 FIESTEL

AES 128,192,256 128 10,12,14 NON- FIESTEL

BLOWFISH 64 32-448 16 FIESTEL

Page 9: Design and Development of Audio Encryption Algorithm with ... · Everyday huge amount of personal, secret, and official information is exchanged over the network. Security of voice

9

Table 2: Comparative Analysis on Different Parameters

Literature Study and initial testing demonstrated the superiority of Blowfish Algorithm as compared to

other Symmetric Algorithm. In the literature work [28-34] implementations of the Blowfish algorithm in

different platforms and situations is demonstrated. Researchers in [39,40,41] tested the Blowfish

Algorithm for Security aspects. So the work targeted is:

Figure 5: Working of MS to BTS Algorithm

Algorithm DES 3DES AES BLOWFISH

Energy

Consumption

Low Highest Medium Lowest

Execution Speed Slow Slowest Medium Fastest

Security Cracked Not cracked,

but very slow

Not Cracked Not Cracked

Encryption/

Decryption time

High Highest Moderate Lowest

Throughput Low Lowest High Highest

Figure 4: Replacing A5 with Modified

Blowfish

Page 10: Design and Development of Audio Encryption Algorithm with ... · Everyday huge amount of personal, secret, and official information is exchanged over the network. Security of voice

10

Working of the end to end Encryption Algorithm is as shown below:

Figure 6: Working of End to End Encryption Algorithm

Testing Parameters

Testing parameters for measuring the efficiency of Encryption Algorithm are:

Encryption/Decryption time

CPU time

CPU load

RAM Utilization

Testing parameters for measuring the Strength of Encryption Algorithm in terms of Cryptographic

Attack are:

• Avalanche Effect

• Correlation Coefficient

1. Average of Encryption/Decryption Time:

Average of encryption/decryption time is calculated by considering the time taken by the CPU for

encrypting the original audio file and decrypting the encrypted file to get back the original file. It only

Page 11: Design and Development of Audio Encryption Algorithm with ... · Everyday huge amount of personal, secret, and official information is exchanged over the network. Security of voice

11

counts the time consumed for encryption and decryption. Time taken for reading data from the file

and similarly time taken to write back the data in the output file is not considered.

2. CPU Time:

It is the time interval between the file submitted for encryption and getting the original file back after

decryption.

3. CPU Load:

CPU Load considers the percentage of CPU utilization.

4. RAM Utilization:

Amount of RAM needed to load the code and for the whole process of encryption and decryption is

calculated as RAM utilization.

Attack Testing:

Confusion Property is designed to hide the relationship between the Plaintext and Ciphertext. It will

discourage the attacker who attempts to locate the key using ciphertext.

Correlation Coefficient is used to determine the Confusion effect of the Block Cipher.

Diffusion is supposed to disseminate the plaintext statistics through the ciphertext. It will discourage

the attacker to attempting to locate the plaintext using the ciphertext statistics.

Avalanche effect is used to find out the Diffusion effect.

5. Avalanche Effect

Avalanche Effect=Number of changed Bits in Ciphertext

𝑁𝑢𝑚𝑏𝑒𝑟 𝑜𝑓 𝐵𝑖𝑡𝑠 𝑖𝑛 𝐶𝑖𝑝ℎ𝑒𝑟𝑡𝑒𝑥𝑡

A good cipher must have avalanche effect > 50%. One bit change in plaintext or key should produce a

significant change in at least half of the bits in the ciphertext. In other words a minimum change in the

input message should be amplified and produces a maximum change in the output message.

6. Correlation Coefficient

Page 12: Design and Development of Audio Encryption Algorithm with ... · Everyday huge amount of personal, secret, and official information is exchanged over the network. Security of voice

12

It deals with the dependency of how a single bit change in Plaintext ripples change is Ciphertext. It is

a measure of how the two variables affect each other. It is a measurement of the degree of the linear

relationship between two variables. It is a number between -1 and 1. Value 1 shows increasing linear

relationship. Value -1 shows decreasing linear relationship. Values 0 and ±0.3 indicates weak positive

or negative relationship. Values ±0.3 and ±0.7 indicates moderate positive or negative relationship.

Values ±0.7 and ±1 indicates strong positive or negative relationship

6. Experiments and Results:

6.1 Training set:

Proposed Algorithm is initially tested for Text file, Image file and Audio File. Since the prime concern is

for Audio File, the proposed algorithm is tested for different types of Audio file formats like .mp3, .aicc,

.aiff, .amr, .wav,.wma. For the experiment different size of audio files are used like 1.03 MB, 2.12 MB,

3.02 MB, 4.45 MB, and 7.64 MB etc. Proposed algorithm is tested for 5 different male and female

voices. The proposed algorithm is also tested for different block size and key size of 64 bits, 128 bits, 192

bits and 256 bits.

6.2 Software Used:

Proposed algorithm is implemented on Java Platform. An implementation Platform is developed in Java.

Coding is done for DES (CBC, CFB mode), AES (CBC, CFB mode) and Blowfish with all its 12 cases. A

Prototype of the system is implemented. A Personal computer (PC) is used with 2.40 GHz intel core i3

processor, Microsoft Windows 7 Operating System (64 bit) and 3.0 GB of RAM is used. Cryptographic

library functions from java ready cryptography library are called.

6.3 Results:

Result Analysis for DES, AES and Blowfish Original

Before implementation of the Proposed Modified Blowfish Algorithm, a test bed is created to find out the

best performing algorithm for encryption and decryption of Audio File. Data Encryption Standard (DES),

Advanced Encryption Standard (AES) and Blowfish Algorithm are tested for Text file, Image file and

Audio file. DES and AES are implemented in both CBC and CFB mode. Experimental results of Average

of Encryption/decryption time is shown in Table 3. Graphical representation of it is given in Figure 7. In

Table 4 results for encryption time and decryption time is demonstrated.

Page 13: Design and Development of Audio Encryption Algorithm with ... · Everyday huge amount of personal, secret, and official information is exchanged over the network. Security of voice

13

Table 3: Experimental Result for Average of Encryption and Decryption time for Text, Image and

Audio File in ms (Milli Seconds)

Algorithm Audio File-1.70 MB Image File-893 KB Text File-26.6 KB

DES: CBC Mode 176.5 155.5 6.4

DES: CFB Mode 172.5 149.5 7

AES: CBC Mode 112 118 9.5

AES: CFB Mode 100 97.5 5.65

Original Blowfish 67.4 53.5 0

Fig 7: Comparative Analysis of Average of Enc/Dec time for Text, image and Audio File

Table 4: Experimental Result for Encryption and Decryption time for Text, Image and Audio File

in ms (Milli Seconds)

Algorithm

Audio File-1.70 MB Image File-893 KB Text File-26.6 KB

Enc Time

Dec Time

Avg. Time

Enc Time

Dec Time

Avg. Time

Enc Time

Dec Time

Avg. Time

DES: CBC Mode 177 176 176.5 105 101 155.5 6.4 6.4 6.4

DES: CFB Mode 183 162 172.5 109 81 149.5 6 8 7

AES: CBC Mode 109 115 112 83 70 118 8 11 9.5

AES: CFB Mode 105 95 100 73 49 97.5 9.3 2 5.65

Original Blowfish 59.1 75.7 67.4 33 41 53.5 0 0 0

0

20

40

60

80

100

120

140

160

180

200

Audio File-1.70 MB Image File-893 KB Text File-26.6 KB

Aver

age

of

En

c/D

ec T

ime

File Types

Comparative Analsis of Average of Encryption and Decryption Time

DES: CBC Mode

DES: CFB Mode

AES: CBC Mode

AES: CFB Mode

Original Blowfish

Page 14: Design and Development of Audio Encryption Algorithm with ... · Everyday huge amount of personal, secret, and official information is exchanged over the network. Security of voice

14

Result Analysis for Blowfish and 12 cases for Advanced Parameters

Comparative analysis of Original Blowfish with the newly designed 12 cases is done on following

parameters: Encryption/Decryption Time, RAM utilization, CPU Load and CPU time.

An Audio file of Size 1.7 Mb and Type .mp3 is used for testing. Original Blowfish and the newly

designed 12 cases are tested for

– Block Size 64 bit and Key size 64 bit.

– Block Size 128 bit and Key size 128 bit.

– Block Size 192 bit and Key size 192 bit.

– Blowfish with Block Size 256 bit and Key size 256 bit.

Blowfish 64: Table 5: Experimental Result for Average of Encryption/Decryption time for Original

Blowfish and Proposed 12 cases in ms (Milli Seconds)

Blowfish Original + Proposed 12 Cases (Block Size 64 bit and Key size 64 bit)

Algorithm Average of Enc/Dec

time in ms CPU Time

in Sec CPU Load Output file Size in MB

Average RAM Utilized in MB

Blowfish Original 70 4.35 0.17 1.7 5.06

Case 1 80 3.73 0.16 1.7 5.16

Case 2 86 3.62 0.28 1.7 5.16

Case 3 65.3 3.60 0.20 1.7 5.16

Case 4 64.2 3.61 0.15 1.7 5.16

Case 5 56.8 3.65 0.14 1.7 5.16

Case 6 47.05 3.62 0.13 1.7 5.16

Case 7 52.6 3.58 0.16 1.7 5.16

Case 8 49.55 3.58 0.20 1.7 5.16

Case 9 50.35 3.57 0.20 1.7 5.16

Case 10 41.6 3.58 0.14 1.7 5.16

Case 11 48.4 3.53 0.14 1.7 5.16

Case 12 42.5 3.52 0.16 1.7 5.16

Figure 8: Comparative Analysis of Average of Encryption and Decryption time for Original Blowfish and

Proposed 12 cases for 64 bit

050

100

Blo

wis

h…

Cas

e 1

Cas

e 2

Cas

e 3

Cas

e 4

Cas

e 5

Cas

e 6

Cas

e 7

Cas

e 8

Cas

e 9

Cas

e1

0

Cas

e1

1

Cas

e1

2

Ave

rage

of

Encr

ypti

on

an

d

De

cryp

tio

n T

ime

Original Blowfish and Proposed 12 cases

Comparative Analysis of Average of Enc/Dec time in ms for 64 bit

Page 15: Design and Development of Audio Encryption Algorithm with ... · Everyday huge amount of personal, secret, and official information is exchanged over the network. Security of voice

15

Blowfish 128:

Table 6: Experimental Result for Average of Encryption and Decryption time for Original Blowfish

and Proposed 12 cases (128 bit) in ms (Milli Seconds)

Blowfish Original + Proposed 12 Cases (Block Size 128 bit and Key Size 128 bit)

Algorithm Average of Enc/Dec

time in ms CPU Time in

Sec CPU Load

Output file Size in MB

Average RAM Utilized in MB

Blowfish Original 39.9 2.19 0.19 1.7 3.36

Case 1 39.9 2.18 0.20 1.7 3.36

Case 2 48.0 2.14 0.16 1.7 3.36

Case 3 44.7 2.48 0.14 1.7 3.36

Case 4 41.4 2.45 0.13 1.7 3.36

Case 5 50.2 2.50 0.14 1.7 3.36

Case 6 58.0 2.52 0.18 1.7 3.36

Case 7 41.0 2.54 0.19 1.7 3.36

Case 8 60.6 2.53 0.13 1.7 3.36

Case 9 39.0 2.23 0.16 1.7 3.36

Case 10 32.1 2.20 0.15 1.7 3.36

Case 11 47.0 2.53 0.18 1.7 3.36

Case 12 63.5 2.52 0.15 1.7 3.36

Figure 9: Comparative Analysis of Average of Encryption and Decryption time for Original

Blowfish and Proposed 12 cases for 64 bit

0.010.020.030.040.050.060.070.0

Blowish…

Cas

e 1

Cas

e 2

Cas

e 3

Cas

e 4

Cas

e 5

Cas

e 6

Cas

e 7

Cas

e 8

Cas

e 9

Cas

e 1

0

Cas

e 1

1

Cas

e 1

2

Ave

rgae

of

Encr

ypti

on

an

d D

ecr

ypti

on

ti

me

in m

s

Blowfish Original and Proposed 12 cases

Comparative Analysis of Average of Enc/Dec time in ms for 128 bit

Page 16: Design and Development of Audio Encryption Algorithm with ... · Everyday huge amount of personal, secret, and official information is exchanged over the network. Security of voice

16

Blowfish 192

Table 7: Experimental Result for Average of Enc/Dec time for Original Blowfish and Proposed 12

cases (192 bit) in ms (Milli Seconds)

Blowfish Original + Proposed 12 Cases (Block size 192 bit and key size 192 bits)

Algorithm Average of Enc/Dec

time in ms CPU Time in

Sec CPU Load Output file Size in MB

Average RAM Utilized in MB

Blowfish Original 41.0 2.60 0.15 1.7 2.76

Case 1 33.1 1.93 0.21 1.7 2.76

Case 2 37.3 1.84 0.12 1.7 2.76

Case 3 48.7 1.79 0.13 1.7 2.76

Case 4 45.3 1.84 0.08 1.7 2.76

Case 5 52.8 1.90 0.13 1.7 2.76

Case 6 29.5 1.84 0.00 1.7 2.76

Case 7 39.1 1.87 0.10 1.7 2.76

Case 8 45.2 1.84 0.15 1.7 2.76

Case 9 31.3 1.84 0.18 1.7 2.76

Case 10 26.3 1.87 0.13 1.7 2.76

Case 11 35.6 2.75 0.15 1.7 2.76

Case 12 51.2 2.71 0.13 1.7 2.76

Figure 10: Comparative Analysis of Average of Encryption and Decryption time for Original

Blowfish and Proposed 12 cases for 64 bit

0.0

10.0

20.0

30.0

40.0

50.0

60.0

Ave

rage

of

Encr

ypti

on

an

d D

ecr

ypti

on

tim

e in

ms

Blowfish Original and Proposed 12 cases

Comparative Analysis of Average of Enc/Dec Time in ms for 192 bit

Page 17: Design and Development of Audio Encryption Algorithm with ... · Everyday huge amount of personal, secret, and official information is exchanged over the network. Security of voice

17

Blowfish 256

Table 8: Experimental Result for Average of Enc/Dec time for Original Blowfish and Proposed 12

cases (256 bit) in ms (Milli Seconds)

Blowfish Original + Proposed 12 Cases (Block Size 256 bits and Key size 256 bits)

Algorithm Average of Enc/Dec

time in ms CPU Time in

Sec CPU Load Output file Size in MB

Average RAM Utilized in MB

Blowfish Original 47.3 1.41 0.00 1.7 2.40

Case 1 33.7 1.87 0.16 1.7 2.40

Case 2 48.2 2.01 0.21 1.7 2.40

Case 3 50.3 2.12 0.16 1.7 2.40

Case 4 43.4 2.16 0.13 1.7 2.40

Case 5 44.0 2.17 0.13 1.7 2.40

Case 6 59.1 2.17 0.14 1.7 2.40

Case 7 47.1 1.42 0.00 1.7 2.40

Case 8 52.7 1.44 0.00 1.7 2.40

Case 9 30.9 1.42 0.10 1.7 1.20

Case 10 45.4 1.42 0.00 1.7 2.40

Case 11 50.1 1.42 0.00 1.7 2.40

Case 12 48.4 1.41 0.09 1.7 2.40

Figure 11: Comparative Analysis of Average of Encryption and Decryption time for Original

Blowfish and Proposed 12 cases for 64 bit

0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

Blo

wis

h O

rigi

nal

Cas

e 1

Cas

e 2

Cas

e 3

Cas

e 4

Cas

e 5

Cas

e 6

Cas

e 7

Cas

e 8

Cas

e 9

Cas

e 1

0

Cas

e 1

1

Cas

e 1

2

Ave

rage

of

Encr

ypti

on

an

d D

ecr

ypti

on

tim

e in

ms

Blowfish Original and Proposed 12 cases

Comaprative Analysis of Average of Enc/Dec Time in ms for 256 bit

Page 18: Design and Development of Audio Encryption Algorithm with ... · Everyday huge amount of personal, secret, and official information is exchanged over the network. Security of voice

18

Table 9: Comparative Summary for Original Blowfish and Proposed 12 cases for Block size and

Key size of 64, 128, 192 and 256 bits

Comparative Summary

Average of

Enc/Dec time in

ms

CPU Time in

Sec

CPU

Load

Output file

Size in MB

Average RAM

Utilized in MB

Blowfish Original + 12 Cases 64 bit

min 41.6 3.516 0.1265 1.7 5.06

case 10 case 12 case 6 Same Org

Blowfish Original + 12 Cases 128 bit

min 32.1 2.141 0.13 1.7 3.36

case 10 case 2 case 4 Same Same

Blowfish Original + 12 Cases 192 bit

min 26.3 1.792 0 1.7 2.76

case 10 case 3 case 6 Same Same

Blowfish Original + 12 Cases 256 bit

min 30.9 1.406 0 1.7 2.40

Case 9

The proposed Algorithm will consider a block size of 128 bit and key size of 128 bit. Form the newly

designed 12 cases, for 128 bit block size and key size, proposed case 10 is giving the best performance.

So for further analysis and comparisons I will compare original blowfish with proposed case 10 named as

Modified Blowfish Algorithm.

Blowfish Algorithm and Proposed Modified Blowfish Algorithm have been tested for:

1. Different File Formats (.aac, .aiff, .wma, .wav, .mp3, .amr)

2. Different Male and Female Voice Types.

3. Audio file of different size

The Proposed Algorithm is also compared with A5 algorithm. It is tested for Cryptographic Attacks.

5.7 Comparative analysis of Blowfish with A5

Page 19: Design and Development of Audio Encryption Algorithm with ... · Everyday huge amount of personal, secret, and official information is exchanged over the network. Security of voice

19

7. Contribution from Research Work:

1. The Implemented algorithm is performing better as compared to the existing cryptographic

algorithms in terms of Average of Encryption and Decryption time for Audio File.

2. The Implemented algorithm is also tested against the existing A5 algorithm used in GSM and

performance is very good.

3. Security of the Algorithm is tested.

4. It can be used to replace the existing A5 algorithm in the GSM.

5. It can be adopted by the Mobile Phone developers and Network operators for secure

communication.

8. Conclusion:

The proposed work improves the speed and security of the data communicated using Modified

Blowfish Algorithm. It can be used to replace the existing A5 algorithm in the mobile devices without

doing any changes in the network scenario. It can also be applied for end to end security. In the end to

end security scenario key exchange is done by the Diffie Hellman Key exchange algorithm. The

proposed work could be tested for the live network scenario.

9. List of Paper Published/Presented:

1. “A Study on Current Scenario of Audio Encryption” – International Journal of Computer

Applications-Volume 116-No 7, April 2015. ISSN: 0975-8887.

2. “Audio encryption with AES and Blowfish”-International Journal for Research in Applied

Science and Engineering Technology-Volume 4- Issue XI, November 2016. ISSN: 2321-9653.

3. “Design and Development of Improved Blowfish for Encryption and Decryption of Audio

File” in International Journal of Research in Advent Technology, Special Issue, (ICSACST-

2019) at Christ University, Bengaluru E-ISSN: 2321-9637. March 2019.

4. “Security of Voice over GSM Network using Modified Blowfish Algorithm” Special Issue

on ”Computational Algorithms for different wireless and optical technology standards"

for International Journal of Sensors, Wireless Communications and Control (Bentham

Science- Scopus Indexed) August 2019.

Page 20: Design and Development of Audio Encryption Algorithm with ... · Everyday huge amount of personal, secret, and official information is exchanged over the network. Security of voice

20

10. References: 1. Constantin Daniel Oancea, Member, IEEE, “GSM Infrastructure Used for Data Transmission”, 2011 the 7

th

International Symposium on Advanced Topics in Electrical Engineering, Bucharest,May,12-

14,2011.

2. Petr Bouška, Martin Drahanský, “Communication Security in GSM Networks”, 2008 International

Conference on Security Technology, pp. 248-251, 2008.

3. C. Lo and Y. Chen, “Secure communication mechanisms for gsm networks,” Consumer Electronics, IEEE

Transactions on, vol. 45, no. 4, pp. 1074–1080, 1999.

4. Mohsen Tooran and Ali Asghar Beheshti Shirazi, “Solutions to the GSM Security Weaknesses”, The

Second International Conference on Next Generation Mobile Applications, Services, and

Technologies(NGMAST'08), pp.576-581, University of Glamorgan, Cardiff, UK, Sep. 2008.

5. Khaled Merit and Abdelazziz Ouamri, “Securing Speech in GSM Networks using DES with Random

Permutation and Inversion Algorithm”, International Journal of Distributed and Parallel Systems (IJDPS)

Vol.3, No.4, July 2012.

6. Daya Sagar Gupta, G. P. Biswas, “Securing Voice Call Transmission over Cellular Communication”, 3rd

International Conference on Recent Trends in Computing 2015 (ICRTC-2015), Elsevier, Science direct.

7. E.Vrentzos, G.Kostopoulos, O.Koufopaiou, “Hardware Implementation of The A5/3 and A5/4 GSM

encryption algorithms”

8. Oliver Damsgaard Jensen, Kristoffer Alvern Anderson, “A5 Encryption in GSM”, 2017

9. Vadim Bulavintesv, Alexander Semenov, Oleg zaikin, Stephen Kochemazov,“A Bitslice implementation of

the Anderson’s Attack on A5/1” Research Article for Open Eng. 2018

10. Neetesh Saxena, Narendra S. Chaudhuri, “NPA: Protocol for Secure Communication in GSM Cellular

Network”, 10th Annual IEEE CCNC – Wireless Communication Track-2013.

11. Gianluigi Biancucci, Andrea Claudi, Aldo Franco Dragoni, “Secure Data and Voice Transmission Over

GSM Voice Channel: Applications for Secure Communications”, 4th International Conference on

Intelligent Systems, Modelling and Simulation, pp. 230-233, 2013.

12. Aruna Chouhan, Sukriti Singh, “Real Time Secure End To End Communication over GSM Network”,

2015 International Conference on Energy Systems and Applications (ICESA 2015), Pune, India 30 Oct - 01

Nov, 2015,IEEE.

13. Lin Shaofeng, Guo Chaoping, Ni Lin, Kou Wanli, Zheng Minjiao, “The Research of Encryption Algorithm

for Voice Communication of Mobile Station”, International Conference in Intelligent Transportation, Big

Data and Smart City, IEEE 2016.

14. N.N. Katugampala, K.T. Al-Naimi, S. Villette, and A.M. Kondoz, “REAL-TIME END-TO-END SECURE

VOICE COMMUNICATIONS OVER GSM VOICE CHANNEL”, 13th European Signal Processing

Conference (EUSIPCO’05), Turkey, Sep-2005.

15. Mehmet Akif Ozkan, Berna Ors, Gokay Saldamli, “Secure Voice Communication via GSM Network”,

ELECO 2011 7th International Conference on Electrical and Electronics Engineering, 1-4 December,

Bursa, TURKEY.

16. Vikas Kaul, Dr. V. A. Bharadi, P. Choudhari, Dhvani Shah, Dr. S. K. Narayankhedkar, “Security

Enhancement for Data Transmission in 3G/4G Networks”, 2015 International Conference on Computing

Communication Control and Automation, IEEE.

17. Heshem A. El Zouka, “PROVIDING END-TO-END SECURE COMMUNICATIONSIN GSM

NETWORKS”, International Journal of Network Security & Its Applications (IJNSA) Vol.7, No.4, July

2015.

18. Yohan Suryanto1, Kalamullah Ramli, “ Implementation and Performance Analysis of Reliable and Secure

End to End Voice Encryption over Public Mobile Network Based on Frequency Domain Using Dual

Processor in FPGA Platform”, Yohan Suryanto et al , Int.J.Computer Technology & Applications,Vol 5

(1),103-111, ISSN:2229-6093,IJCTA,2014.

19. Saad Islam, Fatima Ajmal, Salman Ali, Jawad Zahid and Adnan Rashdi, “Secure End-to-End

Communication over GSM and PSTN Networks”, IEEE,2009.

Page 21: Design and Development of Audio Encryption Algorithm with ... · Everyday huge amount of personal, secret, and official information is exchanged over the network. Security of voice

21

20. Walid Y. Zibideh, Mustafa M. Matalgah, “Energy Consumption Analysis for a class of Symmetric

Encryption Algorithm”, IEEE 2014.

21. Aamer Nadeem, Dr M. Younus Javed, “A Performance Comparison of Data Encryption Algorithms”,

IEEE, 2005.

22. Sourabh Chandra, Siddhartha Bhattacharyya, Smita Paira, Sk Safikul Alam, “A Study and Analysis on

Symmetric Cryptography”, IEEE, 2014.

23. Bijoy Kumar Mandal, Debnath Bhattacharyya, Samir Kumar Bandyopadhyay, “Designing and

Performance Analysis of a Proposed Symmetric Cryptography Algorithm”, IEEE,2013.

24. Rashmi A. Gandhi, Dr. Atul M. Gosai “A Study on Current Scenario of Audio Encryption ”, International

Journal of Computer Applications, Volume 116, No.7, April 2015.

25. Diaa Salama, Hatem Abdual Kader, and Mohiy Hadhoud, “Evaluating the effects of Symmetric Encryption

Algorithms on Power Consumption for different Data Types”, International Journal of Network Security,

Vol.11, No.2, Page.78-87, Sep 2010

26. O P Verma, Ritu Agarwal, Dhiraj Dafouti, Shobha Tyagi, “Performance Analysis Of Data Encryption

Algorithms”, IEEE, 2011.

27. Tingyuan Nie, Chuanwang Song, Xulong Zhi, “Performance Evaluation of DES and Blowfish

Algorithms”, IEEE,2010.

28. Chaitali Haldankar, Sonia Kuwelkar, “ IMPLEMENTATION OF AES AND BLOWFISH ALGORITHM

“, International Journal of Research in Engineering and Technology , Volume:03, Issue:03, May 2014.

29. Ramesh .A, Suruliandi.A “ Performance Analysis of Encryption for information security”, IEEE 2013

30. S. Pavithra, E.Ramadevi, “Throughput Analysis of Symmetric Algorithms”, International Journal of

Advanced Networking and Applications, Volume-4, Issue-2, Pages:1574-1577, 2012.

31. Vaibhav poonia and Dr. Narendra Singh Yadav, “Analysis of the modified blowfish Algorithm in different

cases with various parameters” , International Conference on Advanced Computing and communication

Systems, Jan 05-07,2015, IEEE2015.

32. Brian Cody, Justin Madigan, Spencer MacDonald, Kenneth W. Hsu, “High Speed SOC Design for

Blowfish Cryptographic Algorithm”, IEEE, 2007.

33. Russell K. Meyers and Ahmed H. Desoky, “An Implementation of the Blowfish Cryptosystem”, IEEE,

2008.

34. Swagata Roy Chatterjee, Soham Majumder, Bodhisatta Pramanik, Mohuya Chakraborty, “FPGA

Implementation of Pipelined Blowfish Algorithm”, Fifth International Symposium on Electronic System

Design, IEEE, 2014.

35. Khelf Mohamed, Ouslim Mohamed, Hamoudi.M, Masmoudi.M, “QoS evaluation in VoIP software with

and without Blowfish encryption module”, IEEE.

36. Amaal A. Abd El-Sadek, Talaat A. El-Garf, Mohammed M. Fouad, “Speech Encryption Applying a

Modified Blowfish Algorithm”, IEEE, 2014.

37. Spencer Davis, Brandon Jones, Hai Jiang, « Portable Parallelized Blowfish Via RenderScript”, IEEE, 2015.

38. Kuraniawan Nur Prestyo ST., Yudha Purwanto, ST., MT., Denny Darlis, S.Si, MT., “An Implementation of

Data Encryption for Internet of Things Using Blowfish Algorithm on FPGA”, 2nd

International Conference

on Information and Communication Technology, IEEE, 2014.

39. Jiali Bian, Bei Lu, Jian Kuang, ” A New Hierarchical File Encryption System Based On Smartphone”, 2nd

International Conference on Computer Science and Network Technology, IEEE, 2012.

40. Ashwak Alabaichi, Faudziah Ahmad, Ramlan Mahmod , Mohmood S. Mechee, “Randomness Analysis on

Blowfish Block Cipher using ECB and CBC Mode”, Journal of Applied Sciences, Asian Network for

Scientific Information, 2013.

41. Ashwak Alabaichi, Faudziah Ahmad, Ramlan Mahmod , “ Securtiy Analysis of Blowfish Algorithm”,

IEEE, 2013.