The Proposed Development of Prototype with Secret Messages Model in Whatsapp Chat

ABSTRACT

. The authentication flowchart [8] Figure 2. Abstracted chat client user interface, including examples of user, system, and action messages [5] Figure 2 shows a chatting process in a group in a textual application. This plaintext is likely dangerous to be attacked by outsider trying to achieve those information through smartphone. In WA is already using end-to-end encryption, but we need to provide additional system outside the WA system, so users feel confident that the messages sent are not be tapped by the attackers. Next, our research focuses on how to disguise messages sent through chatting application in smartphone system used by many users nowadays. The plaintext sent will be encrypted, thus the text saved in the operator server will not be easily re-encrypted by the attackers. The text saved can be textual [17] or images [18]. Development of a model of secret messages through WA voice as in the crypt analysis of two pads in case of compressed speech [19].

RELATED WORK
In the development of the proposed model there are several related studies directly or indirectly, so some prior research is needed to produce a prototype suitable for the security needs of the next secret message. In [20] developing message authentication with a correlated setup for sending messages as done by Bob and Alice. Meanwhile, [21] analyzed the recently-proposed block cipher using cellular automata (CA) and the self-invertible function. This paper performs insecurity analysis of the cipher, because of its conjugate nature.
The results of this study also build the decryption process without knowledge of the secret key so it needs to be fixed. In [8] developing a secret-key cipher based on a non-linear structured code with the distinguishing feature that its error generator is implemented by a non-linear combination of code words of two linear complementary codes. In [2] developing a practical reputation system for pervasive social chatting with Mobile Ad Hoc Network (MANET) on smartphone. In this study [22], a general attack scenario was given in order to conduct security analyses of chaos based crypto systems with proposed cryptanalysis of a new images so it can be used in the MANET.
In the proposed algorithm, the secret key detected between sender and receiver was used in the calculation of system parameters of chaotic map group. In [1] perform a social graph based text mining framework for chat log investigation for presents a unified social graph based text mining framework. Further, it needs to be made for secure group-based mobile chat protocol [23]. For example, huge mobile The effect of coaching and feedback on online chats in this case also found in the study [25]. Coaching occurred before each chat, and feedback was provided immediately afterwards. The findings suggest that over time, the frequency of higher-order thinking will increase more in a group that receives coaching and feedback than in a group that does not receive coaching and feedback. In addition, the findings suggest that the Community of Inquiry framework has benefits beyond its use in course design, facilitation, and assessment to include serving as a guide to coaching [25]. New steganography algorithm to conceal a large amount of secret message using hybrid adaptive neural networks with modified adaptive genetic algorithm [26]. Meanwhile, a high capacity text steganography scheme based on LZW compression and color coding [27].
The steganographic system thus embeds secret content in cover media (like text, image, audio, and video) so that its existence is not detected to an eavesdropper. Impostor Detection Through Chat Analysis also done by [3] for verify the identity of the person on the basis of his writing style. In this case we also need a model that can integrate the system in sending text. Some other research results such as in [14] and [23].
In [28] an encryption algorithm with an error detection technique and technique has been successfully verified and synthesized using Xilinx by Spartan-3E FPGA. [14] present a scheme which allows an arbitrary 2-qubit quantum state teleportation between two remote parties with control of many agents in a network, while on [23] using Mobile Chat (MC) and Mobile Chat System (MCS) for services plays a very important role in current social networking. Thus, both the Secure End-to-End Secure Mobile Chat (SE2E-MC) scheme and Secure Group-based Mobile Chat (SG-MC) scheme are provided the suitable solutions. The design and implementation of secure chatting application with end to end encryption [29] and combining steganography and cryptography on android platform to achive a high-level security [30].

PROPOSED METHOD
Data security model is needed to encrypt client message before sending through the provider service, thus the message has been encrypted when it is received by the server. Plaintext and plainimage sent to the other users will not be easily read with application such as WA before decrypted to plaintext. In this method, we prepare some material consisting of cryptography model such as vigenere cipher and transposition to encrypt the message sent and to smartphone as sender and receiver.
In this case, we try to do encryption by using vigenere model, transposition and web service for testing whether the messages we sent can be encrypted in the smartphone system. Meanwhile, we use web service for integrating data from client in the smartphone so that the message written in the client application can be accepted by WA chat system. This method is proposed to secure textual message and images sent through chatting system hence the users of chatting application feel more secured. The steps of proposed model in developing prototype with secret message model as shown in Figure 3 we are working on is used for message sending process written by sender and then received by others. In Figure 3 we can see the process in (a) private area and (b) public area in a proposed method consisting of sender (1) as the secret message sender. Sender (1) fill in the textual data or images as plaintext which is then encrypted before sending so that the message sent to the smartphone is in the form of ciphertext. The message is then sent to the providers (2) plaintext as the origin, (3) step for message encryption process (MEP) then sent to the chatting system in the smart phone (4) receive encrypted message to send to receivers through providers. (5) ISP only receive message than cannot be read the originality and then sent to the receivers as the destination (6).
The receiver (6) of encrypted message then input the encrypted message into cryptosystem in the smartphone to be encrypted (7) by message decryption process (MDP) based on private key of the sender key, thus the origin text as the plaintext (8) will be returned in the receiver system (9). Receiver (9) receives the message in the form of ciphertext which is next will be decrypted to see the original message. It can be seen that the message sent from the sender to the receiver through provider is a cryptography method encrypted message. In the proposed method we are working on, there are some methods in building prototype of secret message in chatting. They are:

Message encryption and decryption with vigenere cipher model
Vigenere Cipher is a polyalphabetic substitution cipher algorithm. It means, for every similar letter in a plaintext is not substituted by a letter but another letter based on the key used to encrypt. Vigenere Cipher is the simple form of polyalphabetic substitution code [31]. Vigenere Cipher is very well-known because it is very easy to understand and implement [32]. Vigenere Cipher can be seen on Figure 4. To perform Vigenere cipher encryption, draw a downward vertical line from the plaintext in the vigenere square. Then draw an across line from the key letter to the right. The letter at the intersection shows the ciphertext letter with addition [31]. Mathematically, the encryption process is expressed as follow (1): Decryption is done oppositely by drawing across line from key letter to the chipertext letter, then draw a vertical line from ciphertext letter to the plaintext letter with subtraction [31]. The decryption process can be written mathematically as follow (2)

Message encryption and decryption with transposition model
Transposition model is basically making ciphertext by changing plaintext object-object position without shifting the plaintext object, thus in this technique another character is not needed [33]. Reading the matrix value column by column based on the key used is done to get the ciphertext by using this transposition technique [34]. Chiper example of character transposition technique from "Universitas Mulawarman Samarinda" plaintext and using of frame 9x4 is shown by Figure 5. Plaintext is arranged to the right then downward by "4 3 1 5 2 6" key, so the cipher result will follow the key downward as "ITANI ESAAD NILAR USUMA VAWSN RMRMA". Encryption for data disguising by using key is known as encryption process. The key used is alphanumerical (a-z, A-Z, 0-9). Figure 6 shows the encryption and decryption process as shown in Figure 7 at frame 9x4 pixel.  Figure 6 shows a process of encryption and decryption horizontally, meanwhile Figure 7 shows the process vertically. This model is used to jumble the number of pixel in the images so it cannot be easily recognized when it is encrypted and can be recognized when it is decrypted.

Integration with web service model
In this step we carry out data integration by using web service to achieve secret message through sender and receiver in which architecture is oriented in doing service producing different solution in business logic with various application [15].Web service technology combines the message sent through WhatsApp (WA) web to be disguised. The model of architecture in the web service is shown in Figure 8

Autentification model
An authentication system with a correlated setup is a pair of algorithms (MAC, VER) for a source space Ѕ, a message space M and a distribution over × × , specified as follows [20]. a. Setup (1 ). Initially, is taken from × × according to Alice receives , Bob receives and Oscar receives . b. Authentication MAC( , Ѕ). Upon S ∈ S, Alice uses her secret key to run algorithm MAC to generate a message M= MAC ( , S). c. Verification VER( , M). Upon M ∈ M, Bob uses his secret key to run algorithm VER to compute − =VER(Yn, M) ∈ S ∪{⊥}, where − =⊥ means that Bob accepts − as an authenticated source state while − =⊥ means a reject.

Security model
An authentication system (MAC, VER) with a correlated setup for source space S. distribution × × is t-order secure if the following holds. If user attack doesn't attack, then Pr( − =S|S=S) is × × is t-order secure if the following holds [20]. a. Authentication attack has only a negligible success probability in nin the following attack: b. After seeing messages M1, ···, M (≤t) from sender to receiver, attacker constructs a new message M * such that VER (Yn, M * ) is distinct from ⊥, S1, ···, S, where Siis the source state in .

Implementation result
We recognize that the WA system already has end to end encryption. However, we need to propose other systems outside the system to give the WA users more confidence in the messages they send. Our research focuses on encrypting and decrypting text and images in the secret message through chatting to be applied in WA application and has successfully tested in android system application, thus the text will be first encrypted before sending to the WA application. The implementation of the secret message is shown in Figure 9. Figure 9 shows our test results for sending secret messages into private and group chats. The result of encryption process of sentence THIS PLAINTEXT is done by using key sony sonysonys, i.e. Text for T is encrypted to key S and so on, as the result the step of encryption can be seen as follows:

Discussion result
There are some weaknesses we need to improve in the model we are working on. For instance, how to not separate secret message system and WA so that the model developed does not need to carry out two steps of the process; message sent encryption and message received decryption in the system and sending a message to the WA system. We assume it is more practical and safe to have the model separately with the WA so that users do not need to encrypt the message they do not intend to if it does not require a secret thing.

CONCLUSION
The model developed has produced a new prototype in chatting application to add secret message as text security. The model can be used by users of the application in their smartphone. As the result, users will be more secure in using social chatting application than before. In the next research, we are trying to integrate the prototype into social chatting application in web so that the model can be used immediately and embedded in WA system.