New Cryptographic Algorithms for Enhancing Security of Voice Data

• Main Author: Al-Hazaimeh, Obaida Mohammad Awad
• Format: Thesis
• Language: English; English
• Published: 2010
• Subjects: QA75 Electronic computers. Computer science; TK5101-6720 Telecommunication
• Online Access: http://etd.uum.edu.my/2740/1/Obaida_Mohammad_Awad_Al-Hazaimeh.pdf; http://etd.uum.edu.my/2740/2/1.Obaida_Mohammad_Awad_Al-Hazaimeh.pdf; http://etd.uum.edu.my/2740/; http://lintas.uum.edu.my:8080/elmu/index.jsp?module=webopac-l&action=fullDisplayRetriever.jsp&szMaterialNo=0000772230
• Institution: Universiti Utara Malaysia

A real-time application Voice over Internet Protocol (VoIP) is the technology that enables voice packets transmission over internet protocol (IP). Security is of concern whenever open networks are to be used. In general, the real-time applications suffer from packet latency and loss due to the nature of IP network. Cryptographic systems may be used to achieve VoIP security, but their impact on the Quality of Services (QoS) should be minimized. Most of the known encryption algorithms are computationally expensive resulting in a significant amount of time added to packet delay. VoIP is usually used by public users resulting in a key exchange problem and a trusted intermediate authority normally takes this responsibility. In this research, VoIP security was enhanced via a proposed cryptographic system. The proposed solution consists of a simple, but strong encryption/decryption algorithm as well as an embedded method to exchange the keys between the users. In this research, a new keys is generated in a random fashion and then used to encrypt each new voice packet to strengthen the security level. Key exchange is carried out by inserting the key with the ciphered voice packet that depends on the table of the key positions at the sender and receiver sides, and the target receiver is the only one who is able to extract the key. The encryption process in this research is divided into three main stages: key generation, encryption process, and key insertion process. The decryption process on the other hand is divided into two main stages: key extraction process, and decryption process. The proposed solution was implemented and tested and the results showed that the required time for the security processes is minimized compared to some known algorithms such as AES_Rijndael algorithm. Furthermore, the analysis has proved that the security level has a direct relationship to the key length and the voice packet size in that large packet size requires more processing time. Finally, the implementation result in this research shows the average time needed to encrypt and decrypt a voice packet size using a proposed algorithm with the long key of 1024-bits is much smaller than AES_Rijndael algorithm with a short key length of 128-bits.