Asian Research Thesis Index

Abstract

The internet is getting credit for being a backbone of information sharing these days. As the
demand of the transmission of information on the un-trusted network increases, people
demanding the more secure mechanism to share data on hybrid networks. Now a day, the Ib
contents contain multimedia which demands security along with a maximum compression ratio
of transmission over the networks. The aim of this research paper is twofold, one the novel
lightIight encryption method is proposed which is based on Permutation rounds with exclusive
OR (P-XOR) and second the implementation of expert systems to auto detects three different
security levels for hybrid devices/networks. P-XOR is a symmetric stream cipher with three keys
of 128 bits along with five rounds. As per the need of heterogeneous devices and networks, three
security levels are implemented on the compression stage of H.264 bit-streams. Three cipher
algorithms, i.e. exclusive OR (XOR) for low level security, Advanced Encryption Standard
(AES) for high level security and newly proposed P-XOR for medium level security are selected
by the implementation of fuzzy rule based expert system. The proposed expert system
intelligently selected the best security level by judging the device and network requirements for
H.264 video transmission. The proposed framework has resolved the key problem of selecting
the same level of hybrid infrastructure, i.e. the devices like mobiles, tablets, laptops and servers
with differing requirements like energy, storage, and resolution and frame rate. The results on
different sample videos have confirmed the validity of the proposed expert system for selection
of different security levels. As a research contribution, the implementation of proposed
framework has provided the minimum computational cost due to selective encryption, better
performance in terms of selection time by the expert system, video format compliance and
without network overhead for encrypted videos on said security levels.

Add/Update Thesis