A hybrid scheme for wireless physical layer security based on encryption and channel pre-compensation /

Physical layer security is an emerging research area that explores the possibility of achieving perfect-secrecy for data transmission among intended network by exploiting the characteristics of the wireless channel, such as fading or noise, to provide security for wireless transmissions. Physical la...

Full description

Saved in:
Bibliographic Details
Main Author: Tahir, Mohammad
Format: Thesis
Language:English
Published: Gombak, Selangor : Kulliyyah of Engineering, International Islamic University Malaysia, 2011
Subjects:
Online Access:http://studentrepo.iium.edu.my/handle/123456789/4671
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Physical layer security is an emerging research area that explores the possibility of achieving perfect-secrecy for data transmission among intended network by exploiting the characteristics of the wireless channel, such as fading or noise, to provide security for wireless transmissions. Physical layer security is not meant to replace existing higher layer security schemes but calls for the enhancement of the security level by combining with existing security methods. Therefore in this context a hybrid scheme that provides security in the physical layer, by combining the traditional cryptography and physical layer characteristics is proposed in this dissertation. The aim of combining these two techniques at physical layer is to deprive the eavesdropper from receiving the correct information in first place. This aim is achieved by two step process known as encryption and pre-compensation. Here, encryption refers to rotation of data symbols in original constellation based upon the key-sequence that is generated by the AES in counter mode. Encryption transforms the original data constellation of modulated symbols into higher order constellation. After encrypting the original constellation pre-compensation is applied, pre-compensation refers to compensation of data for channel variation before transmission i.e. compensation for phase and attenuation at transmitter. These two operations hide the original signal constellation due to which eavesdropper cannot demodulate the data as demodulation require knowledge signal constellation for correct demodulation. Hence eavesdropper is forced to perform blind channel estimation first in order to find encrypted constellation followed by task of finding key-sequence to decrypt the encrypted constellation. The key-sequence can be found correctly only when the effect of precompensation cancellation results in decoded bits that are error free (because to perform cryptanalysis the data should be error free otherwise it will result in erroneous result). Performance of the scheme is analysed in MATLAB® using bit error rate (BER) as performance metric, higher the BER for eavesdropper better is the security. Results under different settings showed (perfect and imperfect pre-compensation, different training duration of pilot symbols) that the eavesdropper suffers from high BER whereas the BER of legitimate receiver (intended receiver) depends upon the accuracy of CSI used for pre-compensation at transmitter. The performance was also evaluated for image and audio transmission, which also showed similar results of eavesdropper suffering from high BER. Due to high BER the received image and audio are completely unintelligible and no useful information can be extracted. Finally the performance of hybrid scheme in existing standard IEEE 802.11a was evaluated and it was found that system performs better due the effect of encryption and precompensation in terms of BER and security i.e. legitimate receiver suffers from a lower BER whereas the eavesdropper suffers from high BER.
Item Description:Abstracts in English and Arabic.
"A dissertation submitted in fulfilment of the requirement for the degree of Master of Science in Communication Engineering."--On t.p.
Physical Description:xiv. 103 leaves : ill. charts ; 30cm.
Bibliography:Includes bibliographical references (leaves 89-93).