Conference proceeding
A new jamming technique for secrecy in multi-antenna wireless networks
2010 IEEE International Symposium on Information Theory, pp.2513-2517
06/2010
DOI: 10.1109/ISIT.2010.5513559
Abstract
We consider the problem of secure wireless communication in the presence of an eavesdropper when the transmitter has multiple antennas, using a variation of the recently proposed artificial noise technique. Under this technique, the transmitter sends a pseudo-noise jamming signal to selectively degrade the link to the eavesdropper without affecting the desired receiver. The previous work in the literature focuses on ideal Gaussian signaling for both the desired signal and the noise signal. The main contribution of this paper is to show that the Gaussian signaling model has important limitations and propose an alternative "induced fading" jamming technique that takes some of these limitations into account. Specifically we show that under the Gaussian noise scheme, the eavesdropper is able to recover the desired signal with very low bit error rates when the transmitter is constrained to use constant envelope signaling. Furthermore, we show that an eavesdropper with multiple antennas is able to use simple, blind constant-envelope algorithms to completely remove the Gaussian artificial noise signal and thus defeat the secrecy scheme. We propose an alternative scheme that induces artificial fading in the channel to the eavesdropper, and show that it outperforms the Gaussian noise scheme in the sense of causing higher bit error rates at the eavesdropper and is also more resistant to constant modulus-type algorithms.
Details
- Title: Subtitle
- A new jamming technique for secrecy in multi-antenna wireless networks
- Creators
- Omar Bakr - University of California, BerkeleyRaghuraman Mudumbai - University of California, Berkeley
- Resource Type
- Conference proceeding
- Publication Details
- 2010 IEEE International Symposium on Information Theory, pp.2513-2517
- Publisher
- IEEE
- DOI
- 10.1109/ISIT.2010.5513559
- ISSN
- 2157-8095
- eISSN
- 2157-8117
- Language
- English
- Date published
- 06/2010
- Academic Unit
- Electrical and Computer Engineering
- Record Identifier
- 9984197327502771
Metrics
11 Record Views