The physical layer describes what happens when an electromagnetic wave leaves our device.  To obtain physical layer security (PLS) is to exploit the physical properties of the channel to secure our information against eavesdropper. At the Smart Internet Lab, we use a range of approaches to achieve this. 

Security at physical layer is by no means intended to replace cryptographic security, but rather, it affords an additional protection layer.

What happens when we send a message to our bank, our parents, or our doctor? Once it leaves our PC or our mobile, we need to be reassured it will arrive at our recipient securely, without being read en route by an eavesdropper. Not only people, but also machines need to communicate securely – with the Internet of Things taking off, 'x' machines will be online by 2021, all talking to each other. NHS scans, medical history, diplomatic interactions - we need these to be secure. 

Communication security is traditionally provided by methods such as shared secret keys. Such techniques take place in the upper network layers. In contrast, physical layer security exploits the inherent randomness of the wireless medium.   In a wireless channel, this variability is present due to multipath propagation and unpredictable fluctuations in the signal strength, Doppler and delay domains.  Physical layer security has information-theoretic foundations and dates to the original work of Shannon. In the basic setup, the inherent randomness of the physical medium may be leveraged to generate secret keys at both communications end points, without any explicit key exchange.  Physical Layer Security is also concerned with establishing limits of perfectly secure communications, i.e. with zero probability of eavesdropping. A complementary theme is a search for practical techniques which increase the secrecy rates. Beamforming, a technique originally devised to direct the signal towards the intended receiver(s) and simultaneously supress the signal to unintended uses, can also be used in the context of increasing the security rates.  

Traditional a signal processing technique that can be employed in order to direct the signal towards the intended receiver(s). That is, instead of transmitting the signal isotopically, the signal follows a fixed direction to reach the legitimate receiver, avoiding the geographical position of the. Therefore, as long as the malicious eavesdropper is in a different direction, she will not be able to infer the message. This can also be used in reverse: to jam the location of an eavesdropper by transmitting noise at her to jamming techniques are employed. In this context, jamming refers to transmitting noise towards the eavesdropper in order to decrease her encoding capabilities. 

We can also use the fundamental physical characteristics of the wireless channel -such as noise and channel variations- to provide secure communications.  Part of the Communication Systems and Networks (CSN) research group of the University of Bristol examines keyless secure transmission via signal design and signal processing techniques in order to provide confidentiality. 

Furthermore, we can also exploit the underlying fundamental quantum nature of the communications channel, down to the level of individual photons. By using this, we can use the unique properties of single photons to securely exchange keys between ourselves and our interlocutor: using these on the optical part of the network we can demonstrate provably secure communications.