Security in the exchange of information has been primarily treated as an inherently applied subject, despite the theoretical formulation of perfect secrecyby Shannon in 1946. In actual networks, security commonly relies on cryptographic algorithms implemented at upper layers of the protocol stack. Recently, a compelling complementary approach for enhancing the securing of wireless systems has risen from the area of information theory and has become a focal point of research in the wireless community.
The breakthrough concept of physical layer security is to exploit the characteristics of the wireless medium such as fading or noise to ensure secrecy in wireless transmissions. Seminal earlier analyses that investigated security aspects of the wiretap channel and the broadcast channel with confidential messages have established that a noisy communication channel can offer opportunities for perfectly secure exchange of information.
In this setting, the performance measure of interest, the secrecy capacity (SC), was defined as the largest communication rate for which encoding schemes exist that simultaneously guarantee reliability in the exchange of information with a legitimate user and perfect secrecy with respect to an eavesdropper. It has been demonstrated that the SC is strictly positive when the wiretap channel is on average a degraded version of the main channel. Similar results were obtained for wireless fading channels and multiple input multiple output (MIMO) systems.
The APLOE project contributed important results in four major topics of research on physical layer security (PLS): In particular, the APLOE PIOF-GA-2010-274723 project generated significant outputs in the following areas of research:
- Physical layer encryption for OFDM systems – Masked-OFDM systems: The first line of research of the APLOE project concerned the investigation of physical layer encryption for orthogonal frequency division multiplexing (OFDM) signals. We proposed a scheme termed Masked-OFDM in which OFDM signals are encrypted through faster than Nyquist signalling. The proposed scheme compromised neither the bandwidth efficiency nor the error performance of the underlying OFDM in additive white Gaussian noise channels (AWGN) and slow fading channels, at the cost of increased transmission power.
- Security based on asymmetric reception – The APLOE research team has established that in networks with asymmetries in (i) the mobility of the nodes, (ii) the number of legitimate users and eavesdroppers, (iii) the circuit imperfections of the intended destination and eavesdropping receivers, it is possible to employ PLS techniques while retaining transmission rates compatible with actual applications. Building on these results, the MC fellow currently investigates PLS techniques in multiuser networks with multiple adversaries for delay constrained applications. In particular, the focus of the current researches is on secure waterfilling techniques for networks with causal access to the channel state information.
- Security through cooperation and adaptation – The APLOE research team has characterized the achievable secrecy rates of discrete alphabet systems (e.g. M-ary quadrature amplitude modulation (M-QAM) systems). These results have been employed towards the stochastic characterization of the optimal jamming signals, in the framework of interference assisted secret communications.
- Security in active attacks – The PLS literature examined predominantly passive attacks, i.e., interception of a wiretap or a broadcast channel. The APLOE MC fellow has initiated and coordinated investigations regarding the achievable secrecy rates in systems with active attacks. In the investigated scenarios the active eavesdropper realizes the attack by providing false feedback (such scenarios are subcases of the Byzantine type of attacks).