In this letter, we investigate the performance of the secrecy capacity in an amplify-and-forward dual-hop network for both distributed beamforming (DBF) and opportunistic relaying (OR) techniques. We derive the capacity scaling for two sets; U untrustworthy aggressive relays cooperating together with a wire-tapper to intercept the message, and T trustworthy relays, for a large number of nodes. We prove that the capacity scaling in the DBF is bounded by a value depending on a ratio between the number of the trustworthy and the untrustworthy aggressive relays. Finally, we show that DBF is better than OR whose capacity scaling is proved to be upper-bounded by a value tending to zero when the untrustworthy relays are aggressive. Simulation results confirm our analytical derivations.
In this letter, we investigate the performance of the secrecy capacity in an amplify-and-forward dual-hop network for both distributed beamforming (DBF) and opportunistic relaying (OR) techniques. We derive the capacity scaling for two sets; U untrustworthy aggressive relays cooperating together with a wire-tapper to intercept the message, and T trustworthy relays, for a large number of nodes. We prove that the capacity scaling in the DBF is bounded by a value depending on a ratio between the number of the trustworthy and the untrustworthy aggressive relays. Finally, we show that DBF is better than OR whose capacity scaling is proved to be upper-bounded by a value tending to zero when the untrustworthy relays are aggressive. Simulation results confirm our analytical derivations.
This paper presents an investigation of the secrecy performance of amplify-and-forward (AF) distributed beamforming (DBF) and opportunistic relaying (OR) through K untrustworthy relays. The capacity scaling and achievable diversity order of a secure communication are derived for K → ∞ and in high signal-to-noise-ratio (SNR) regimes, respectively. The DBF is shown to convey confidential messages with a capacity scaling of 1/2 log K regardless of the trustworthiness of the relays, even without intended jamming (IJ), where 1/2 log K corresponds to the maximal capacity scaling of cooperative relaying in non-eavesdropping environments. However, using multiple untrustworthy relays worsens the secrecy performance of the OR, and as K increases, the secrecy capacity converges to zero. When secrecy outage performance with IJ is considered, the achievable diversity order for both the DBF and OR is limited to one. However, only the DBF exploits the benefits using multiple untrustworthy relays, and its secrecy outage probability decreases exponentially as K increases.
Wireless networks have experienced rapid evolutions toward sustainability, scalability and interoperability. The digital economy is driven by future networked societies to a more holistic community of intelligent infrastructures and connected services for a more sustainable and smarter society. Furthermore, an enormous amount of sensitive and confidential information, e.g., medical records, electronic media, financial data, and customer files, is transmitted via wireless channels. The implementation of higher layer key distribution and management was challenged by the emergence of these new advanced systems. In order to resist various malicious abuses and security attacks, physical layer security (PLS) has become an appealing alternative. The basic concept behind PLS is to exploit the characteristics of wireless channels for the confidentiality. Its target is to blind the eavesdroppers such that they cannot extract any confidential information from the received signals. This thesis presents solutions and analyses to improve the PLS in wireless networks. \n \nIn the second chapter, we investigate the secrecy capacity performance of an amplify-andforward (AF) dual-hop network for both distributed beamforming (DBF) and opportunistic relaying (OR) techniques. We derive the capacity scaling for two large sets; trustworthy relays and untrustworthy aggressive relays cooperating together with a wire-tapper aiming to intercept the message. We show that the capacity scaling in the DBF is lower bounded by a value which depends on the ratio between the number of the trustworthy and the untrustworthy aggressive relays, whereas the capacity scaling of OR is upper bounded by a value depending on the number of relays as well as the signal to noise ratio (SNR). \n \nIn the third chapter, we propose a new location-based multicasting technique, for dual phase AF large networks, aiming to improve the security in the presence of non-colluding passive eavesdroppers. We analytically demonstrate that the proposed technique increases the security by decreasing the probability of re-choosing a sector that has eavesdroppers, for each transmission time. Moreover, we also show that the secrecy capacity scaling of our technique is the same as for broadcasting. Hereafter, the lower and upper bounds of the secrecy outage probability are calculated, and it is shown that the security performance is remarkably enhanced, compared to the conventional multicasting technique. \n \nIn the fourth chapter, we propose a new cooperative protocol, for dual phase amplify-andforward large wireless sensor networks, aiming to improve the transmission security while taking into account the limited capabilities of the sensor nodes. In such a network, a portion of the K relays can be potential passive eavesdroppers. To reduce the impact of these untrustworthy relays on the network security, we propose a new transmission protocol, where the source agrees to share with the destination a given channel state information (CSI) of source-trusted relay-destination link to encode the message. Then, the source will use this CSI again to map the right message to a certain sector while transmitting fake messages to the other sectors. Adopting such a security protocol is promising because of the availability of a high number of cheap electronic sensors with limited computational capabilities. For the proposed scheme, we derived the secrecy outage probability (SOP) and demonstrated that the probability of receiving the right encoded information by an untrustworthy relay is inversely proportional to the number of sectors. We also show that the aggressive behavior of cooperating untrusted relays is not effective compared to the case where each untrusted relay is trying to intercept the transmitted message individually. \n \nFifth and last, we investigate the physical layer security performance over Rayleigh fading channels in the presence of impulsive noise, as encountered, for instance, in smart grid environments. For this scheme, secrecy performance metrics were considered with and without destination assisted jamming at the eavesdropper’s side. From the obtained results, it is verified that the SOP, without destination assisted jamming, is flooring at high signal-to-noise-ratio values and that it can be significantly improved with the use of jamming.
This paper studies the secrecy capacity scaling in an Amplify-and-Forward (AF) dual-phase large network containing K relays. In our model, part of these K relays are assumed to be potential eavesdroppers. Before transmitting the message in the first phase, the multi-antennas source divides it to partial messages and multicasts each part to a different disjoint sector. During the second phase, the K relays use the distributed beamforming (DBF) technique to retransmit the message to the destination. We investigate the ergodic secrecy capacity considering two different behaviours of the untrustworthy relays; the passive behaviour, when the untrustworthy relays work separately from each other to intercept the signal, and the aggressive one, when the untrusted relays collaborate to hijack the message. As demonstrated, our location-based multicasting scenario is significantly increasing the security compared to the recent works that employ broadcasting schemes. Additionally, it also increases the secrecy capacity scaling remarkably. Finally, our analytical derivations are confirmed by the simulation.
This paper investigates the problem of secure communication for amplify-and-forward (AF) systems with untrustworthy relay nodes. To achieve positive secrecy rate, the destination-based jamming (DBJ) technique is applied. We first focus on the single-relay scenario, for which the closed-form expression for the lower bound of the ergodic secrecy capacity (ESC) is derived. Afterward, we extend the DBJ method to the multi-relay scenario and propose a secure relay selection scheme that can maximize the achievable secrecy rate. Based on the extreme-value theory, the approximate ESC is characterized. Our work reveals an interesting result that, when the relay nodes are untrustworthy, the system performance worsens as the number of relays increases.
This paper investigates the ergodic secrecy capacity (ESC) of multiple-antenna amplify-and-forward relay systems, where one eavesdropper can wiretap the relay. To reveal the capability of the multiple- antenna relay in improving the secrecy performance, we derive new tight closed-form lower bounds of the ESC for two secure transmission schemes: artificial noise aided precoding (ANP) and eigen-beamforming (EB). We also derive the lower bound of the ESC for ANP with a large antenna array at the relay, and investigate their corresponding asymptotic performance in the high and low SNR regimes. Based on the asymptotic analysis, we optimally allocate the power to the information signal and the artificial noise. Both the analysis and simulation results indicate that, in the moderate-to-high SNR regime, ANP achieves considerable performance gain over EB, while in the low SNR regime, EB outperforms ANP with equal power allocation. As SNR grows large, the ESC of EB approaches a constant only related to the number of relay antennas. Moreover, in the high SNR regime, it is optimal to allocate around half of total power to artificial noise for ANP.
This paper investigates the ergodic secrecy capacity (ESC) of multiple-antenna amplify-and-forward relay systems, where one eavesdropper can wiretap the relay. To reveal the capability of the multiple- antenna relay in improving the secrecy performance, we derive new tight closed-form lower bounds of the ESC for two secure transmission schemes: artificial noise aided precoding (ANP) and eigen-beamforming (EB). We also derive the lower bound of the ESC for ANP with a large antenna array at the relay, and investigate their corresponding asymptotic performance in the high and low SNR regimes. Based on the asymptotic analysis, we optimally allocate the power to the information signal and the artificial noise. Both the analysis and simulation results indicate that, in the moderate-to-high SNR regime, ANP achieves considerable performance gain over EB, while in the low SNR regime, EB outperforms ANP with equal power allocation. As SNR grows large, the ESC of EB approaches a constant only related to the number of relay antennas. Moreover, in the high SNR regime, it is optimal to allocate around half of total power to artificial noise for ANP.