Secrecy Performance Analysis of Distributed CDD Based Cooperative Systems with Jamming

Kyeong Jin Kim, Hongwu Liu, Marco Di Renzo, Philip V. Orlik, H. Vincent Poor

TL;DR

This paper introduces a novel physical layer security technique for cooperative wireless systems by combining distributed cyclic delay diversity (dCDD) with a strategically chosen "sentinel" transmitter that actively jams eavesdroppers. This approach enhances data confidentiality by degrading the signal quality for illegitimate users while improving it for the intended recipient, offering a practical solution for secure wireless communications.

What Problem Does This Paper Address?

In modern wireless communication, ensuring the confidentiality of transmitted data is a significant challenge. Unlike traditional encryption methods that operate at higher network layers, physical layer security aims to exploit the inherent characteristics of wireless channels to secure communications. A key problem is that signals broadcast by transmitters can be intercepted by unintended users (eavesdroppers or ERs) as easily as by the intended legitimate user (LR). While cooperative communication schemes can improve signal strength for the LR, they can also inadvertently boost the signal for ERs. Existing methods for enhancing physical layer security, such as relay selection or artificial noise transmission, often rely on explicit channel state information (CSI), which itself can be intercepted, compromising security. This paper addresses the need for a robust physical layer security mechanism that doesn't require explicit CSI feedback and effectively combats eavesdropping in cooperative systems.

Methods

The core innovation of this paper lies in the integration of distributed cyclic delay diversity (dCDD) with a sentinel transmitter jamming strategy.

  • dCDD: This technique, applied to cyclic-prefix single-carrier (CP-SC) systems, allows multiple cooperative transmitters to transmit the same data symbol with carefully chosen delays. These delays are designed to align the signals at the receiver, effectively converting a multi-input single-output (MISO) channel into an ISI-free single-input single-output (SISO) channel, thereby improving the received signal-to-noise ratio (SNR) at the legitimate user. The dCDD scheme is advantageous because it does not require explicit CSI feedback.
  • Sentinel Transmitter Jamming: The paper proposes selecting one of the dCDD transmitters to act as a "sentinel transmitter." This sentinel transmitter's primary role is to transmit a jamming signal specifically targeted at the eavesdropper (ER). The selection criterion for this sentinel transmitter is crucial: it's chosen based on having the strongest channel gain to the ER. By doing so, the jamming signal is most effective in degrading the signal-to-interference-plus-noise ratio (SINR) at the ER. The remaining dCDD transmitters function as data transmitters, sending the actual information to the LR.

The paper derives new closed-form expressions for the SNR at the LR and the SINR at the ER, considering frequency-selective fading channels. It then analyzes the secrecy outage probability, a key metric for physical layer security, which quantifies the probability that the secrecy capacity falls below a desired rate.

Datasets

Not discussed in the paper. The evaluation is based on Monte-Carlo simulations.

Key Insights and Ideas

  • Synergistic Security: Combining dCDD with a jamming sentinel transmitter creates a powerful synergy. dCDD improves the legitimate signal, while the sentinel's jamming actively degrades the eavesdropper's signal, leading to enhanced secrecy.
  • CSI-Free Security: The dCDD approach's inherent lack of reliance on explicit CSI feedback is a significant advantage for security, as it removes a potential vulnerability.
  • Optimal Jamming Strategy: Selecting the sentinel transmitter based on the strongest channel to the eavesdropper is an intuitive and effective strategy for maximizing jamming impact.
  • Trade-off Management: The system aims to increase the LR's SNR while decreasing the ER's SINR, demonstrating a sophisticated trade-off management for security.
  • Diversity Gain: The dCDD scheme, even with one transmitter acting as a jammer, still provides a diversity gain for the legitimate user, contributing to a more robust communication link.

Key Results

The paper presents analytical derivations for the SNR at the legitimate user and the SINR at the eavesdropper in frequency-selective fading channels. It then derives a closed-form expression for the secrecy outage probability.

  • Improved Secrecy: Simulations demonstrate that the proposed dCDD-based sentinel transmitter jamming scheme significantly improves secrecy performance, leading to lower secrecy outage probabilities compared to scenarios without jamming or with less optimal sentinel selection.
  • Diversity Gain Verification: The analysis confirms that the diversity gain is primarily influenced by the legitimate user's channels and the number of data transmitters, and is not significantly affected by the jamming strategy itself.
  • Impact of Jamming Power: Increasing the jamming power ratio relative to the data transmission power leads to a lower secrecy outage probability, as expected.
  • Sentinel Selection Effectiveness: The proposed selection of the sentinel transmitter based on the strongest channel to the ER is shown to be superior to selecting the second-best or least-effective channel transmitter.

Limitations and Future Work

  • Assumptions: The analysis assumes ideal backhaul connections between the control unit (CU) and transmitters, and that the jamming signal is known at the CU and LR. The paper also assumes single-antenna transmitters for simplicity.
  • Complexity: While the dCDD scheme itself is designed to avoid explicit CSI feedback, the selection of the sentinel transmitter still requires some knowledge of the ER channels, which is assumed to be available at the CU.
  • Future Work: The authors suggest further investigation into the impact of imperfect CSI and unreliable backhaul connections on the proposed system's performance. Exploring more advanced jamming strategies or adaptive sentinel selection could also be fruitful research directions.

Summary updated

Abstract

International audience
BibTeX
@inproceedings{kim2018secrecy,
  author = {Kyeong Jin Kim and Hongwu Liu and Marco Di Renzo and Philip V. Orlik and H. Vincent Poor},
  title = {{Secrecy Performance Analysis of Distributed CDD Based Cooperative Systems with Jamming}},
  booktitle = {ICC},
  year = {2018},
  doi = {10.1109/icc.2018.8422679},
}
Figure