A Coalitional Formation Game for Physical Layer Security of Cooperative Compressive Sensing Multi-Relay Networks

Jialun Li, Shuai Chang, Xiaomei Fu, Liang Zhang, Yishan Su, Zhigang Jin2018 · DOI

TL;DR

This paper proposes a novel approach to enhance physical layer security in wireless networks by combining cooperative compressive sensing (CCS-AF) with a coalitional game theory framework for optimal relay selection. The key innovation is a "Max-Coalition order" that prioritizes overall network secrecy capacity, leading to more efficient relay cooperation and improved security compared to traditional methods.

What Problem Does This Paper Address?

Traditional wireless security often relies on complex encryption, which can be computationally intensive and difficult to manage in resource-constrained networks like wireless sensor networks (WSNs). Physical layer security offers an alternative by exploiting the inherent characteristics of wireless channels to secure communication without encryption keys. However, achieving robust physical layer security in multi-relay networks presents challenges. Specifically, the effectiveness of security can be limited by channel conditions, and selecting the optimal subset of relays for cooperation is crucial for maximizing secrecy capacity. Existing relay selection schemes often require centralized control, which is impractical for distributed WSNs. Furthermore, simply selecting all relays doesn't guarantee optimal security, as the number of relays can impact amplification coefficients and overall performance.

Methods

The paper introduces a Cooperative Compressive Sensing Amplify-and-Forward (CCS-AF) scheme. This scheme integrates compressive sensing (CS) for efficient signal acquisition with an amplify-and-forward (AF) relaying strategy. The core of the proposed solution lies in a coalitional game theory framework for relay selection. Instead of traditional "Pareto order" which focuses on individual player utility, this paper defines a novel "Max-Coalition order." This order prioritizes the formation of coalitions that maximize the overall network secrecy capacity, rather than individual relay gains.

The coalition formation process involves:

  • Merge-and-Split Rules: Relays can merge to form larger coalitions if it improves secrecy capacity, or split existing coalitions if it leads to better outcomes.
  • Iterative Algorithm: A distributed algorithm guides the relays through neighbor discovery, secrecy capacity calculation, and iterative coalition formation (merging and splitting) until a stable, optimal coalition structure is achieved.

Datasets

Not discussed in the paper. The simulations utilize a simulated network environment with specified parameters for sources, relays, eavesdroppers, and their locations.

Key Insights and Ideas

  • Secrecy Capacity is Not Monotonic with Relay Count: The paper demonstrates that simply increasing the number of relays doesn't always lead to higher secrecy capacity. There's an optimal number of relays that maximizes security, influenced by factors like source-relay distances and amplification coefficients.
  • "Max-Coalition Order" for Global Optimization: The novel "Max-Coalition order" is a significant departure from individualistic game theory approaches. By focusing on the collective secrecy capacity, it enables more effective cooperation and better overall network security.
  • Self-Organization and Adaptability: The coalitional game approach allows relays to self-organize and adapt to dynamic network conditions, such as relay mobility, without centralized control.
  • Compressive Sensing as a Security Enabler: The paper leverages CS not just for efficient data acquisition but also as a component of physical layer security, where the measurement matrix can contribute to encryption.

Key Results

  • The proposed CCS-AF scheme with the "Max-Coalition order" significantly improves the secrecy capacity of cooperative multi-relay networks compared to using all relays or traditional Pareto-based selection.
  • Simulations show that the algorithm can adapt to relay mobility, dynamically forming new coalitions to maintain optimal secrecy.
  • The "Max-Coalition order" leads to higher overall network utility (secrecy capacity) than the Pareto order in the simulated scenarios.
  • The paper demonstrates that increasing relay power (PR) generally improves system performance and the number of selected relays, up to a certain point.
  • Conversely, increasing source power (Ps) beyond a certain threshold can negatively impact secrecy capacity due to increased eavesdropper capabilities.

Limitations and Future Work

  • CSI Accuracy: The paper notes that the accuracy of channel state information (CSI) is critical for the coalition formation algorithm. Significant CSI errors could lead to suboptimal coalition selection.
  • Computational Complexity: While the algorithm aims for efficiency by stopping when a better coalition is found, the split operation's complexity can grow exponentially with the number of relays in a coalition.
  • Future Work: The authors suggest that future research could explore more complex network topologies, different relaying strategies, and investigate the impact of imperfect CSI on the proposed algorithm. Further analysis of the trade-offs between secrecy capacity, energy efficiency, and other performance metrics would also be valuable.

Summary updated

BibTeX
@misc{li2018coalitional,
  author = {Jialun Li and Shuai Chang and Xiaomei Fu and Liang Zhang and Yishan Su and Zhigang Jin},
  title = {{A Coalitional Formation Game for Physical Layer Security of Cooperative Compressive Sensing Multi-Relay Networks}},
  year = {2018},
  doi = {10.3390/s18092942},
}
Figure