Energy Efficiency Optimization for Secure Transmission in MISO Cognitive Radio Network with Energy Harvesting

Miao Zhang, Kanapathippillai Cumanan, Jeyan Thiyagalingam, Wei Wang, Alister G. Burr, Zhiguo Ding, Octavia A. DobreZenodo 2017 · DOI

TL;DR

This paper tackles the challenge of optimizing both security and energy efficiency in wireless communication networks that also harvest energy. The researchers propose novel algorithms to maximize "Secrecy Energy Efficiency" (SEE) in a cognitive radio network, ensuring confidential data transmission while minimizing power consumption and meeting energy harvesting demands.

What Problem Does This Paper Address?

Modern wireless communication systems are pushing for higher data rates and lower latency, but this comes at a significant energy cost. Simultaneously, ensuring the security of transmitted data is paramount, especially in shared spectrum environments like cognitive radio (CR) networks. Adding to this complexity, many devices now rely on energy harvesting (EH) to power themselves, creating a multi-objective optimization problem. Specifically, this paper addresses the challenge of maximizing the "Secrecy Energy Efficiency" (SEE) in a MISO (Multiple-Input Single-Output) CR network. This means finding the best way for a transmitter with multiple antennas to send secure information and energy to a receiver, while minimizing interference to a primary user and ensuring the energy receiver harvests enough power. The problem is made more difficult by the fact that the energy receiver might also act as an eavesdropper, and the exact channel conditions might not always be perfectly known.

Methods

The core of this paper lies in developing sophisticated optimization techniques to solve a non-convex problem. The authors formulate the SEE maximization as a non-linear fractional programming problem. To tackle its inherent non-convexity, they employ a combination of:

  • Non-linear Fractional Programming: This technique is used to transform the fractional objective function into a more manageable form.
  • Difference of Concave (DC) Functions Programming: This method is applied to handle objective functions that can be expressed as the difference of two concave functions, a common characteristic of non-convex optimization problems.
  • Dinkelbach's Algorithm: This iterative algorithm is used in conjunction with the above techniques to find the optimal solution for the SEE maximization problem.
  • Semidefinite Relaxation (SDR): For scenarios with imperfect channel state information (CSI), SDR is employed to relax the rank-one constraint, making the problem more tractable.
  • S-Procedure: This technique is used to convert robust optimization problems with ellipsoidal channel uncertainties into convex ones.

The paper proposes iterative algorithms that converge to the optimal SEE under various CSI assumptions (perfect, statistical, and imperfect).

Datasets

Not discussed in the paper. The paper focuses on theoretical modeling and simulation-based validation using randomly generated channel sets.

Key Insights and Ideas

  • The SEE Trade-off: The paper highlights that maximizing SEE is a complex balancing act. Increasing secrecy rates or energy harvesting requirements can negatively impact overall energy efficiency, and vice-versa.
  • ER as Eavesdropper: The assumption that the energy receiver (ER) could also be an eavesdropper adds a crucial layer of security consideration to the energy harvesting process.
  • CSI Matters: The accuracy of channel state information (CSI) significantly impacts the achievable SEE. Perfect CSI yields the best performance, while imperfect CSI necessitates more robust designs and potentially higher transmit power.
  • Iterative Solutions are Key: Due to the non-convex nature of the optimization problems, iterative algorithms are essential for finding practical solutions. The proposed algorithms demonstrate good convergence properties.
  • Robustness is Crucial: In real-world scenarios, perfect CSI is rare. The paper's exploration of statistical and imperfect CSI demonstrates the importance of robust designs that can handle channel uncertainties.

Key Results

The paper presents several key findings through simulations:

  • Algorithm Convergence: The proposed iterative algorithms demonstrate fast convergence to the optimal SEE across different CSI assumptions and channel conditions.
  • Performance Gains: The SEE maximization scheme consistently outperforms power minimization and secrecy rate maximization schemes, especially when considering the joint objectives.
  • Impact of Parameters:
    • Target Secrecy Rate: Higher target secrecy rates generally lead to lower achievable SEE, as more power is needed to meet the secrecy demands.
  • EH Requirement: A higher EH requirement at the ER also tends to decrease SEE, as more power must be allocated for energy transfer.
  • Transmit Power: Increasing transmit power can improve secrecy rate but doesn't necessarily improve SEE beyond a certain point, especially if the problem is already feasible.
  • Number of Antennas: Increasing the number of transmit antennas at the SU-Tx generally improves SEE by providing more degrees of freedom for better signal design.
  • Energy Conversion Ratio: A higher energy conversion ratio at the ER leads to better SEE, as less power is wasted in the conversion process.
  • Robustness Benefits: The robust designs for statistical and imperfect CSI show that while performance might be slightly lower than with perfect CSI, they offer a more realistic and reliable solution in the face of channel uncertainties.

Limitations and Future Work

  • Computational Complexity: While the proposed algorithms are iterative, the computational complexity can still be significant, especially for large systems. The paper acknowledges this and provides complexity analyses.
  • Simplified EH Model: The paper primarily uses a linear EH model, although it mentions the existence of non-linear EH models. Future work could explore optimization with more complex non-linear EH models.
  • Specific CR Scenarios: The paper focuses on an underlay MISO CR network. Future research could extend these concepts to other CR architectures (e.g., overlay) or different wireless network types.
  • Practical Implementation: While simulations are provided, real-world implementation challenges such as hardware imperfections, feedback delays, and more complex interference models could be explored.

Summary updated

Abstract

This paper investigates a secure energy efficiency (SEE) optimization problem in a multiple-input single-output (MISO) underlay cognitive radio (CR) network. In particular, a multi-antenna secondary transmitter (SU-Tx) simultaneously sends secured information and energy to a secondary receiver (SU-Rx) and an energy receiver (ER), respectively, in the presence of a primary receiver (PU-Rx). It is assumed that the SU-Rx, ER and PU-Rx are each equipped with a single antenna. In addition, the SU-Tx should satisfy constraints on maximum interference leakage to the PU-Rx and minimum harvested energy at the ER. In this CR network, we consider the transmit covariance matrix design with the assumption of perfect channel state information (CSI) at the SU-Tx. In addition, it is assumed that the ER is a potential passive eavesdropper due to broadcast nature of wireless transmission. On the other hand, we consider the worst-case scenario that ER's energy harvesting requirement is only satisfied when it performs only energy harvesting without intercepting or eavesdropping information intended for the SU-Rx. We formulate this transmit covariance matrix design as a SEE maximization problem which is a non-convex problem due the non-linear fractional objective function. To realize the solution for this non-convex problem, we utilize the non-linear fractional programming and difference of concave (DC) functions approaches to reformulate into a tractable form. Based on these techniques and the Dinkelbach's method, we propose iterative algorithms to determine the solution for the original SEE maximization problem. Numerical simulation results are provided to demonstrate the performance of the proposed transmit covariance matrix design and convergence of the proposed algorithms.
BibTeX
@inproceedings{zhang2017energy,
  author = {Miao Zhang and Kanapathippillai Cumanan and Jeyan Thiyagalingam and Wei Wang and Alister G. Burr and Zhiguo Ding and Octavia A. Dobre},
  title = {{Energy Efficiency Optimization for Secure Transmission in MISO Cognitive Radio Network with Energy Harvesting}},
  booktitle = {Zenodo},
  year = {2017},
  doi = {10.5281/zenodo.3825234},
}
Figure