Joint Spectrum Allocation and Structure Optimization in Green Powered Heterogeneous Cognitive Radio Networks

Ali Shahini, Nirwan AnsariarXiv 2017 · arXiv

TL;DR

This paper tackles the challenge of maximizing data throughput in green-powered cognitive radio networks by jointly optimizing how users access spectrum and how they manage their energy harvesting and transmission time. They propose a novel two-stage approach that breaks down a complex problem into manageable subproblems, leading to near-optimal solutions with a closed-form expression for energy harvesting ratios.

What Problem Does This Paper Address?

The ever-increasing demand for wireless data has led to a crowded radio spectrum. Cognitive Radio (CR) networks offer a solution by allowing secondary users (SUs) to opportunistically access underutilized spectrum licensed to primary users (PUs). However, efficiently managing these networks, especially when SUs rely on energy harvesting for power, presents significant challenges. This paper focuses on heterogeneous CR networks where SUs have different needs (real-time vs. non-real-time) and must balance energy harvesting, spectrum sensing, and data transmission within limited time slots. The core problem is to find the optimal way to allocate spectrum (sub-channels) and structure the time slots to maximize the total data rate for all SUs while respecting PU interference limits and SU rate requirements.

Methods

The authors tackle this complex problem using a two-stage approach:

  1. Sub-channel Allocation: They first address the problem of assigning sub-channels to SUs. To simplify the initial mixed-integer non-linear programming (MINLP) problem, they propose a sub-channel allocation scheme based on an "Energy Figure of Merit" (EFM). This scheme prioritizes real-time (RT) SUs and aims to satisfy their rate requirements, effectively removing integer constraints.
  2. Structure Optimization: After sub-channel allocation, the problem is reformulated into a convex optimization task. Since a simple closed-form solution for optimal harvesting ratios isn't directly available, they derive a near-optimal closed-form solution using the Lambert W function. They also employ an iterative gradient method based on Lagrangian dual decomposition as an alternative approach.

Datasets

Not discussed in the paper. The paper relies on simulations to evaluate its proposed methods.

Key Insights and Ideas

  • Joint Optimization is Crucial: The paper highlights that simply optimizing spectrum allocation or time slot structure in isolation is insufficient. A joint approach is necessary to achieve optimal performance in energy-constrained CR networks.
  • Time-Switching is Key for Energy Harvesting: The time-switching model, where each time slot is divided into energy harvesting, spectrum sensing, and data transmission phases, is fundamental to managing energy-harvesting SUs. The trade-off between harvesting more energy and having less time for transmission is a central theme.
  • Prioritizing Real-Time Users: The proposed sub-channel allocation scheme gives priority to real-time SUs, ensuring their critical data transmission needs are met first, which is essential for maintaining quality of service.
  • Lambert W Function for Closed-Form Solutions: The application of the Lambert W function to derive a near-optimal closed-form solution for harvesting ratios is a novel and elegant mathematical contribution.
  • Convexity Simplifies Optimization: Proving the convexity of the optimization problem after sub-channel allocation is a significant step, enabling the use of efficient convex optimization techniques.

Key Results

  • The proposed sub-channel allocation scheme outperforms existing methods, especially when the number of available sub-channels is limited.
  • The derived closed-form solution for harvesting ratios achieves a performance gap of less than 3.5% compared to optimal solutions.
  • Simulations demonstrate that the achievable sum rate increases with the number of available sub-channels and decreases slightly with higher rate constraints for RT users.
  • The system performance is shown to be sensitive to PU interference thresholds, with lower thresholds limiting SU transmission power and thus sum rate.

Limitations and Future Work

  • Perfect CSI Assumption: The paper assumes perfect knowledge of Channel State Information (CSI) between SUs and the Access Point (AP), which may not always hold in real-world scenarios.
  • Simplified Interference Model: While interference is considered, the model might be simplified, and more complex interference scenarios could be explored.
  • Power Allocation Beyond Scope: The paper mentions that power allocation for assigned sub-channels is beyond the scope of this work, suggesting it as a potential area for future research.
  • Further Exploration of Lambert W Function Applications: The authors proved a lemma that could be useful for other similar problems, hinting at broader applications of their mathematical approach.
  • Real-World Deployment Challenges: Future work could investigate the practical implementation challenges and performance in dynamic and more complex heterogeneous CR network environments.

Summary updated

BibTeX
@misc{shahini2017joint,
  author = {Ali Shahini and Nirwan Ansari},
  title = {{Joint Spectrum Allocation and Structure Optimization in Green Powered Heterogeneous Cognitive Radio Networks}},
  year = {2017},
  eprint = {1705.04289},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/1705.04289},
}
Figure