Dynamic Flow-Adaptive Spectrum Leasing with Channel Aggregation in Cognitive Radio Networks

Lei Jiao, Xiang Xiao, Fanzi Zeng, Zhenzhen Hu, Lei Jiao 0001Sensors 2020 · DOI

TL;DR

This paper introduces a novel dynamic spectrum leasing strategy for cognitive radio networks that adaptively adjusts leased channels based on primary user (PU) traffic. By combining this adaptive leasing with channel aggregation for secondary users (SUs), the proposed approach significantly improves spectrum utilization and network capacity while reducing blocking and forced termination probabilities for SUs.

What Problem Does This Paper Address?

The ever-increasing demand for wireless services has led to a critical shortage of available spectrum. Cognitive radio networks (CRNs) offer a solution by allowing secondary users (SUs) to dynamically access licensed spectrum without disrupting primary users (PUs). However, traditional CRN approaches often overlook SU quality of service (QoS), leading to frequent spectrum handoffs and transmission limitations for SUs. Existing spectrum leasing schemes, while improving SU QoS, often use fixed leasing strategies that are inefficient in dynamic environments. This paper addresses the challenge of dynamically allocating spectrum resources in CRNs to balance the QoS requirements of both PUs and SUs, particularly in environments with fluctuating traffic demands.

Methods

The core of this work is the Dynamic Flow-Adaptive Spectrum Leasing (DFSL) strategy. Key components include:

  • Adaptive Channel Adjustment Algorithm: This algorithm dynamically adjusts the number of leased channels (L) based on the real-time number of ongoing and buffered PU flows. It categorizes PU traffic into four levels (ultra-low, low, medium, high) and uses predefined thresholds to determine the optimal number of leased channels, ensuring PUs have sufficient resources while maximizing leased spectrum for SUs.
  • Channel Aggregation for SUs: In the leased spectrum band (L-CRN), SUs employ dynamic channel aggregation. This allows a single SU flow to utilize multiple adjacent or non-adjacent spectrum holes simultaneously, enhancing their transmission efficiency and adapting to available resources.
  • Priority and Opportunistic Access: The strategy incorporates both priority access (in L-CRN for SUs) and opportunistic access (in N-CRN for PUs and SUs when available), providing flexibility during traffic surges.
  • Continuous Time Markov Chain (CTMC) Modeling: A CTMC model is developed to mathematically analyze the system's performance, capturing the complex state transitions based on flow arrivals and departures. This model allows for theoretical analysis of key performance metrics.

Datasets

Not discussed in the paper. The evaluation relies on simulations based on defined parameters and theoretical analysis.

Key Insights and Ideas

  • Dynamic Leasing is Crucial: Fixed spectrum leasing is inefficient. Adapting the leased spectrum based on real-time PU demand is key to balancing PU and SU needs and improving overall spectrum utilization.
  • Channel Aggregation Empowers SUs: Allowing SUs to aggregate multiple channels in the leased band significantly boosts their throughput and resilience, especially when PU activity is high.
  • Trade-offs Exist: While DFSL improves SU performance, it can lead to a slight decrease in PU network capacity compared to schemes that don't involve leasing. This highlights the inherent trade-off in spectrum sharing.
  • Thresholds Matter: The performance of the adaptive leasing algorithm is sensitive to the chosen thresholds for PU traffic levels, impacting SU blocking probability.

Key Results

The proposed DFSL strategy demonstrates significant improvements across several metrics:

  • Improved Spectrum Utilization and Network Capacity: DFSL effectively enhances the overall network capacity, particularly for SUs.
  • Reduced SU Blocking Probability: The adaptive nature of DFSL, coupled with channel aggregation, substantially lowers the probability of SU flows being blocked.
  • Reduced SU Forced Termination Probability: By prioritizing leased channels for SUs and limiting PU interruptions to the unleased spectrum (N-CRN), the forced termination probability for SUs is effectively minimized.
  • Comparison to Existing Schemes: DFSL outperforms fixed spectrum leasing (FSL) and dynamic channel aggregation without leasing (DCA) in terms of SU performance metrics. While traffic-adaptive spectrum leasing (TASL) shows comparable capacity, DFSL offers better adaptability.

Limitations and Future Work

  • Energy Efficiency: The authors acknowledge that energy efficiency was not considered in this work and is planned as future research.
  • Realistic Channel Conditions: The current model assumes ideal channel sensing and ignores interference uncertainty, which could influence real-world performance.
  • Simulation vs. Real-world: While the CTMC model is validated by simulations, a more comprehensive system-level simulation or real-measurement-based study is suggested for future work to capture more realistic scenarios.
  • Price Function: The paper proposes a price function for spectrum leasing, but its optimization and impact on PU incentives could be further explored.

Summary updated

BibTeX
@article{jiao2020dynamic,
  author = {Lei Jiao and Xiang Xiao and Fanzi Zeng and Zhenzhen Hu and Lei Jiao},
  title = {{Dynamic Flow-Adaptive Spectrum Leasing with Channel Aggregation in Cognitive Radio Networks}},
  journal = {Sensors},
  year = {2020},
  doi = {10.3390/s20133800},
}
Figure