Bi-Criteria Radio Spectrum Sharing With Subspace-Based Pareto Tracing

Area of Science:

• Wireless Communications
• Signal Processing
• Network Engineering

Background:

• Radio spectrum is a finite resource, necessitating efficient sharing between wireless technologies.
• Coexistence of Long-Term Evolution (LTE) License-Assisted Access (LAA) and Wi-Fi in unlicensed bands is crucial for meeting increasing demand.
• Optimizing the performance of both LAA and Wi-Fi systems simultaneously presents a significant challenge.

Purpose of the Study:

• To develop a technique for simultaneously optimizing the performance of coexisting LTE-LAA and Wi-Fi systems in shared unlicensed spectrum.
• To identify the key parameters that govern the optimal coexistence of these wireless technologies.
• To provide a computationally efficient method for approximating optimal parameter sets.

Main Methods:

• A novel technique to continuously estimate the Pareto frontier of parameter sets that maximize convex combinations of network throughputs.
• Application of active subspaces, a dimensionality reduction method, to identify the most relevant parameters for optimization.
• Formulation and solution of a reduced-dimension convex optimization problem.

Main Results:

• The near-optimal parameter set for coexistence is primarily determined by two physically relevant parameters.
• The active subspace method effectively reduces the complexity of the optimization problem.
• The proposed technique yields approximations that significantly outperform random grid search in optimizing coexistence performance.

Conclusions:

• A dimensionality reduction approach using active subspaces can efficiently identify optimal parameters for wireless system coexistence.
• This method enhances the explainability and performance of optimizing shared spectrum usage between LTE-LAA and Wi-Fi.
• The findings offer a practical solution for improving the performance of coexisting wireless networks in unlicensed bands.