[Paper Review] Decentralized Opportunistic Access for D2D Underlaid Cellular Networks

Zheng Chen, Marios Kountouris|arXiv (Cornell University)|Jul 19, 2016

Advanced MIMO Systems Optimization34 references3 citations

TL;DR

This paper proposes a decentralized, SIR-aware access control scheme for D2D underlaid cellular networks that combines signal-to-interference ratio (SIR)-based link activation with cellular exclusion regions (guard zones). Using stochastic geometry, it derives analytical expressions for coverage probabilities and optimizes the SIR threshold and guard zone radius to maximize D2D area spectral efficiency while maintaining cellular uplink coverage, achieving significant throughput gains over state-of-the-art schemes.

ABSTRACT

We propose a decentralized access control scheme for interference management in D2D (device-to-device) underlaid cellular networks. Our method combines SIR-aware link activation with cellular exclusion regions in a case where D2D links opportunistically access the licensed cellular spectrum. Analytical expressions and tight approximations for the coverage probabilities of cellular and D2D links are derived. We characterize the impact of the guard zone radius and the SIR threshold on the D2D area spectral efficiency and cellular coverage. A tractable approach was proposed in order to find the SIR threshold and guard zone radius, which maximize the area spectral efficiency of the D2D communication while ensuring sufficient coverage probability for cellular uplink users. Simulations validate the accuracy of our analytical results and show the performance gain of our proposed scheme compared to existing state-of-the-art solutions.

Motivation & Objective

• Address the challenge of interference management in dense D2D underlaid cellular networks where D2D links share the same licensed spectrum as cellular users.
• Develop a fully decentralized access control mechanism that does not rely on centralized coordination or channel state information exchange.
• Simultaneously optimize the SIR threshold and guard zone radius to maximize D2D area spectral efficiency (ASE) under a minimum cellular coverage probability constraint.
• Provide tractable analytical expressions for coverage probabilities of both D2D and cellular links using stochastic geometry tools.
• Validate the proposed scheme's performance gain over baseline and state-of-the-art access control methods through extensive simulations.

Proposed method

• Model the locations of cellular base stations (BSs) and D2D transmitters as independent homogeneous Poisson point processes (PPPs).
• Introduce a guard zone (exclusion region) around each cellular uplink receiver to reduce co-channel interference from D2D transmissions.
• Implement SIR-aware link activation: D2D links access the spectrum only if their SIR to the serving BS exceeds a threshold $ p_s $, which controls the access probability.
• Derive the Laplace transform of the aggregate interference at the typical cellular receiver using the Poisson-Voronoi tessellation approximation for the Voronoi cell shape.
• Use stochastic geometry to derive tight analytical approximations for the coverage probabilities of both D2D and cellular links.
• Formulate a decoupled optimization problem to jointly find the optimal SIR threshold $ G^* $ and guard zone radius $ ho^* $ that maximize D2D ASE while satisfying a cellular coverage constraint.

Experimental results

Research questions

• RQ1 How does the combination of SIR-aware activation and guard zones affect the coverage probability of D2D and cellular links in a decentralized D2D underlaid network?
• RQ2 What is the optimal SIR threshold and guard zone radius that maximize D2D area spectral efficiency while preserving a minimum cellular uplink coverage probability?
• RQ3 How does the proposed decentralized scheme compare in performance to baseline schemes without access control and to schemes using only guard zones or only SIR-based access?
• RQ4 To what extent can the analytical framework based on stochastic geometry accurately predict the system performance in realistic network deployments?
• RQ5 What is the impact of key system parameters—such as D2D density, cellular density, and path loss exponent—on the optimal access strategy?

Key findings

• The proposed scheme achieves substantial throughput gains over baseline and state-of-the-art schemes, with simulations showing significant improvements in D2D area spectral efficiency.
• The optimal SIR threshold $ p_s^* $ is derived using the Lambert W function, resulting in $ p_s^* riangleq rac{ ext{W}ig(eta^{2/eta} ho ig)}{ ho} $, where $ ho $ depends on network densities and path loss.
• The optimal guard zone radius $ ho^* $ is determined by balancing D2D throughput gain against the loss in spatial reuse, with simulations showing that larger guard zones improve cellular coverage at the cost of reduced D2D access opportunities.
• The analytical coverage probability approximations for both D2D and cellular links are shown to be highly accurate, with errors less than $ 10^{-5} $, validating the stochastic geometry model.
• The decoupled optimization approach enables a tractable solution to the joint optimization of $ p_s^* $ and $ ho^* $, making the scheme suitable for real-time decentralized deployment.
• Even when the primary objective is to maximize D2D ASE under cellular coverage constraints, the optimal parameters $ p_s^* $ and $ ho^* $ also improve the average sum rate of the D2D network.