The Device-to-Device Reuse Maximization Problem With Power Control

Abstract

We investigate the problem of how to maximize the frequency reuse factor in a cellular network with device-to-device links and dynamic power control. Recent works in the literature mostly assume fixed powers, which we find to enable only around 70% of achievable reuse, or limit the reuse factor a priori . The problem is formulated as a mixed-integer nonlinear problem and identified as non-deterministic polynomial-time complete. Nevertheless, we present a full analysis of the involved effects and identify the main driver of complexity, that a priori , the effect of binary link activation and deactivation on the spectral radius of the relative gain matrix cannot be properly bounded. We identify a core parameter that we call relative interference pressure and that reflects the severity of interference caused by a link with dynamic power control. Based on relative interference pressure, we propose two solution algorithms, the worst pressure shutdown and the $\tilde {\varepsilon }$ Pressure Packing algorithm. Both algorithms are thoroughly evaluated with simulations and compared with state-of-the-art methods. The proposed algorithms maximize reuse nearly optimally, enabling the remaining 30% of frequency reuse.