Analysis and Optimization of Probabilistic Caching in Micro/Millimeter Wave Hybrid Networks With Dual Connectivity

Abstract

Dual connectivity (DC) is a promising solution to boost system capacity and enhance mobility robustness in the fifth generation networks, allowing mobile users (MUs) to aggregate radio-frequency resources simultaneously from at least two kinds of serving base stations. In this paper, we focus on a novel cache-enabled heterogeneous network, where MUs can flexibly schedule its traffic between traditional sub-6 GHz ( $\mu $ Wave) macro cells and dense millimeter-wave (mmWave) small cells. We propose a comprehensive framework of mathematical models for the $\mu $ Wave and mmWave hybrid caching network based on stochastic geometry, and then derive general and tractable expressions of the average successful delivery probability (SDP) for different networks with reasonable approximations. Furthermore, we optimize the caching probabilities for maximizing the SDP in the hybrid network with DC. The problem is non-convex and we solve it numerically using simulated annealing particle swarm optimization algorithm. At last, the simulation results show the effectiveness of our analytical and optimal conclusions.

Keywords

• Dual connectivity
• hybrid caching networks
• stochastic geometry
• average successful delivery probability
• non-convex optimization