Robust joint user association and resource partitioning in heterogeneous cloud RANs with dual connectivity

Abstract

Dual connectivity (DC) allows users to connect to both a macrocell and a small cell simultaneously and is introduced to combat constrained fronthaul links while improving throughput and user mobility. In this paper, we investigate the problem of joint user association and resource allocation in heterogeneous cloud radio access networks (H-CRAN) with DC. The problem is formulated as a non-convex mixed integer non-linear program with constraints on non-ideal fronthaul links and quality of service (QoS) requirements. We propose a block coordinate descent (BCD)-based algorithm that successively solves associated subproblems and prove its convergence. We also utilize successive convex approximation (SCA) to deal with binary association variables. Moreover, we discuss a realistic case in which imperfect knowledge of channel state information causes the user’s rate to be uncertain at the time of decision making. Then, we extend our model to a robust optimization problem by introducing protection functions for constraints. We introduce parameters to address the trade-off between robustness and performance and derive probability bounds of constraint violation. Our numerical results validate the effectiveness of both our nominal and robust algorithms in uncertain environments, confirm the convergence of our proposed algorithms, and examine the cost of robustness.