Joint user association and power allocation in ultra-dense mmWave networks: a multi-connectivity approach

TL;DR

This paper proposes a joint optimization framework for user association and power allocation in ultra-dense mmWave networks, leveraging multi-connectivity to improve individual user rates amidst severe signal blockages and high handover rates.

Contribution

It introduces a novel multi-connectivity enabled user association and power allocation method, formulated as a non-convex optimization problem, and develops an iterative algorithm to solve it effectively.

Findings

The proposed algorithm improves downlink rates for users.

Numerical results validate the effectiveness of the joint optimization approach.

Abstract

In ultra-dense millimeter wave (mmWave) networks, mmWave signals suffer from severe path losses and are easily blocked by obstacles. Meanwhile, ultra-dense deployment causes excessive handovers, which reduces the data link reliability. To alleviate the above issues, the novel technology, known as multi-connectivity enabled user association (MCUA) is incorporated in this letter. We aim to jointly optimize MCUAs and downlink (DL) power allocations (PAs) to maximize the DL rate of each user simultaneously, rather than total. This is a non-convex nonlinear 0-1 mixed integer multi-objective optimization problem and quite complicated. To solve it, we first use the weighted sum method to scalarize it as a single-objective optimization problem (SOOP), and then relax the binary association variables to real ones. Considering that the relaxed SOOP is still non-convex, we perform a series of transformations upon it and make it a differential of convex programming. Finally, we develop an iterative algorithm based on the convex-concave procedure to solve the SOOP. Numerical results are presented to demonstrate the effectiveness of the proposed algorithms.