User-Centric Joint Transmission in Virtual-Cell-Based Ultra-Dense Networks

TL;DR

This paper evaluates user-centric joint transmission in ultra-dense networks, demonstrating that optimal power allocation with MRT enhances user throughput and signal quality without causing excessive interference.

Contribution

It introduces a power-constrained joint transmission scheme with MRT in UDNs, analyzing its performance using stochastic geometry models and highlighting its advantages over existing methods.

Findings

High SNR achieved with MRT-based JT

MRT precoding is essential for effective JT

User throughput improves with optimal power allocation

Abstract

In ultra-dense networks (UDNs), distributed radio access points (RAPs) are configured into small virtual cells around mobile users for fair and high-throughput services. In this correspondence, we evaluate the performance of user-centric joint transmission (JT) in a UDN with a number of virtual cells. In contrast to existing cooperation schemes, which assume constant RAP transmit power, we consider a total transmit power constraint for each user, and assume that the total power is optimally allocated to the RAPs in each virtual cell using maximum ratio transmission (MRT). Based on stochastic geometry models of the RAP and user locations, we resolve the correlation of transmit powers introduced by MRT and derive the average user throughput. Numerical results show that user-centric JT with MRT provides a high signal-to-noise ratio (SNR) without generating severe interference to other co-channel users. Moreover, we show that MRT precoding, while requiring channel-state-information (CSI), is essential for the success of JT.