Interleaved Training and Training-Based Transmission Design for Hybrid Massive Antenna Downlink

TL;DR

This paper introduces an interleaved training and transmission design for hybrid massive antenna systems that reduces training overhead while maintaining outage performance, applicable to both single-user and multi-user scenarios.

Contribution

It proposes a novel interleaved training scheme that adapts training length to channel conditions and jointly optimizes beam assignment and data transmission for multi-user systems.

Findings

Achieves outage performance comparable to full-training schemes.

Significantly reduces training overhead in both SU and MU systems.

Provides analytical expressions for training length and outage probability.

Abstract

In this paper, we study the beam-based training design jointly with the transmission design for hybrid massive antenna single-user (SU) and multiple-user (MU) systems where outage probability is adopted as the performance measure. For SU systems, we propose an interleaved training design to concatenate the feedback and training procedures, thus making the training length adaptive to the channel realization. Exact analytical expressions are derived for the average training length and the outage probability of the proposed interleaved training. For MU systems, we propose a joint design for the beam-based interleaved training, beam assignment, and MU data transmissions. Two solutions for the beam assignment are provided with different complexity-performance tradeoff. Analytical results and simulations show that for both SU and MU systems, the proposed joint training and transmission designs achieve the same outage performance as the traditional full-training scheme but with significant saving in the training overhead.