Pulse Shaping Diversity to Enhance Throughput in Ultra-Dense Small Cell Networks

TL;DR

This paper introduces pulse shaping diversity through distinct transmit filters at adjacent antennas to improve MIMO channel conditions, significantly enhancing SINR and throughput in ultra-dense small cell networks.

Contribution

It proposes a novel pulse shaping technique with relative offsets at antennas and a fractional equalizer to mitigate ISI, improving MIMO performance in dense networks.

Findings

Median SINR improved by 11 dB

Throughput doubled at the UE

Effective in 2x2 MIMO with 50 m ISD

Abstract

Spatial multiplexing (SM) gains in multiple input multiple output (MIMO) cellular networks are limited when used in combination with ultra-dense small cell networks. This limitation is due to large spatial correlation among channel pairs. More specifically, it is due to i) line-of-sight (LOS) communication between user equipment (UE) and base station (BS) and ii) in-sufficient spacing between antenna elements. We propose to shape transmit signals at adjacent antennas with distinct interpolating filters which introduces pulse shaping diversity eventually leading to improved SINR and throughput at the UEs. In this technique, each antenna transmits its own data stream with a relative offset with respect to adjacent antenna. The delay which must be a fraction of symbol period is interpolated with the pulse shaped signal and generates a virtual MIMO channel that leads to improved diversity and SINR at the receiver. Note that non-integral sampling periods with inter-symbol interference (ISI) should be mitigated at the receiver. For this, we propose to use a fractionally spaced equalizer (FSE) designed based on the minimum mean squared error (MMSE) criterion. Simulation results show that for a 2x2 MIMO and with inter-site-distance (ISD) of 50 m, the median received SINR and throughput at the UE improves by a factor of 11 dB and 2x, respectively, which verifies that pulse shaping can overcome poor SM gains in ultra-dense small cell networks.