[Paper Review] Preliminary Results on 3D Channel Modeling: From Theory to Standardization

Abla Kammoun, Hajer Khanfir+3 more|arXiv (Cornell University)|Dec 1, 2013

Millimeter-Wave Propagation and Modeling15 references4 citations

TL;DR

This paper presents the ongoing 3GPP standardization effort for 3D channel modeling to support elevation beamforming and Full Dimensional MIMO (FD-MIMO) in LTE. It introduces a 3D channel model that accounts for vertical angular spread and UE height, demonstrating improved geometry factor and interference suppression over 2D models, particularly with optimal downtilt angles of 9°–12° depending on antenna spacing.

ABSTRACT

Three dimensional beamforming (3D) (also elevation beamforming) is now gaining a growing interest among researchers in wireless communication. The reason can be attributed to its potential to enable a variety of strategies like sector or user specific elevation beamforming and cell-splitting. Since these techniques cannot be directly supported by current LTE releases, the 3GPP is now working on defining the required technical specifications. In particular, a large effort is currently made to get accurate 3D channel models that support the elevation dimension. This step is necessary as it will evaluate the potential of 3D and FD(Full Dimensional) beamforming techniques to benefit from the richness of real channels. This work aims at presenting the on-going 3GPP study item "Study on 3D-channel model for Elevation Beamforming and FD-MIMO studies for LTE", and positioning it with respect to previous standardization works.

Motivation & Objective

• To bridge the gap between theoretical 3D channel modeling and 3GPP standardization for elevation beamforming.
• To present the current status and technical direction of the 3GPP TSG-RAN-WG1 study item on 3D channel modeling.
• To evaluate the performance of the proposed 3D channel model against existing 2D models using key metrics like geometry factor and coupling gain.
• To support accurate system-level evaluation of 3D beamforming and FD-MIMO by providing a standardized, realistic 3D channel model.

Proposed method

• The paper outlines the 3GPP 3D channel model framework, including large-scale parameters such as azimuth and elevation angle spreads, path loss, and delay spread.
• It defines key performance metrics: coupling gain (CLm) as the difference between RSRP of the strongest cell and transmit power, and geometry factor (GF) as the ratio of strongest RSRP to sum of all other RSRPs.
• The model incorporates 3D UE deployment with varying heights, enabling realistic inter-cell interference and beamforming gain evaluation.
• Phase 1 calibration uses slow fading only, while Phase 2 and 3 include fast fading and require CDFs of angular spreads, eigenvalues, and system-level performance metrics.
• Simulations compare the 3D model with the 2D ITU model, using CDFs of geometry factor under different vertical antenna spacings (0.5λ and 0.8λ).
• Optimal downtilt angles (9°–12°) are determined based on trade-offs between coupling gain and interference reduction.

Experimental results

Research questions

• RQ1 How does the 3D channel model improve interference management and beamforming gain compared to 2D models in urban macro (UMa) scenarios?
• RQ2 What is the optimal downtilt angle for 3D beamforming when vertical antenna spacing is 0.5λ or 0.8λ?
• RQ3 How do large-scale parameters such as azimuth and elevation angular spreads affect system-level performance in 3D beamforming?
• RQ4 What are the key calibration requirements for 3D channel models in Phase 2 and Phase 3 of the 3GPP standardization process?
• RQ5 To what extent does 3D UE deployment reduce inter-cell interference compared to 2D models?

Key findings

• The 3D channel model achieves better geometry factor performance than the 2D ITU model, especially due to reduced inter-cell interference from UEs on higher floors.
• For vertical spacing of 0.5λ, a 12° downtilt achieves optimal performance, balancing coupling gain and interference reduction.
• For 0.8λ spacing, a 9° downtilt yields the best performance, as higher downtilt angles degrade coupling gain due to beam narrowing.
• The 3D model shows significant gains in geometry factor, indicating improved signal-to-interference ratio and beamforming efficiency.
• Phase 2 and 3 calibrations require CDFs of angular spreads, eigenvalues, and system-level metrics like cell edge throughput and spectrum efficiency.
• The 3D model enables more accurate evaluation of FD-MIMO and 3D beamforming techniques by incorporating realistic 3D propagation effects and UE height variations.