# Measurement Grid Optimization for OTA Testing of 5G Smart Watches

## TL;DR

This paper optimizes measurement grids for 5G smart watch OTA testing to significantly reduce test time while maintaining acceptable accuracy.

## Contribution

A novel MU-based grid optimization method for 5G smart watch OTA testing is proposed, reducing test time by up to 75%.

## Key findings

- Reducing grid points from 62 to 26 saves 60% test time with 0.1 dB uncertainty increase for TIS and TRP.
- Further reducing grid points from 266 to 62 saves 75% test time with the same uncertainty increase.
- Experimental verification confirms the feasibility of the proposed grid optimization method.

## Abstract

Over-the-air (OTA) testing is crucial for optimizing wireless performance of 5G smart watches and improving their user experience. However, the current required test time is so long that it is almost impossible to complete the entire OTA testing without recharging and repositioning, which is unacceptable for the industry. Therefore, test-time reduction is significant. The objective of this work is to optimize measurement grids for OTA testing of 5G smart watches, which balance accuracy with efficiency. In this research, passive patterns from a typical 5G commercial smart watch are measured at different bands as reference patterns, which represent general radiation properties of 5G commercial smart watches. The effect of various coarse grids on OTA testing precision is characterized quantitatively by analyzing their accuracy in reconstructing reference patterns. The related measurement uncertainty (MU) terms are then evaluated and determined quantitatively based on statistical analysis. According to the derived MU limits for grid configurations, reducing grid points from currently required 62 (30/30) to 26 (45/45), and from 266 (15/15) to 62 (30/30) could save roughly 60% and 75% of the test time, respectively, with an uncertainty increase of 0.1 dB for both Total Isotropic Sensitivity (TIS) and Total Radiated Power (TRP) testing, which is considered acceptable. Furthermore, the feasibility of the proposed MU analysis and recommended grids have been experimentally verified.