# Beam Entropy of 5G Cellular Millimetre-Wave Channels

## TL;DR

This paper analyzes beam entropy in 5G mm-wave channels across various scenarios and frequencies, highlighting its implications for beam training and secure communications in future cellular systems.

## Contribution

It provides the first comprehensive study of beam entropy values for 5G mm-wave channels using NYUSIM across multiple scenarios and frequencies.

## Key findings

- Lower beam entropy channels are suitable for memory-assisted beam training.
- High beam entropy channels can enhance secure key generation.
- Beam entropy varies significantly across scenarios and frequencies.

## Abstract

In this paper, we obtain and study typical beam entropy values for millimetre wave (mm-wave) channel models using the NYUSIM simulator for frequencies up to 100 GHz for fifth generation (5G) and beyond 5G cellular communication systems. The beam entropy is used to quantify sparse MIMO channel randomness in beamspace. Lower relative beam entropy channels are suitable for memory-assisted statistically-ranked (MarS) and hybrid radio frequency (RF) beam training algorithms. High beam entropies can potentially be advantageous for low overhead secured radio communications by generating cryptographic keys based on channel randomness in beamspace, especially for sparse multiple input multiple output (MIMO) channels. Urban micro (UMi), urban macro (UMa) and rural macro (RMa) cellular scenarios have been investigated in this work for 28, 60, 73 and 100 GHz.