Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Enabling ISAC in Real World: Beam-Based User Identification with Machine Learning

Umut Demirhan, Ahmed Alkhateeb

TL;DR

This letter formulates this user identification problem and develops two solutions, a baseline model-based solution that maps the angle of the object from the radar scene to communication beams and a scalable deep learning solution that is agnostic to the number of candidate objects.

Abstract

Leveraging perception from radar data can assist multiple communication tasks, especially in highly-mobile and large-scale MIMO systems. One particular challenge, however, is how to distinguish the communication user (object) from the other mobile objects in the sensing scene. This paper formulates this \textit{user identification} problem and develops two solutions, a baseline model-based solution that maps the objects angles from the radar scene to communication beams and a scalable deep learning solution that is agnostic to the number of candidate objects. Using the DeepSense 6G dataset, which have real-world measurements, the developed deep learning approach achieves more than $93.4\%$ communication user identification accuracy, highlighting a promising path for enabling integrated radar-communication applications in the real world.

Enabling ISAC in Real World: Beam-Based User Identification with Machine Learning

TL;DR

This letter formulates this user identification problem and develops two solutions, a baseline model-based solution that maps the angle of the object from the radar scene to communication beams and a scalable deep learning solution that is agnostic to the number of candidate objects.

Abstract

Leveraging perception from radar data can assist multiple communication tasks, especially in highly-mobile and large-scale MIMO systems. One particular challenge, however, is how to distinguish the communication user (object) from the other mobile objects in the sensing scene. This paper formulates this \textit{user identification} problem and develops two solutions, a baseline model-based solution that maps the objects angles from the radar scene to communication beams and a scalable deep learning solution that is agnostic to the number of candidate objects. Using the DeepSense 6G dataset, which have real-world measurements, the developed deep learning approach achieves more than communication user identification accuracy, highlighting a promising path for enabling integrated radar-communication applications in the real world.

Paper Structure

This paper contains 9 sections, 10 equations, 4 figures, 1 table.

Table of Contents

  1. Introduction
  2. System Model
  3. Communication Model
  4. Radar Model
  5. Problem Definition: User Identification with Beam Information
  6. Proposed Solutions
  7. Dataset
  8. Results
  9. Conclusion

Figures (4)

  • Figure 1: The system model, where a radar-equipped mmWave basestation communicates with a user, while radar data to aid the communications is collected from all the available targets in the environment.
  • Figure 2: The proposed DNN solution. Along with the communication beam index, the radar estimated state of each user is respectively fed to the DNN. The resulting outputs are collected together to select the communication target.
  • Figure 3: A sample from the dataset is shown with the corresponding camera image and range-Doppler map. In this sample, the beam with the index $28$ corresponds to the optimal beam. The targets are determined using this visual, based on a sequence of samples by tracking through time.
  • Figure 4: The visualization of the data samples and the baseline solutions. Most of the samples follow a linear pattern with a relatively large variance.