A lightweight beamforming method based on unsupervised MLP

By employing a lightweight beamforming method based on unsupervised MLP, the beamformer and phase shift matrix are optimized, solving the adaptability problem of beamforming in dynamic environments, reducing computational complexity, improving the sensing signal-to-noise ratio, and achieving stable optimization.

CN121333366BActive Publication Date: 2026-06-02XIDIAN UNIV +1
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
XIDIAN UNIV
Filing Date
2025-11-07
Publication Date
2026-06-02

AI Technical Summary

Technical Problem

In existing technologies, beamforming optimization methods are difficult to adapt to dynamically changing communication environments, and neural network-based methods have high computational complexity and difficulty in designing loss functions.

Method used

A lightweight beamforming method based on unsupervised MLP is adopted. By constructing the target optimization problem of communication signal-to-noise ratio and sensing signal-to-noise ratio, the beamformer and phase shift matrix are optimized. The lightweight network of multilayer perceptron is used for training to obtain the phase control vector and generate the target phase shift matrix to maximize the sensing signal-to-noise ratio of the echo signal.

Benefits of technology

It effectively balances communication and sensing performance, reduces computational complexity, and improves the sensing signal-to-noise ratio of echo signals, thereby achieving stable beamforming optimization.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN121333366B_ABST
    Figure CN121333366B_ABST
Patent Text Reader

Abstract

The application discloses a kind of lightweight beamforming methods based on unsupervised MLP, belong to wireless communication field, including: according to the ISAC signal sent by base station, the perception signal-to-noise ratio of communication signal-to-noise ratio and ISAC signal echo signal at target is constructed;According to ISAC signal and communication signal-to-noise ratio, the target optimization problem of perception signal-to-noise ratio is constructed;Solving target optimization problem obtains target beamformer;Channel parameters in communication signal-to-noise ratio and perception signal-to-noise ratio are input into the lightweight network based on multilayer perceptron trained, and phase control vector is obtained;According to phase control vector, target phase shift matrix is obtained;According to target beamformer and target phase shift matrix, the perception signal-to-noise ratio of maximum echo signal is obtained.The application can effectively balance communication and sensing performance, both have lower computational complexity, and have better performance, improve the perception signal-to-noise ratio of echo signal.
Need to check novelty before this filing date? Find Prior Art