A Spectrum Map Construction Method Based on Weighted Dynamic Assignment Tensor Ring Completion

By modeling multi-frequency spectrum data as a three-dimensional spectrum tensor, and utilizing tensor ring decomposition and completion algorithms under different scenarios, the problems of low accuracy and poor robustness in spectrum map construction are solved, and high-precision spectrum map construction under sparse data conditions is achieved.

CN118885957BActive Publication Date: 2026-06-30NAT UNIV OF DEFENSE TECH
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
NAT UNIV OF DEFENSE TECH
Filing Date
2024-07-04
Publication Date
2026-06-30

AI Technical Summary

Technical Problem

Existing technologies suffer from low accuracy and poor robustness in constructing spectrum maps when target frequency spectrum data is sparse in different scenarios.

Method used

A method based on weighted dynamic allocation of tensor rings is adopted to model multi-frequency spectrum data as a three-dimensional spectrum tensor. Low-rank TR factors are extracted through tensor ring decomposition. Different completion algorithms are designed for path loss and shadow fading scenarios. The alternating least squares tensor ring completion algorithm and the tensor ring weighted optimization algorithm are used for completion, and weighted fusion is performed to construct a spectrum map.

Benefits of technology

It improves the robustness and accuracy of the spectrum map under different scenario conditions, and can dynamically adapt to the sparse target frequency spectrum data to accurately construct the spectrum map.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118885957B_ABST
    Figure CN118885957B_ABST
Patent Text Reader

Abstract

This invention relates to the field of electromagnetic spectrum management technology and proposes a spectrum map construction method based on weighted dynamic allocation tensor ring completion. The method includes: modeling multi-frequency spectrum data as a target spectrum tensor; performing tensor ring decomposition on the target spectrum tensor data to extract TR factors, and generating a spectrum map generation model based on the TR factors; establishing different completion algorithms for path loss and shadow fading scenarios to complete the missing spectrum data in the spectrum map generation model; and weighted fusion of the completion results to construct the final spectrum map. This invention employs different algorithms to complete the missing spectrum data in the spectrum map generation model for different scenarios, thereby improving the robustness of the algorithm under different scenario conditions. Furthermore, by weighted fusion of the completion results from the two algorithms to construct the final spectrum map, it can dynamically adapt to two key factors affecting the construction accuracy, thus accurately constructing the spectrum map.
Need to check novelty before this filing date? Find Prior Art

Citation Information

Patent Citations

  • Method and device for controlling adaptive optimization switching for heterogeneous converged wireless network

    CN102427595A

  • Method for constructing radio frequency spectrum map

    CN116112106A