Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Extended-Target Classification and Localization for Near-Field ISAC

Zongyao Zhao, Zhaolin Wang, Lincong Han, Jing Jin, Yuanwei Liu, Kaibin Huang

Abstract

Near-field integrated sensing and communication (ISAC) enables object-level sensing from distance-dependent array responses, yet most existing near-field methods still rely on point-target models and realistic extended targets remain largely unexplored. In this paper, joint target classification and range-azimuth localization are studied from channel responses of realistic extended targets. A dual-branch inference framework is proposed. Semantic and geometric branches are used for classification and localization, respectively. Cross-task attention is introduced after task-specific encoding so that complementary cues can be exchanged without forcing full feature sharing from the input stage. To improve localization on the same backbone, uncertainty-aware regression and a physics-guided structured objective are adopted, including planar consistency, peak-response regularization, and geometry-coupling constraints. Training and evaluation data are generated from full-wave electromagnetic scattering simulations of voxelized vehicle targets with randomized heading angles, material contrasts, and placements. The compared variants show that cross-task attention mainly benefits classification, while uncertainty-aware and structured supervision are needed to recover strong localization performance on the same backbone. Under the adopted shared-OFDM benchmark, the proposed framework reaches the best joint operating point with fewer sensing tones for the same target performance region.

Extended-Target Classification and Localization for Near-Field ISAC

Abstract

Near-field integrated sensing and communication (ISAC) enables object-level sensing from distance-dependent array responses, yet most existing near-field methods still rely on point-target models and realistic extended targets remain largely unexplored. In this paper, joint target classification and range-azimuth localization are studied from channel responses of realistic extended targets. A dual-branch inference framework is proposed. Semantic and geometric branches are used for classification and localization, respectively. Cross-task attention is introduced after task-specific encoding so that complementary cues can be exchanged without forcing full feature sharing from the input stage. To improve localization on the same backbone, uncertainty-aware regression and a physics-guided structured objective are adopted, including planar consistency, peak-response regularization, and geometry-coupling constraints. Training and evaluation data are generated from full-wave electromagnetic scattering simulations of voxelized vehicle targets with randomized heading angles, material contrasts, and placements. The compared variants show that cross-task attention mainly benefits classification, while uncertainty-aware and structured supervision are needed to recover strong localization performance on the same backbone. Under the adopted shared-OFDM benchmark, the proposed framework reaches the best joint operating point with fewer sensing tones for the same target performance region.
Paper Structure (26 sections, 46 equations, 9 figures, 4 tables)

This paper contains 26 sections, 46 equations, 9 figures, 4 tables.

Table of Contents

  1. Introduction
  2. System Model and Problem Formulation
  3. Near-Field Extended-Target Scattering Model
  4. Dipole Excitation and Incident Field
  5. Volume Integral Equation and Total-Field Transfer
  6. Receiver Model and Channel Matrix Construction
  7. Channel Tensor Representation
  8. Communication Model and Resource Partition
  9. Joint Classification and Localization Task
  10. Proposed Joint Learning Framework
  11. Overall Architecture and Dual-Branch Feature Extraction
  12. Cross-Task Attention for Semantic-Geometric Interaction
  13. Uncertainty-Aware Task Supervision
  14. Physics-Guided Structured Regularization and Optimization
  15. Numerical Results
  16. ...and 11 more sections

Figures (9)

  • Figure 1: Illustration of near-field point-target and extended-target sensing.
  • Figure 2: Illustration of the near-field extended-target sensing scenario.
  • Figure 3: Overall system framework of the proposed method.
  • Figure 4: Joint-learning gain map relative to the single-task references.
  • Figure 5: Performance versus the number of selected subcarriers $K_s$.
  • ...and 4 more figures

Theorems & Definitions (1)

  • Remark 1