Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

JSSAnet: Theory-Guided Subchannel Partitioning and Joint Spatial Attention for Near-Field Channel Estimation

Zhiming Zhu, Shu Xu, Chunguo Li, Yongming Huang, Luxi Yang

Abstract

The deployment of extremely large-scale antenna array (ELAA) in sixth-generation (6G) communication systems introduces unique challenges for efficient near-field channel estimation. To tackle these issues, this paper presents a theory-guided approach that incorporates angular information into an attention-based estimation framework. A piecewise Fourier representation is proposed to implicitly encode the near-field channel's inherent nonlinearity, enabling the entire channel to be segmented into multiple subchannels, each mapped to the angular domain via the discrete Fourier transform (DFT). Then, we develop a joint subchannel-spatial-attention network (JSSAnet) to extract the spatial features of both intra- and inter-subchannels. To guide theoretically the design of the joint attention mechanism, we derive upper and lower bounds based on approximation criteria and DFT quantization loss mitigation, respectively. Following by both bounds, a JSSA layer of an attention block is constructed to assign independent and adaptive spatial attention weights to each subchannel in parallel. Subsequently, a feed-forward network (FFN) of an attention block further captures and refines the residual nonlinear dependencies across subchannels. Moreover, the proposed JSSA map is linearly computed via element-wise product combining large-kernel convolutions (DLKC), maintaining strong contextual learning capability. Numerical results verify the effectiveness of embedding sparsity information into the attention network and demonstrate JSSAnet achieves superior estimation performance compared with existing methods.

JSSAnet: Theory-Guided Subchannel Partitioning and Joint Spatial Attention for Near-Field Channel Estimation

Abstract

The deployment of extremely large-scale antenna array (ELAA) in sixth-generation (6G) communication systems introduces unique challenges for efficient near-field channel estimation. To tackle these issues, this paper presents a theory-guided approach that incorporates angular information into an attention-based estimation framework. A piecewise Fourier representation is proposed to implicitly encode the near-field channel's inherent nonlinearity, enabling the entire channel to be segmented into multiple subchannels, each mapped to the angular domain via the discrete Fourier transform (DFT). Then, we develop a joint subchannel-spatial-attention network (JSSAnet) to extract the spatial features of both intra- and inter-subchannels. To guide theoretically the design of the joint attention mechanism, we derive upper and lower bounds based on approximation criteria and DFT quantization loss mitigation, respectively. Following by both bounds, a JSSA layer of an attention block is constructed to assign independent and adaptive spatial attention weights to each subchannel in parallel. Subsequently, a feed-forward network (FFN) of an attention block further captures and refines the residual nonlinear dependencies across subchannels. Moreover, the proposed JSSA map is linearly computed via element-wise product combining large-kernel convolutions (DLKC), maintaining strong contextual learning capability. Numerical results verify the effectiveness of embedding sparsity information into the attention network and demonstrate JSSAnet achieves superior estimation performance compared with existing methods.

Paper Structure

This paper contains 23 sections, 2 theorems, 37 equations, 12 figures, 2 tables.

Table of Contents

  1. introduction
  2. System Model and Problem Formulation
  3. System Model
  4. Channel Model
  5. Near-field Channel Estimation Formulation
  6. Piecewise Linear Approximation
  7. The Piecewise Linear Approximation of the Near-field ARV
  8. Discussion on Number of Subarrays
  9. Joint Subchannel-Spatial-Attention Network for Near-Field Channel Estimation
  10. The Structure Design of JSSAnet
  11. DFT and IDFT Operations
  12. The Implementation of JSSAB
  13. JSSA layer
  14. FFN layer
  15. Implementation of Refine Tail
  16. ...and 8 more sections

Key Result

Theorem 1

If the piecewise Fourier vector can approximately present the near-field ARV, the partitioned number $M$ must satisfy where $r_\text{min}$ is the minimum feasible distance in communication system.

Figures (12)

  • Figure 1: The near-field uplink channel with spherical-wave and with several scatters.
  • Figure 2: The ELAA is partitioned uniformly into $M$ subarrays. The EM waves impinge on the subarray with planar wavefront, while the wavefronts between subarrays exhibit spherical waves.
  • Figure 3: The similarity metric between $\mathbf{b}(\bm{\theta})$ and $\mathbf{a}(\theta_0, r_0)$. The carrier frequency $f_c$, the spatial AoA $\theta_0$ and distance $r_0$ are $60$ GHz, $\sin(\pi/12)$ and $15$ meters, respectively.
  • Figure 4: The normalized power distributions of $\mathbf{c}_{k,\ell}$ for the $\ell$-th path component at $f_k=60$ GHz, where $N_\text{BS}=256$, $\phi_\ell=15 ^\circ$ and $r_\ell=20$ meters. (a) The case of $M=2$; (b) The case of $M=4$.
  • Figure 5: Decomposition of the proposed JSSAnet: Overall Architecture and functional Modules.
  • ...and 7 more figures

Theorems & Definitions (2)

  • Theorem 1
  • Theorem 2