CSISensing：面向6G感知的信道状态信息技术

CSISensing 定义了将原始 Wi-Fi 与 5G-Advanced 信道估计转化为经过校准的CSI数据流的信号处理流水线，可达到类雷达级感知精度，支撑无设备手势识别、入侵检测与6G通感一体化系统。

CSISensing Joint Radar-Communication Architecture

ISAC Core: CSI Sensing Engine — IEEE 802.11bf & Subcarrier Phase Extractor

6G ISAC Nodes & Beamforming Subsystems

  • Phase Sanitization Filter
  • CFO / SFO Compensator
  • Spatial Gesture Classifier
  • Doppler Angle-of-Arrival
  • Wi-Fi 7 Multi-Link Channel

Architectural Layers & Waveform Pipelines

  • 第一层：太赫兹与毫米波有源相控阵射频前端 (GaN/InP PA / True-Time-Delay (TTD) / 140 GHz D-Band Active Arrays): Ultra-wideband signal generation, beam squint mitigation, and simultaneous transmission of communication and radar probing waves.
  • 第二层：通感一体化双功能波形与基带DSP引擎 (OTFS / Cyclic Prefix OFDM / Ambiguity Function Shaping / 2D-FFT): Orthogonal modulation in delay-Doppler domains, low peak-to-average power ratio coding, and deterministic clutter cancellation.
  • 第三层：通感融合特征提取与4D雷达点云解算 (Range-Doppler Matrix / MUSIC & ESPRIT DOA / CFAR Detection): High-resolution angle of arrival estimation, velocity extraction, ghost artifact elimination, and 4D spatial coordinate synthesis.
  • 第四层：分布式多基地协同感知网与6G感存算融合服务 (3GPP Rel-19 SBA / Sensing Function (SF) / NEF APIs / RIS Control): Cooperative radar data fusion across distributed cells, intelligent surface reflection control, and exposure of real-time environmental digital twins.

Sensing Telemetry & Benchmarks: <2.5 cm (距离分辨率) | ±0.05 m/s (速度测量精度) | 0.1° (测角精度（DOA）) | 100 Hz (感知刷新率)

CSISensing：从原始信道估计到校准后CSI数据的转化路径

消费级 Wi-Fi 与 5G-Advanced 射频芯片可输出逐子载波的信道估计数据，经相位噪声与时钟偏移的净化处理后，即可成为一种高分辨率感知手段。CSISensing 明确了将此类CSI数据转化为可靠的类雷达级测量所需的校准方法、特征提取流程与标准对齐要求，且无需专用雷达硬件。

Core Engineering Areas

Technical Publications