Accurate CSI forecasting is essential for mitigating channel aging and supporting reliable downlink transmission in high-mobility wireless systems. In FDD systems, forecasting future DL CSI from historical UL observations requires joint temporal extrapolation and cross-band inference. However, defining historical UL as the input and future DL as the target does not specify how elapsed time and UL-to-DL band separation should be represented in a pretrained predictor's positional space. FAR-CSI establishes an FDD-aware representation of channel forecasting within a joint temporal and operating-band positional field. Within this field, historical UL observations and future DL targets occupy segregated regions according to their temporal and operating-band roles, preserving the distinction between observed UL context and unobserved DL prediction targets. This representation is instantiated through a unified structured routing rule that governs encoder positional assignment, decoder-field reconstruction, and future-DL target selection, ensuring the same FDD geometry is preserved from contextual encoding to prediction readout. Under matched backbone, initialization, and training conditions, FDD-aware routing reduces NMSE by 21.44% relative to generic routing. Across the evaluated FDD baselines, FAR-CSI achieves higher CSI prediction accuracy and MRT-based spectral efficiency, while retaining its predictive advantage across mobility conditions, prediction horizons, reduced training-data regimes, and noisy historical observations.
Figures & tables
Fig. 1: Overall architecture of FAR-CSI for historical-UL-to-future-DL forecasting. Historical UL CSI is normalized and partitioned into local time-frequency-antenna patches. FDD-aware structured routing then assigns the observed UL tokens and unobserved channel regions to a joint temporal and operating-band positional field. Contextual representations of the observations are inserted into their designated decoder positions, while all remaining positions are mask-initialized. Reconstruction outputs located in the future-DL target region are selected, unpatched, and restored to the physical CSI amplitude scale.
Routing Strategy
NMSE
Generic Routing
0.19134±0.00631
FDD-Aware Routing
0.15025±0.00223
TABLE II: Matched FDD routing ablation on the clean test set. Each run uses its best-validation checkpoint. NMSE values are reported as mean ± standard deviation over three paired seeds.
Fig. 2: FDD NMSE across user velocities from 10 to 100 km/h under clean-input evaluation.
Fig. 3: TDD NMSE across user velocities from 10 to 100 km/h under clean-input evaluation.
Fig. 4: FDD NMSE versus forecast horizon for the four future prediction steps.
Fig. 5: TDD NMSE versus forecast horizon for the four future prediction steps.
Fig. 6: FDD NMSE under different fractions of downstream training data. FAR-CSI and Transformer are evaluated using the indicated training fractions, while LLM4CP is shown only as a separate full-data reference.
Fig. 7: FDD NMSE versus historical-input SNR. Noise is applied only to the observed history, while prediction targets remain clean.
Fig. 8: TDD NMSE versus historical-input SNR. Noise is applied only to the observed history, while prediction targets remain clean.
Accurate channel state information (CSI) prediction is essential for improving the reliability and spectral efficiency of massive MIMO-OFDM systems in high-mobility scenarios. Existing deep learning methods struggle to jointly capture short-term local variations and long-range nonlinear dependencies in CSI sequences. To address this challenge, we propose ChannelKAN, a hybrid CNN-KAN channel prediction model with multi-scale frequency domain information enhancement. The key insight is that CNNs and Kolmogorov-Arnold Networks (KANs) are naturally complementary: CNNs extract intra-time-step local spatial-frequency correlations, while KANs with learnable Chebyshev polynomial activations fit inter-time-step nonlinear temporal evolution in a holistic manner. Specifically, a dual-domain expansion module first generates complementary frequency-domain and delay-domain CSI representations. A multi-scale frequency information enhancement module then retains dominant spectral components at multiple scales to strengthen key features and suppress noise. Next, a CNN-KAN feature extraction module captures local correlations via cascaded convolutions and models long-range dependencies via Chebyshev KAN layers. Finally, a dual-domain fusion module adaptively integrates features from both branches to produce the prediction. Experiments on 3GPP-compliant QuaDRiGa datasets demonstrate that ChannelKAN outperforms RNN, LSTM, GRU, CNN, and Transformer baselines in normalized mean square error (NMSE), spectral efficiency (SE), and bit error rate (BER) across various velocities and signal-to-noise ratios. Ablation studies further confirm the effectiveness of each proposed module.
Nanqing Jiang, Zhangyao Song, Tao Guo +2
School of Cyber Science and Engineering, Southeast University, Nanjing, China
Channel state information (CSI) prediction forecasts future channel states from historical observations, supporting transceiver design and resource allocation with reduced measurement overhead. However, existing deep learning predictors mainly encode channel knowledge in model parameters, which can limit their ability to represent diverse patterns of channel dynamics across scenes. In this paper, we propose RAG-CSI, a retrieval-enhanced Transformer framework that augments prediction with relevant historical channel knowledge. Specifically, we construct a global--local CSI knowledge store containing representative observations across scenes and historical observations from the current serving scene. We further develop joint state--dynamics retrieval (JSDR), which jointly characterizes channel states, phase dynamics, and amplitude dynamics to retrieve relevant CSI sequences. A Transformer predictor then uses cross-attention to fuse the retrieved knowledge with the current CSI history and predict future CSI with limited additional model parameters. Experiments on a multi-scene CSI dataset show overall improvements exceeding 10% across all five evaluation metrics over the best baseline for each metric, including a 15.42% reduction in normalized mean squared error.
Xinyue Jiang, Tong Li, Shiceng Zhang +2
College of Computer Science and Electronic Engineering, Hunan University, Changsha, China. · Xidian University, Xian, China. · Peng Cheng Laboratory, Shenzhen 518066, China.
This paper considers channel state information (CSI) prediction using neural networks for time-varying multi-antenna channels from noisy historical observations. The main challenge is to exploit temporal correlation while distinguishing channel evolution from observation noise. We propose JEPA-kMMVAE, which combines a joint-embedding predictive architecture with a k-memory Markov variational autoencoder to predict future CSI through temporally structured latent representations. The framework supports direct prediction and residual corrections relative to the latest latent state and noisy CSI observation. Simulation results show that the residual configuration outperforms the evaluated baselines at moderate and high input CSI signal-to-noise ratios, while sequence baselines remain more accurate under stronger noise.
Ericsson Research, Stockholm, Sweden · Department of Computer and Information Science, Linköping University, Linköping, Sweden · Department of Electrical Engineering, Linköping University, Linköping, Sweden