# MambaCSP: Hybrid-Attention State Space Models for Hardware-Efficient Channel State Prediction

arXiv: https://arxiv.org/abs/2604.21957 · HTML: https://mlsanity.com/paper/2604.21957/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Core SSM architectures for language, one of 10 directions in State-space sequence models (S4, Mamba).
- Builds on 8 of its 14 references that are in the map.
- First posted on arXiv 23 Apr 2026.

## Placed in

- State-space sequence models (S4, Mamba) / [Core SSM architectures for language](https://mlsanity.com/ssm/core-ssm-architectures-for-language/)

## Builds on (most cited first, 8 of 8)

- [Mamba: Linear-Time Sequence Modeling with Selective State Spaces](https://mlsanity.com/paper/2312.00752/) (2023)
- [gpt-oss-120b & gpt-oss-20b Model Card](https://mlsanity.com/paper/2508.10925/) (2025)
- [LLM4CP: Adapting Large Language Models for Channel Prediction](https://mlsanity.com/paper/2406.14440/) (2024)
- [Csi-LLM: A Novel Downlink Channel Prediction Method Aligned with LLM Pre-Training](https://mlsanity.com/paper/2409.00005/) (2024)
- [Exploring the Potential of Large Language Models for Massive MIMO CSI Feedback](https://mlsanity.com/paper/2501.10630/) (2025)
- [Large AI Models for Wireless Physical Layer](https://mlsanity.com/paper/2508.02314/) (2025)
- [Deep UL2DL: Channel Knowledge Transfer from Uplink to Downlink](https://mlsanity.com/paper/1812.07518/) (2018)
- [Joint Partitioning and Placement of Foundation Models for Real-Time Edge AI](https://mlsanity.com/paper/2512.01039/) (2025)

## Built on by (most cited first, 1 of 1)

- [C-MambaPose: A Physics-Informed Complex Mamba Framework for Cross-Environment WiFi Human Pose Estimation](https://mlsanity.com/paper/2606.13700/) (2026)

## Abstract (arXiv)

Recent works have demonstrated that attention-based transformer and large language model (LLM) architectures can achieve strong channel state prediction (CSP) performance by capturing long-range temporal dependencies across channel state information (CSI) sequences. However, these models suffer from quadratic scaling in sequence length, leading to substantial computational cost, memory consumption, and inference latency, which limits their applicability in real-time and resource-constrained wireless deployments. In this paper, we investigate whether selective state space models (SSMs) can serve as a hardware-efficient alternative for CSI prediction. We propose MambaCSP, a hybrid-attention SSM architecture that replaces LLM-based prediction backbones with a linear-time Mamba model. To overcome the local-only dependencies of pure SSMs, we introduce lightweight patch-mixer attention layers that periodically inject cross-token attentions, helping with long-context CSI prediction. Extensive MISO-OFDM simulations show that MambaCSP improves prediction accuracy over LLM-based approaches by 9-12%, while delivering up to 3.0x higher throughput, 2.6x lower VRAM usage, and 2.9x faster inference. Our results demonstrate that hybrid state space architectures provide a promising direction for scalable and hardware-efficient AI-native CSI prediction in future wireless networks.