Pith. sign in

REVIEW 2 cited by

MambaJSCC: Adaptive Deep Joint Source-Channel Coding with Generalized State Space Model

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2409.16592 v1 pith:7INVTEBC submitted 2024-09-25 cs.IT cs.AIcs.LGmath.IT

MambaJSCC: Adaptive Deep Joint Source-Channel Coding with Generalized State Space Model

classification cs.IT cs.AIcs.LGmath.IT
keywords stategssmchannelmambajsccadaptationcomputationalgeneralizedjscc
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Lightweight and efficient neural network models for deep joint source-channel coding (JSCC) are crucial for semantic communications. In this paper, we propose a novel JSCC architecture, named MambaJSCC, that achieves state-of-the-art performance with low computational and parameter overhead. MambaJSCC utilizes the visual state space model with channel adaptation (VSSM-CA) blocks as its backbone for transmitting images over wireless channels, where the VSSM-CA primarily consists of the generalized state space models (GSSM) and the zero-parameter, zero-computational channel adaptation method (CSI-ReST). We design the GSSM module, leveraging reversible matrix transformations to express generalized scan expanding operations, and theoretically prove that two GSSM modules can effectively capture global information. We discover that GSSM inherently possesses the ability to adapt to channels, a form of endogenous intelligence. Based on this, we design the CSI-ReST method, which injects channel state information (CSI) into the initial state of GSSM to utilize its native response, and into the residual state to mitigate CSI forgetting, enabling effective channel adaptation without introducing additional computational and parameter overhead. Experimental results show that MambaJSCC not only outperforms existing JSCC methods (e.g., SwinJSCC) across various scenarios but also significantly reduces parameter size, computational overhead, and inference delay.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Feature Importance-Aware Deep Joint Source-Channel Coding for Computationally Efficient and Adjustable Image Transmission

    cs.IT 2025-04 accept novelty 7.0

    FAJSCC is a new deepJSCC architecture for images that achieves better transmission performance with lower complexity than prior models and enables independent encoder-decoder compute adjustment.

  2. Mixture of Semantics Transmission for Generative AI-Enabled Semantic Communication Systems

    cs.IT 2025-09 conditional novelty 5.0

    A semantic communication system that transmits image ROIs at high fidelity and transmits the background as text, reconstructing the image with a diffusion model.