aiqbench analysis
This research presents an uncertainty-aware multi-task learning framework that jointly optimizes modulation recognition and SINR estimation, addressing the distinct error characteristics of each task through shared representations and task-specific adapters. By integrating classification entropy with predicted regression variance, the model enables selective inference that significantly improves accuracy in both matched and unseen channel conditions while reducing signal-to-interference-plus-noise ratio errors. The approach demonstrates substantial performance gains over conventional multi-task learning methods, offering a robust solution for intelligent receivers operating under noisy or mismatched environments.
Uncertainty-Aware Multi-Task Learning for Joint Modulation Recognition and SINR Estimation
arXiv:2608.28865v1 Announce Type: cross Abstract: Joint modulation recognition and signal-to-interference-plus-noise ratio (SINR) estimation can reduce duplicated processing in intelligent receivers, but the two tasks have different uncertainty characteristics. This letter proposes an uncertainty-aware multi-task model that transforms each short normalized in-phase/quadrature window into 36 deterministic, label-free statistics, learns a shared representation, and uses task-specific adapters for modulation classification and heteroscedastic SINR regression. A joint uncertainty score combines classification entropy and predicted regression variance to support selective inference. Simulations cover QPSK, 8PSK, 16QAM, and 64QAM under matched additive white Gaussian noise/Rayleigh channels and an unseen frequency-selective Rician channel. Over five independent seeds, the proposed model improves matched and unseen-channel accuracy over conventional multi-task learning by 14.86 and 8.61 percentage points, respectively, while reducing SINR mean absolute error by 1.60 and 1.61 dB. Confidence-based rejection further lowers modulation error under channel mismatch.
Abstract:Joint modulation recognition and signal-to-interference-plus-noise ratio (SINR) estimation can reduce duplicated processing in intelligent receivers, but the two tasks have different uncertainty characteristics. This letter proposes an uncertainty-aware multi-task model that transforms each short normalized in-phase/quadrature window into 36 deterministic, label-free statistics, learns a shared representation, and uses task-specific adapters for modulation classification and heteroscedastic SINR regression. A joint uncertainty score combines classification entropy and predicted regression variance to support selective inference. Simulations cover QPSK, 8PSK, 16QAM, and 64QAM under matched additive white Gaussian noise/Rayleigh channels and an unseen frequency-selective Rician channel. Over five independent seeds, the proposed model improves matched and unseen-channel accuracy over conventional multi-task learning by 14.86 and 8.61 percentage points, respectively, while reducing SINR mean absolute error by 1.60 and 1.61 dB. Confidence-based rejection further lowers modulation error under channel mismatch.
Networking and Internet Architecture (cs.NI); Machine Learning (cs.LG)
arXiv:2608.28865 [cs.NI]
(or arXiv:2608.28865v1 [cs.NI] for this version)
https://doi.org/10.48550/arXiv.2608.28865 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 234 words
Related coverage
AMN: An Adaptive Multi-Scale Fusion Network with Boundary and Uncertainty Modeling for Nuclei Segmentation
arXiv · Jun 9, 2026
FMRFusion: Frequency-Aware Multi-View Representation Learning for Heterogeneous Image Fusion
arXiv · Jun 9, 2026
A Multi-Modal Generative Model for Tomato Disease Leaves Understanding
arXiv · Sep 18, 2026
A Distributed Multi-UGV Exploration Framework With Loop-Aware Planning and Descriptor-Aided Localization in Resource-Limited Environments
arXiv · Jun 10, 2026