Portrait
FIZLab
Dept. of Information and Communication Engineering
Hanbat National University

🤖 Multimodal Large Language Model (MLLM) for Wireless Communication Networks

Closed-loop wireless intelligence with Task Expert ModelsConceptual research architecture. Synthetic photorealistic 3D illustrations depict a physical wireless environment with a base station illuminating a car and two smartphone users, represented only by phones; Camera, a physical radar dish antenna, and RF sensing; a real-time Digital Twin with geometry and channel state; a robot representing the Multimodal LLM for context-aware reasoning; autonomous optimization problem formulation; feasibility checks of domains, constraints, optimality and simulator dry-run; four neural-brain Task Expert Models, each solving its own specialized problem; diverse wireless task solutions illustrated by beamforming, channel estimation, scheduling and localization; network evaluation; and state feedback to the Digital Twin. E1, E2, E3 and E4 denote separate expert models with individual problem inputs and solution outputs. Infeasible formulations return to the large model for revision. Natural-language evaluation feedback supports learning and adaptation of the expert models. The animation illustrates a proposed workflow, not measured latency or performance.

Related Publications:

  • Robust Transmission of Punctured Text with Large Language Model-based Recovery, IEEE Transactions on Vehicular Technology (TVT), vol. 75, no. 1, pp. 1737-1742, Jan. 2026
  • Large Multimodal Model-Empowered Task-Oriented Autonomous Communications: Design Methodology and Implementation Challenges, IEEE Vehicular Technology Magazine (VTM), accepted
  • Adaptive Resource Allocation Optimization Using Large Language Models in Dynamic Wireless Environments, IEEE Transactions on Vehicular Technology (TVT), vol. 74, no. 10, pp. 16630-16635, Oct. 2025

📡 AI Communication

01 / INPUTS

Sensing + network state

Sensing outputsTargets · range · motionChannel stateCSI · channel measurementsData & trafficLoad · demandNetwork contextTopology · interferenceMobility & locationPosition · trajectoryDevice informationHardware · capabilities
02 / INTELLIGENCE

AI-aided communication

AI enginePerceptionReasoningDecisionActionLearningSense → learn → adapt
03 / CONTROL

Adaptive wireless tasks

Channel predictionResource allocationAdaptive beamformingInterference controlMobility & handoverQoE-aware control
↑ Sensing feedbackWaveform · beam · resources ↓
CONNECTED APPLICATIONS

One intelligence layer. Diverse wireless environments.

Massive IoT

Industry · connected devices

RAN & edge cloud

Access · distributed compute

UAV networks

Aerial sensing · coverage

Connected vehicles

V2X · cooperative mobility

Biosensing

Respiration · heart rate

Localization & tracking

Position · trajectory

Communication Sensing & intelligence

Related Publications:

  • Group-wise Semantic Splitting Multiple Access for Multi-User Semantic Communication, IEEE Wireless Communications Letters (WCL), vol. 15, pp. 5194-5198, 2026
  • Multiple Active STAR-RIS-Assisted Secure Integrated Sensing and Communication via Cooperative Beamforming, IEEE Internet of Things Journal (IoTJ), accepted
  • DCFNet: Doppler Correction Filter Network for Integrated Sensing and Communication in Multi-User MIMO-OFDM Systems, IEEE Transactions on Wireless Communications (TWC), doi: 10.1109/TWC.2026.3687605

🎯 Deep Learning-based Sensing

Related Publications:

  • ALERT Open Dataset and Input-Size-Agnostic Vision Transformer for Driver Activity Recognition Using IR-UWB, IEEE Access, vol. 14, pp. 28654-28674, 2026
  • Leveraging Multiple PRF Radar for Target Detection and Sea Clutter Suppression with Deep Learning Network, Proc. International Conference on ICT Convergence (ICTC)