Technical Sessions

DAY 1 (October 21, 2026)

Session 1A: Edge AI and Distributed Learning (10:50 – 12:00)

Session Chair: TBA
Room: TBA

Communication-Efficient In-Network Aggregation for Multi-Job Distributed Deep Learning
Source track: Protocols and Emerging Technologies

Heterogeneity-Aware Subgraph Composition for Fast Decentralized Machine Learning
Source track: Algorithms & Theory

ToCo-Split: Topology-Constrained Communication for Efficient Split Learning in Edge Intelligence
Source track: AI/ML Smart Design

COST: Chain of Speculative Thinking via Communication-Efficient Collaboration over Edge Networks
Source track: AI/ML Smart Design

Session 1B: Vehicular Networking and Perception (10:50 – 12:00)

Session Chair: TBA
Room: TBA

DriveTransGNN-CAN: Stable CAN Transition Graph Neural Network for Vehicle Intrusion Detection
Source track: Security and Privacy

T-RCFusion: Occlusion-Resilient Vehicle Perception in Tunnels With Radar-Camera Fusion
Source track: Systems and Applications

Preventing Safety Monopoly in V2X Open RAN Slicing: A Deterministic, Starvation-Aware Scheduler
Source track: Systems and Applications

HEAVEN: Hybrid Data Sharing for Efficient and Secure Edge-Assisted Vehicular Collaboration
Source track: Systems and Applications

Invited Session 1: Next-Generation Wireless Networks and Radio Systems (13:30 – 14:40)

Session Chair: TBA
Room: TBA

LOCUS: Local Control for Uniformly Scalable Integrated Sensing and Communication Networks
Assigned theme: Next-Generation Wireless Networks and Radio Systems

Reclaiming Stranded Guaranteed Bandwidth for Mobile Core UPFs via In-Network Shadow Pricing
Assigned theme: Next-Generation Wireless Networks and Radio Systems

Waves on the Walls: Empirical Characterization of mmWave Lateral Waves for Enhanced Indoor Coverage
Assigned theme: Next-Generation Wireless Networks and Radio Systems

A Handover Management Approach through Fluid Antenna Systems
Assigned theme: Next-Generation Wireless Networks and Radio Systems

Invited Session 2: Edge Intelligence and Resource-Aware Computing (13:30 – 14:40)

Session Chair: TBA
Room: TBA

Dynamic Distributionally Robust CPU-NPU Scheduling for Vehicular Edge Systems
Assigned theme: Edge Intelligence and Resource-Aware Computing

EdgeVLR: Latency-Aware Routing for VLM Inference in Collaborative Edge Computing
Assigned theme: Edge Intelligence and Resource-Aware Computing

RA-LLMR: Resource-Aware LLM Routing for Maritime Edge-Cloud Inference
Assigned theme: Edge Intelligence and Resource-Aware Computing

Invited Session 3: Sensing for Health and Dynamic Environments (15:10 – 16:05)

Session Chair: TBA
Room: TBA

A Smart-Health Microwave Sensing Node for Non-Invasive Cerebrospinal-Fluid Monitoring
Assigned theme: Sensing for Health and Dynamic Environments

SeCURE: Selective CSI Utilization for Robust Wi-Fi Sensing under Network Dynamics
Assigned theme: Sensing for Health and Dynamic Environments

CoAqua: A Multi-Sonar System for Continuous Swimmer Tracking
Assigned theme: Sensing for Health and Dynamic Environments

Invited Session 4: Aerial, Underwater, and Quantum Networks (15:10 – 16:05)

Session Chair: TBA
Room: TBA

TrOOPER: Transport of Organs and Personnel in Emergency Response using Urban Air Mobility
Assigned theme: Aerial, Underwater, and Quantum Networks

Cost-Optimal Channel Allocation for QoS-Constrained Entanglement Distribution over Heterogeneous Quantum Repeater Chains
Assigned theme: Aerial, Underwater, and Quantum Networks

A Low-Power Underwater Agent Network Stack with Software-Defined Hardware Foundation
Assigned theme: Aerial, Underwater, and Quantum Networks

Short Paper A: Wireless Networks, Security, and Traffic Intelligence (16:10 – 17:30)

Session Chair: TBA

Temporal Split Learning: A Privacy-Preserving IDS Framework for Distributed Medical IoT Networks
Source track: Security and Privacy

LoDge: Localized Privacy-Preserving Data Aggregation
Source track: Security and Privacy

FLight: A Fast and Lightweight Approach for High Accuracy Encrypted Network Traffic Classification
Source track: AI/ML Smart Design

Model Merging for Automatic Modulation Classification in Mobile Networks
Source track: AI/ML Smart Design

Joint BWP Switching and MCS Offset Control for Goodput Maximization in 5G Networks
Source track: Protocols and Emerging Technologies

Semantic-Aware Traffic Shaping for Aerial Edge-Content Delivery Networks
Source track: Protocols and Emerging Technologies

Spatio-Temporal Optimization for Service-Oriented CDNs in LEO Satellite Constellations
Source track: Protocols and Emerging Technologies

TINMAN: Time-Aware Interference-Resilient IoT Multinetworks with Machine Learning
Source track: Systems and Applications

Short Paper B: Sensing, Edge Learning, and Distributed Systems (16:10 – 17:30)

Session Chair: TBA

Passive Polarization Tagging for Multi-User Routing and Micro-Control
Source track: Systems and Applications

RSRW: IMU-Based Chicken Behavior Recognition via Real-Synthetic Reliability Weighting on Edge
Source track: Systems and Applications

Motion-Coupled Sensing: When the State Change Powers Its Own Sensing
Source track: Systems and Applications

AUDibLE: Acoustic-based UAV Detection with Low-Power Embedded Machine Learning
Source track: AI/ML Smart Design

FlickerScope: Vision-Based Transformer Diagnostics via Flicker-Aware Markovian Modeling
Source track: AI/ML Smart Design

Fed-CGLA: A Federated Personalized Learning Framework with Collaboration Awareness
Source track: Systems and Applications

An Optimized Fabric Client Architecture with Dynamic Load Scheduling for Distributed Smart Systems
Source track: Systems and Applications

Revisiting Robotic Sensor Networks: A Budget-Constrained Covering Salesman Approach
Source track: Algorithms & Theory

Poster and Demo Session (17:45 – 19:30)

Session Chair: TBA
Location: Hotel ICON

All 12 papers presented together with the reception. Display numbers are identifiers, not presentation order.

PD01: Achieving Effective XR Interactions via Acoustic Gesture Recognition based on LLMs

PD02: MeshEcho: Online Redundancy Control for Airtime-Constrained LoRa Mesh Networks

PD03: EddyVib: Passive Eddy-Current RF Backscatter for Transformer Vibration Sensing

PD04: Feature Routing for Pipeline DNN Inference in Unreliable UAV Network

PD05: ChargeSafe: An Edge AI System for Real-Time Smartphone Charging Anomaly Detection

PD06: An AI-QMIX-Based Adaptive Weight Optimization Framework for HVAC Model Predictive Control

PD07: Privacy-Preserving On-Device Occupancy Estimation Using Existing CCTV in Convenience Stores

PD08: AIStream: A Live Stream Simulation Training System Based on Mobile Cloud Computing

PD09: A Refined M&V Framework for ESG Reporting of Automated HVAC Control Energy Savings

PD10: Toward State-Aware Directed Greybox Fuzzing for MQTT

PD11: ChickWiz: A Multimodal Situation-Aware Conversational Agent for Smart Poultry Farming

PD12: Bounded-Priority Scheduling for Mixed Collective Traffic in HPC Networks

DAY 2 (October 22, 2026)

Session 3A: Network Intelligence and Open RAN (10:30 – 11:40)

Session Chair: TBA
Room: TBA

SlicePlanner: Logic-Guided Agentic AI for Online QoS-Aware Network Slicing in Open RAN
Source track: Systems and Applications

X-RIC: An Agentic RAN-Intelligent Controller Benchmark for Enhancing Reinforcement Learning
Source track: AI/ML Smart Design

SF-Detector: An Efficient and Accurate Method for Detecting Steady Flows in Network Traffic
Source track: Algorithms & Theory

Decentralized Log-Linear Learning for Periodic Random Access Games
Source track: AI/ML Smart Design

Session 3B: Edge Platforms and Model Efficiency (10:30 – 11:40)

Session Chair: TBA
Room: TBA

An Uncertainty-Aware Resilience Micro-Agent for Causal Observability in the Computing Continuum
Source track: Protocols and Emerging Technologies

FetchRight: Two-Stage Network-Adaptive Image Fetching for Fast Edge Container Provisioning
Source track: Systems and Applications

Revisiting Pruning for Edge AI: Latency, Scalability, and OTA Update Costs
Source track: Systems and Applications

HyRAP: Hybrid Resource-Aware Parallelism for MoE Fine-Tuning on Heterogeneous Clusters
Source track: Systems and Applications

Session 4A: Wireless Sensing and Applications (13:10 – 14:20)

Session Chair: TBA
Room: TBA

Diffusion-Based Co-Channel Interference Resistance for WiFi Sensing
Source track: Systems and Applications

RESA: A Radar-Exclusive Spatial Assistant for Indoor mmWave Wheelchair Guidance
Source track: Systems and Applications

Sensing-Assisted Anti-Blockage Pinching-Antenna Systems For Indoor Immersive Communications
Source track: Algorithms & Theory

Session 4B: UAV, Aerial, and Orbital Networks (13:10 – 14:20)

Session Chair: TBA
Room: TBA

Efficient Resource Allocations for Light-weight UAV Base Stations with Differential QoS Requirements
Source track: Systems and Applications

Latency-Energy Co-Optimized Task Scheduling for Earth Observation in Orbital Edge Computing
Source track: AI/ML Smart Design

Channel-Constrained UAV Swarm Task Allocation via Learned Communication Scheduling
Source track: AI/ML Smart Design

Policy Sharing for Signal Strength-Aware 3D Path Planning in Multi-UAV Systems
Source track: AI/ML Smart Design

Session 2A: Security, Privacy, and Reliable Networked Systems (14:50 – 16:00)

Session Chair: TBA
Room: TBA

CrossZK: Privacy-Preserving Inter-Organizational Verification for Consortium Blockchains
Source track: Security and Privacy

LLM-Assisted Initialization Bug Detection for Smart-System Software Reliability
Source track: Security and Privacy

Secure BATMAN-adv: A Routing-Layer Security Framework for Wireless Mesh Networks
Source track: Security and Privacy

Interaction Graph Anomaly Detection via Semantic Edge Augmentation with Large Language Models
Source track: AI/ML Smart Design

Session 2B: Underwater and Underground Networks (14:50 – 15:42:30)

Session Chair: TBA
Room: TBA

CDM-TSM: Correlation Domain Multiplexed Time Shift Modulation for Underwater AComms
Source track: Protocols and Emerging Technologies

MADRID: Material-Aware Clock Synchronization for Internet-of-Underground Things Deployments
Source track: Protocols and Emerging Technologies

FATHOM: Underwater Acoustic Channel Estimation via Dual Domain Windowed Attention
Source track: AI/ML Smart Design