Full Program

13:30 - 21:30

Conference Registration

09:00 - 09:30

Opening Ceremony

Welcome remarks and official photo session

09:30 - 10:10

Keynote 1: Toward Scalable Generative AI via Mixture of Experts in Mobile Edge Networks and Metaverse

Dusit Niyato
Nanyang Technological University, Singapore

10:10 - 10:50

Keynote 2: Multimodal Scene Understanding and Multimodal Foundation Model

Jungong Han
Tsinghua University, China

10:50 - 11:05

Coffee Break

11:00 - 12:00 Room 1

Session 1: Federated Learning and Large Language Models

  • Tiny Models, Tough Limits
  • GUIDE: A Conversational LLM Framework for Dynamic Deep Reinforcement Learning based Building Energy Management
  • FISPA: A LLM-based Assistance Framework for Fire Investigation
  • 11:00 - 12:00 Room 2

    Session 2: Security Enhancement Schemes

  • Application of Dual-segment Non-integer Byte CRC in High-Reliability Satellite-to-ground TDM Communication Verification
  • Formal Verification of DMA-enabled Memory Isolation in MPU-based Embedded Systems
  • RideStyle: Riding Style Representation from Head-Body Dynamics via Adversarial Learning
  • 11:00 - 12:00 Room 3

    Session 3: Architectures and Algorithms for IoT

    • Unmanned Aerial Vehicle Identity Authentication Method for Open-Set of RF Signal and Cross-domain Scenario
    • DASFO: Dependency-Aware Scalable Function Offloading Framework for Intelligent Connected Vehicles
    • Towards Reliable Distributed Systems for Safety-Critical Vehicle-Road-Cloud Continuum

    11:00 - 12:00 Room 4

    Session 4: Cryptographic Schemes

    • Do not Keep Guessing: Enhanced Secure Cloud Storage with Keyword Search and Deduplication
    • EP-BPRE: Efficient Pairing-free Broadcast Proxy Re-encryption for Secure Data Sharing in Internet of Vehicle

    12:00 - 13:30

    Lunch Break

    13:30 - 15:30 Room 1

    Session 1: Federated Learning and Large Language Models

    • MobileLogZip: Lightweight LLM-Driven Tiered Log Compression for Mobile Computing
    • STrustzoneFL: Secure Federated Learning based on TrustZone
    • A Prototype Network Assisted Training Mechanism for Edge Federated Learning Models
    • Heterogeneous Federated Learning of Data and Models Based on Adaptive Personalized Aggregation Prototype

    13:30 - 15:30 Room 2

    Session 2: Security Enhancement Schemes

    • Robust Reinforcement Learning for Q-Network in Worst-Case Scenarios
    • Improving NTP Time Synchronization Accuracy for NTS Enhancement
    • Wave-to-Mute: In-Cabin Single Gesture Mute Control via Acoustic Sensing
    • SPUAuth: Strengthening Pattern Unlock Authentication with Behavioral Biometrics Against Shoulder Surfing Attacks
    • Multi-Task Skill Learning via Meta-Policy Guided Fine-Tuning

    13:30 - 15:30 Room 3

    Session 3: Architectures and Algorithms for IoT

    • FloodTracker: Towards Truly Wireless and Battery-Free Flood Monitoring with Flute
    • Heuristic-Enhanced Multi-Agent Learning for Intelligent Caching and Offloading in UAV-Enabled SAGIN
    • Joint UAV Deployment, Task Offloading and Resource Allocation for Multi-UAV Collaborative Edge Computing Network
    • Mobility-Aware Resource Scheduling in 6G Space-Air-Ground Integrated Networks: A GNN-Enhanced DRL Approach
    • Using Machine Learning to Take Stay-or-Go Decisions in Data-driven Drone Missions
    • A Trustworthy Task Offloading Scheme based on Tournament and Elitist Genetic Algorithm for Vehicular Networks

    13:30 - 15:30 Room 4

    Session 4: Cryptographic Schemes

    • Privacy-Preserving Cross-Domain Authentication in IoV with Lightweight SM2 Group Signatures
    • Secure and Flexible HD Wallets: Enabling Mnemonic Updates and Threshold-Controlled Key Derivation
    • A Lightweight Multi-Authority KP-ABE Scheme Supporting Large Attribute Universes
    • A Hierarchical Key-Based Dynamic Cross-Domain Authentication Scheme for VANETs
    • A Provably Secure and Fault-Tolerant Verifiable Encryption Scheme with Blockchain-Based Private Key Custody
    • A Dynamic Load Balancing Scheme with Differential Privacy for Traffic Analysis in Security Crowdsourcing Platforms

    15:30 - 16:00

    Coffee Break

    16:00 - 18:00 Room 1

    Session 1: Federated Learning and Large Language Models

    • Location-based LLM (LLLM): Trip Purpose Estimation Using Language Models of Location Data
    • A Low-Resource Vision Transformer Accelerator for Edge Applications
    • Enhancing Heterogeneous Federated Learning via Historical Dual Knowledge Distillation
    • A Lightweight Federated Learning Framework with Dynamic Local Differential Privacy and Transfer Learning
    • A Multi-modality Dynamic Feature Fusion Method for Android Malware Detection
    • Multi-distance Identifies Backdoors Attack in Privacy-Preserving Federated Learning

    16:00 - 18:00 Room 2

    Session 2: Security Enhancement Schemes

    • A Trust-based Incentive Mechanism Combined with Behavioral Economics and Payment Member for Truth Discovery in Mobile Crowdsensing
    • Progressive Weight Loading: Accelerating Initial Inference and Gradually Boosting Performance on Resource-Constrained Environments
    • An Traffic Prediction Scheme for LEO Satellite Networks with Graph Neural Networks
    • Gaze-Aware Voronoi-Based Key Region Optimization for Virtual Reality Text Input

    16:00 - 18:00 Room 3

    Session 3: Architectures and Algorithms for IoT

    • Neuro Evolutionary Swarm Intelligence Framework for Real Time Anomaly Detection in IoT Enabled Industrial Networks
    • A Hierarchical Federated Framework for Task Offloading in Satellite-terrestrial Integrated IoT Networks
    • DPA: An Efficient Cross-Domain Face Anti-Spoofing Framework via Dual-Parallel Adapter Fine-Tuning
    • Mitigating Phishing Attacks Through Multi-Layer Edge Task Offloading For Cyber-Physical Systems

    16:00 - 18:00 Room 4

    Session 4: Cryptographic Schemes

    • Heavy Hitter Identification in Large Domains with Personalized Local Differential Privacy
    • A Blockchain-Assisted Outsourced Signcryption Scheme for Fine-Grained VSN Data Sharing System
    • One More Bit, Much More Security? Differential Fault Analysis of 5-bit S-Box Candidates in the NIST LWC Competition
    • MLM-ABE: A Privacy-Preserving Access Control Scheme for Multi-LLM APIs and Local Sensitive Data

    18:30 - 20:30 No. 1 Nandao, Lingang New City, Pudong New Area, Shanghai

    Social Dinner

    InterContinental Shanghai Lingang New City
    Bus transfer is arranged between the venue and the restaurant.

    09:30 - 12:00

    Visit the Shanghai Astronomy Museum

    Shanghai Astronomy Museum
    Bus transfer is arranged between the venue and the restaurant

    12:00 - 13:30

    Lunch Break

    13:30 - 13:40

    Closing Remarks

    Scroll to Top