Communication-Computing Co-Design for Real-Time Computer Vision Services
Goal
- Maximizing the performance of semantic communications and implementing a real-time computer vision transmission system
Approach
- Joint optimization of communication and computing resources
- Breaking through the tradeoff among fidelity, energy, and delay
Issues
- Maximizing accuracy while ensuring end-to-end latency for real-time intelligent surveillance services
- Minimizing energy consumption while guaranteeing precision for real-time immersive 3D/VR monitoring services
- Minimizing energy-latency costs while ensuring fidelity for real-time interactive XR conferencing services
- Implementation and performance evaluation of a real-time computer vision transmission system in virtual and real environments
Related Projects
- Communication-Computing Resource Optimization and Cooperative Inference Protocols for Real-Time Computer Vision Services Ongoing
- Development of Auto Inline Test System Software: Vision-Based Pattern Inspection and Data Matrix Reading for Production Lines Ongoing
Communications and Networks for Physical AI
Goal
- Designing communications and networks that let many physical AI agents (cameras, robots, drones) share scarce wireless resources and still act in real time, by transmitting the meaning of what they sense instead of raw bits
Approach
- Semantic multiple access: superposing the semantic embeddings of multiple agents on a single shared resource block with learned inter-user positional encoding
- Transformer receiver that separates each agent by query-based attention, without successive interference cancellation
- Over-the-air validation on a software-defined radio (USRP) testbed
Issues
- Semantic multiple access for multi-agent physical AI
- Real-time multi-camera semantic video transmission
- Over-the-air SDR testbed
Related Projects
- Prototype Edge Gateway for Multi-Camera Semantic Video Transmission in Construction-Site Safety Monitoring Based on AI Multi-User Semantic Communication Patents Ongoing
- AI Semantic Communication Research Service Ongoing
Energy-Efficient Protocol for 6G Vertical Communications
Goal
- Maximizing energy efficiency in 6G vertically integrated networks with mobile devices, UAVs, and satellites
Approach
- Wireless energy harvesting UAV communication systems
- Context-aware low-power satellite communication networks
- Energy-efficient integrated terrestrial–aerial–satellite networks
Issues
- Minimizing energy consumption in UAV systems with wireless energy transfer and harvesting
- Maximizing the lifetime of LEO satellite networks while ensuring end-to-end QoS
- Maximizing energy efficiency in 6G integrated terrestrial–aerial–satellite networks
Related Projects
- A Study on Energy-Efficient Communication Protocol in 6G Vertical Mobile Communication Environments
- An Optimal Multi-hop Transmission Scheme for Wireless Powered Communication Networks
- Development of System-Level Simulator for Low-Energy Drone Identification in Dense Drone-Distributed Environments
- Development of System-Level Simulator for Low-Energy Drone Identification
- System Performance Comparison of ALOHA Protocols According to Drone Distribution
- Group Scheduling Method for Efficient Identification in Drone Swarm Environments
Machine Learning-based Communications and Networks
Goal
- Achieve fully autonomic operation, maximum throughput, and minimum cost through machine learning
Approach
- Solve CN & SON issues by applying ML algorithms
Issues
- ML-based optimization of communication networks
- Computing and networking for machine learning services
Related Projects
- Machine Learning-Based Energy-Efficiency Optimization in 5G NR-Based Small Cell Cluster Environments
- Energy-Saving Algorithm with AI-Based Handover Techniques in Dense 5G Small Cell Clusters
Bio-Inspired Self-Organizing Communication and Networks
Goal
- Achieve self-organized networking with low overhead while guaranteeing user quality of service
Approach
- Apply bio-inspired algorithms (e.g., ACO, flocking, firefly synchronization/desynchronization)
Issues
- Bio-inspired routing
- Bio-inspired resource management
- Bio-inspired energy saving
- Bio-inspired fast consensus
- Bio-inspired system modeling and engineering
Related Projects
- Energy-Efficient Communication and Networking Protocols Based on Natural Swarm Intelligence for Internet of Everything
- Radio Resource Management and Routing Protocols Based on Bio-Inspired Algorithms in Tactical Mobile Ad Hoc Networks
- Development of Distributed MAC and Radio Resource Management for Autonomous Networks
Next-Generation Wireless Communications and Networks
Goal
- Maximize data rate and energy efficiency while guaranteeing seamless connectivity
Issues
- Wireless energy harvesting
- Cooperative interference management
- Distributed resource management & access control
- Ultra-low power management
- Seamless connectivity management
Related Projects
- Network Structure and Communication Protocol for Energy-Neutral Operations of Wireless Networks
- Optimization Algorithm for Energy Saving in 5G NR-Based Small Cell Clusters
- Development of Open 5G Standard Model
- Development of Device Collaborative Giga-Level Smart Cloudlet Technology
- Energy-Efficient Machine-to-Machine Communication and Networking Technologies for Logistics Center Management
- Energy-Efficient Communication Protocol Considering Network Delay Variation in Next-Generation Heterogeneous Wireless Networks
- Research on Service Scenarios for 5G Wireless Cellular Networks
- Analysis on Services and Requirements for Proximate Internet
- Research on Core Technology in 4G+ Wireless Environments
- Distributed Resource Management for Large-Scale Networks