Welcome to SJSU interconnect lab!
interconnect lab, led by Professor Genya Ishigaki, researches next-generation telecommunication networks, with an emphasis on developing combinatorial optimization and machine learning techniques to address problems related to resource allocation. Resource allocation problems appear in many networking contexts including reliability, scalability, and profitability. A fundamental question we are trying to answer is about the tradeoff in resource allocation: How does a network operator balance the maximization of resource utilization and the reservation of spare capacity for future demands? Even though the tradeoff has been a paramount agenda of resource optimization research, the new features introduced for next-generation networking, such as dependency among different network layers, diverse network usages, and unpredictable traffic patterns, present additional challenges to answer the question in 5G and Beyond networks.
Our recent research themes include:
- ML-based 5G Network Resource Pricing
- Accountable Autonomous Network Operation with Explainable AI
- Edge Computing-Oriented Federated Learning
If you are interested in our projects, please look at our research page and publication list.
News
CFP: Big Data and Machine Learning for Networking
Professor Ishigaki is organizing the 5th International Workshop on Big Data and Machine Learning for Networking (BDMLN) as a part of ICCCN 2025 in Tokyo, Japan from August 4 to 7, 2025. CFP can be found here.
May 10, 2023 - Award
Rachel was selected as the award recipient of the 2024 Broadcom Scholarship for her exceptional academic performance and contribution to supporting underrepresented students in STEM.
May 9, 2023 - Publication
Three of our papers have been accepted to ICCCN 2024. See the publication section for more details.
April 19, 2024 - Presentation
Austin, Rachel, and Pandu presented at the 19th College of Science Student Research Day and SJSU Innovation Showcase!
August 4, 2023 - Publication
Our paper titled “Container Caching Optimization based on Explainable Deep Reinforcement Learning” has been accepted to IEEE Globecom 2023. See the publication section for more details.
May 23, 2023 - Publication
Our paper titled “Reinforcement Learning with Neural Network-based Deterministic Game Tree Approximation” has been accepted to IEEE BigDataService 2023. See the publication section for more details.