
With the rise of real-time communication applications such as remote conferencing, teleoperation, and cloud-based AR, ensuring consistent and high-quality mobile connectivity has become increasingly critical. Although 5G is positioned as a foundational technology for these services, its performance often fluctuates depending on location, time, and carrier. To address these limitations, our research team has developed a novel multipath redundancy communication framework that enhances the quality of real-time streaming by combining multiple mobile network connections.
Focusing on UDP-based applications like WebRTC, our framework overcomes the limitations of conventional TCP-centric multipath systems. By employing IP-layer tunneling and a receiver-side buffering mechanism for packet reordering, the system maintains smooth real-time playback even across diverse mobile connections. Field tests conducted in urban environments using moving vehicles demonstrated a substantial reduction in packet loss and a significant improvement in bitrate without interfering with application-layer congestion control mechanisms.
In addition to system design, we evaluated the Quality of Service (QoS) of WebRTC-based remote collaboration over real-world 5G conditions. Our experiments analyzed the influence of various environmental and network parameters—such as signal bands, handoffs, and mobility—on media quality. The study also highlighted the potential for performance improvement when utilizing multiple mobile carriers. The collected data has been published openly to support further research on QoS and 5G reliability.
These findings contribute to the broader vision of future-ready mobile infrastructure, supporting use cases like mobile XR, vehicle-to-cloud communication, and emergency response systems. By merging real-time media technologies with adaptive multipath strategies, we aim to create robust, flexible, and high-performance communication platforms for next-generation mobile networks.
@article{ITO2025108157,
title = {A multipath redundancy communication framework for enhancing 5G mobile communication quality},
author = {Koki Ito and Jin Nakazato and Romain Fontugne and Manabu Tsukada and Hiroshi Esaki},
url = {https://www.sciencedirect.com/science/article/pii/S0140366425001148},
doi = {https://doi.org/10.1016/j.comcom.2025.108157},
issn = {0140-3664},
year = {2025},
date = {2025-04-23},
urldate = {2025-04-23},
journal = {Computer Communications},
pages = {108157},
abstract = {As networks increasingly become the backbone of modern society, the demands placed on them by various applications have become more complex. In particular, the demand for high-capacity, low-latency services such as real-time streaming is increasing every year. Although 5G has been deployed to meet these needs, its effectiveness can vary significantly by location and time, and sometimes falls short of requirements. Traditionally, much of the research to improve communication stability has focused on TCP-based systems, which do not translate well to real-time UDP streaming applications. To address the above challenges, we propose a multipath redundant communication framework designed to improve the quality of real-time media streaming. This framework has been tested using multipath redundant communication over two mobile networks with a moving vehicle in an urban environment. Using a real-time streaming application based on WebRTC, our framework demonstrates a significant reduction in packet loss and an increase in bitrate, outperforming existing multipath redundant communication systems without interfering with the application’s congestion control mechanisms.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
@inproceedings{Ito2024,
title = {Enhancing Real-Time Streaming Quality through a Multipath Redundant Communication Framework},
author = {Koki Ito and Jin Nakazato and Romain Fontugne and Manabu Tsukada and Hiroshi Esaki},
doi = {10.23919/IFIPNetworking62109.2024.10619885},
year = {2024},
date = {2024-06-03},
urldate = {2024-06-03},
booktitle = {IFIP/IEEE Networking 2024},
address = {Thessaloniki, Greece},
abstract = {Recently, as networks operate as the infrastructure of modern society, the demands placed on the network by applications have become more complex. In particular, an increasing annual demand for high-capacity and low-latency services, including real-time streaming. 5G services have been launched to meet this demand, but their stability varies de- pending on location and time and can only sometimes be considered sufficient. One method to improve communication stability is multipath redundant communication, and much research has been conducted in this area. However, most of this research has focused on TCP-based communication and cannot be applied to real-time UDP streaming. Hence, we propose a multipath redundant communication framework to improve the quality of real-time media streaming communication. Tunneling at the IP layer in our proposed framework was performed to overcome the limitations of transport layer protocols, which was a challenge for traditional multipath redundant communication systems. Furthermore, to address the packet order inconsistency caused by multipath redundant communication, a buffering mechanism was implemented on the receiving side of our system. Our proposed system was verified using multipath redundant communication and multiple mobile networks from a vehicle moving in an urban area. The experiments used a real-time streaming application based on WebRTC, and the framework significantly reduced packet loss and improved bitrate compared to existing multipath redundant communication systems without interfering with the application’s congestion control.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
@article{Nakazato2023,
title = {WebRTC over 5G: A Study of Remote Collaboration QoS in Mobile Environment},
author = {Jin Nakazato and Kousuke Nakagawa and Koki Ito and Romain Fontugne and Manabu Tsukada and Hiroshi Esaki},
url = {https://link.springer.com/content/pdf/10.1007/s10922-023-09778-5.pdf},
doi = {10.1007/s10922-023-09778-5},
issn = {1573-7705},
year = {2023},
date = {2023-10-24},
urldate = {2023-10-24},
journal = {Journal of Network and Systems Management},
volume = {32},
issue = {1},
abstract = {The increasing demand for remote collaboration and remote working has become crucial to daily life owing to the Covid-19 pandemic and the development of internet-based video distribution services. Furthermore, low-latency remote collaboration, such as teleoperation and support applications designed for in-vehicle environments, has gained considerable attention. The 5G technology is considered as a key infrastructure for remote collaboration. This study aimed to evaluate the actual 5G capability to achieve high quality of service (QoS) for remote collaboration. We designed and implemented a measurement tool to monitor the QoS of remote collaboration under real-world 5G conditions. We performed measurements encompassing the various 5G frequency bands. During these experiments, we employed various tools to obtain detailed mobile signal conditions to analyze the relationship between various environmental factors (e.g. signal quality, band, handoff, geographic conditions, and mobility) and the QoS performance of remote collaboration in a real-world 5G environment. This study elucidated the correlation between the WebRTC performance and various environmental factors as well as the performance improvement potential by leveraging the communication technologies of multiple mobile carriers. The collected data has been made publicly available to foster research on QoS and 5G.},
keywords = {},
pubstate = {published},
tppubtype = {article}
}
digital twins extended reality
digital twins
autonomous driving machine learning
machine learning v2x
autonomous driving v2x
extended reality
v2x
machine learning v2x
We are part of the University of Tokyo’s Graduate School of Information Science and Technology, Department of Creative Informatics and focuses on computer networks and cyber-physical systems
Address
4F, I-REF building, Graduate School of Information Science and Technology, The University of Tokyo, 1-1-1, Yayoi, Bunkyo-ku, Tokyo, 113-8657 Japan
Room 91B1, Bld 2 of Engineering Department, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8656, Japan
Mail: