NeTS: Architectural Foundations for Deployable Low-Latency and Resilient Networks

Modern applications such as virtual reality and connected autonomous systems depend on networks that have low delay and robust operation under changing conditions. However, today’s Internet infrastructure often cannot deliver these requirements, not because of a lack of technical solutions, but because existing solutions are impractical to deploy widely due to incompatibilities with current systems. Rooted in design-for-deployability principles, this project establishes architecturally extensible approaches that enable the evolutionary integration of new capabilities within legacy network infrastructures. By enabling more efficient and reliable communication, the project will support emerging industries such as interactive media, transportation systems, and manufacturing. Collaboration with industry partners ensures that the research reflects real-world deployment conditions and accelerates the transition of new ideas into operational networks. The project also contributes to education and workforce development by engaging students in research, creating open-source educational materials based on project outcomes, and preparing students for careers in advanced networking. The project develops deployable network architectures that enable low-delay and resilient operation and is organized into three thrusts. The first thrust develops demand-anchored architectures for high-frequency spectrum, using analytical and economic models to determine when and how additional capacity can be introduced in alignment with existing demand. The second thrust develops reordering-tolerant communication mechanisms that decouple reliable delivery and congestion control from strict packet ordering, enabling multipath and multi-link operation without requiring expensive in-network resequencing. The third thrust develops end-to-end low-latency transport frameworks that address coexistence across heterogeneous bottlenecks, including new queue management mechanisms, cross-domain traffic steering, and extensions into cellular networks. Across all thrusts, the research combines analytical modeling, protocol and system design, and experimental validation on large-scale NSF-supported platforms such as CloudLab, FABRIC, and POWDER, with ongoing input from industry partners. The outcomes will establish general design principles for deployable networking systems and provide open-source tools and educational resources that support broader adoption and training. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.