DataSys: Data-Intensive Distributed Systems LaboratoryData-Intensive Distributed Systems Laboratory

Illinois Institute of Technology
Department of Computer Science

CS Seminar

Date: January 28th, 2026
Time: 12:50pm - 1:45pm
Room: SB111

Dr. Nik Sultana

Assistant Professor
Computer Science Department
Illinois Institute of Technology

Talk Title

Patchwork: A Traffic Capture and Analysis Platform for Network Experiments on a Federated Testbed

Talk Abstract

Today’s federated network testbeds enable experiments of unprecedented scale and detail, but only provide rudimentary primitives to capture an experiment’s network traffic. Capturing and analyzing traffic is important for diagnosing and debugging research prototypes and for evaluating research, but today’s testbed users divert and duplicate effort to craft custom solutions for their experiments. Building a general, reusable system is made more challenging by the autonomous structure of federated testbeds and by their high capacity links (requiring accelerated processing). This talk describes the design, implementation, and evaluation of Patchwork: a user-provided, open-source, capture and analysis platform that runs on the state-of-the-art FABRIC testbed. Patchwork works both for individual experiments and also for all experiments occurring simultaneously on FABRIC. To attain a general design, Patchwork itself runs as an experiment on FABRIC and did not require modifications to FABRIC. For scalability, Patchwork offloads logic to FPGA NICs on the testbed and uses DPDK. Patchwork has been used by individual FABRIC users and has been running on FABRIC for over a year to produce a testbed-wide analysis of how researchers are collectively using the testbed’s network. This paper (https://packetfilters.cs.iit.edu/patchwork/patchwork.pdf) also presents that analysis and discusses implications for future research on measurement. More information on Patchwork can be found at https://packetfilters.cs.iit.edu/patchwork/.

Speaker Bio

Nik Sultana is an assistant professor of Computer Science at Illinois Institute of Technology. He develops networking techniques to improve cybersecurity and research infrastructure. Before joining Illinois Tech, he was a post-doc at UPenn after completing his PhD at Cambridge University. In 2024, 2023, and 2025 he received VSP awards from the Universities Research Association, and in 2022 he received a Google Research Award.

Back to Seminar Schedule