P4Pir: In-Network Analysis for Smart IoT Gateways: in-network analysis for smart IoT gateways

Mingyuan Zang; Changgang Zheng; Radostin Stoyanov; Lars Dittmann; Noa Zilberman

doi:10.1145/3546037.3546060

P4Pir: In-Network Analysis for Smart IoT Gateways: in-network analysis for smart IoT gateways

Mingyuan Zang, Changgang Zheng, Radostin Stoyanov, Lars Dittmann, Noa Zilberman

University of Oxford

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

Abstract

IoT gateways are vital to the scalability and security of IoT networks. As more devices connect to the network, traditional hard-coded gateways fail to flexibly process diverse IoT traffic from highly dynamic devices. This calls for a more advanced analysis solution. In this work, we present P4Pir, an in-network traffic analysis solution for IoT gateways. It utilizes programmable data planes for in-band traffic learning with self-driven machine learning model updates. Preliminary results show that P4Pir can accurately detect emerging attacks based on retraining and updating the machine learning model.

Original language	English
Title of host publication	Proceedings of the SIGCOMM '22 Poster and Demo Sessions
Publisher	Association for Computing Machinery
Publication date	2022
Pages	46-48
ISBN (Print)	9781450394345
DOIs	https://doi.org/10.1145/3546037.3546060
Publication status	Published - 2022
Event	ACM SIGCOMM 2022 - Beurs van Berlage, Amsterdam, Netherlands Duration: 22 Aug 2022 → 26 Aug 2022

Conference

Conference	ACM SIGCOMM 2022
Location	Beurs van Berlage
Country/Territory	Netherlands
City	Amsterdam
Period	22/08/2022 → 26/08/2022

Keywords

P4
in-network computing
internet of things
machine learning
security

Access to Document

10.1145/3546037.3546060

OpenUrl availability

Full text

Cite this

@inproceedings{e03327b40d644a36bd7befa45eadaae9,

title = "P4Pir: In-Network Analysis for Smart IoT Gateways: in-network analysis for smart IoT gateways",

abstract = "IoT gateways are vital to the scalability and security of IoT networks. As more devices connect to the network, traditional hard-coded gateways fail to flexibly process diverse IoT traffic from highly dynamic devices. This calls for a more advanced analysis solution. In this work, we present P4Pir, an in-network traffic analysis solution for IoT gateways. It utilizes programmable data planes for in-band traffic learning with self-driven machine learning model updates. Preliminary results show that P4Pir can accurately detect emerging attacks based on retraining and updating the machine learning model.",

keywords = "P4, in-network computing, internet of things, machine learning, security",

author = "Mingyuan Zang and Changgang Zheng and Radostin Stoyanov and Lars Dittmann and Noa Zilberman",

year = "2022",

doi = "10.1145/3546037.3546060",

language = "English",

isbn = "9781450394345",

pages = "46--48",

booktitle = "Proceedings of the SIGCOMM '22 Poster and Demo Sessions",

publisher = "Association for Computing Machinery",

address = "United States",

note = "ACM SIGCOMM 2022 ; Conference date: 22-08-2022 Through 26-08-2022",

}

P4Pir: In-Network Analysis for Smart IoT Gateways: in-network analysis for smart IoT gateways. / Zang, Mingyuan; Zheng, Changgang; Stoyanov, Radostin et al.
Proceedings of the SIGCOMM '22 Poster and Demo Sessions. Association for Computing Machinery, 2022. p. 46-48.

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

TY - GEN

T1 - P4Pir: In-Network Analysis for Smart IoT Gateways

T2 - ACM SIGCOMM 2022

AU - Zang, Mingyuan

AU - Zheng, Changgang

AU - Stoyanov, Radostin

AU - Dittmann, Lars

AU - Zilberman, Noa

PY - 2022

Y1 - 2022

N2 - IoT gateways are vital to the scalability and security of IoT networks. As more devices connect to the network, traditional hard-coded gateways fail to flexibly process diverse IoT traffic from highly dynamic devices. This calls for a more advanced analysis solution. In this work, we present P4Pir, an in-network traffic analysis solution for IoT gateways. It utilizes programmable data planes for in-band traffic learning with self-driven machine learning model updates. Preliminary results show that P4Pir can accurately detect emerging attacks based on retraining and updating the machine learning model.

AB - IoT gateways are vital to the scalability and security of IoT networks. As more devices connect to the network, traditional hard-coded gateways fail to flexibly process diverse IoT traffic from highly dynamic devices. This calls for a more advanced analysis solution. In this work, we present P4Pir, an in-network traffic analysis solution for IoT gateways. It utilizes programmable data planes for in-band traffic learning with self-driven machine learning model updates. Preliminary results show that P4Pir can accurately detect emerging attacks based on retraining and updating the machine learning model.

KW - P4

KW - in-network computing

KW - internet of things

KW - machine learning

KW - security

U2 - 10.1145/3546037.3546060

DO - 10.1145/3546037.3546060

M3 - Article in proceedings

SN - 9781450394345

SP - 46

EP - 48

BT - Proceedings of the SIGCOMM '22 Poster and Demo Sessions

PB - Association for Computing Machinery

Y2 - 22 August 2022 through 26 August 2022

ER -

P4Pir: In-Network Analysis for Smart IoT Gateways: in-network analysis for smart IoT gateways

Abstract

Conference

Keywords

Access to Document

OpenUrl availability

Fingerprint

Cite this