Defending Against IoT-Enabled DDoS Attacks at Critical Vantage Points on the Internet

Devkishen Sisodia
Date and time: 
Wed, May 25 2022 - 1:00pm
Location: 
Zoom
Speaker(s):
Devkishen Sisodia
University of Oregon
Host/Committee: 
  • Jun Li (Chair)
  • Joseph Sventek
  • Lei Jiao
  • Ramón Alvarado (Philosophy)

The number of Internet of Things (IoT) devices continues to grow every year. Unfortunately, with the rise of IoT devices, the Internet is also witnessing a rise in the number and scale of IoT-enabled distributed denial-of-service (DDoS) attacks. However, there is a lack of network-based solutions targeted directly for IoT networks to address the problem of IoT-enabled DDoS. Unlike most security approaches for IoT which focus on hardening device security through hardware and/or software modification, which in many cases is infeasible, we introduce network-based approaches for addressing IoT-enabled DDoS attacks. We argue that in order to effectively defend the Internet against IoT-enabled DDoS attacks, it is necessary to consider network-wide defense at critical vantage points on the Internet. This dissertation is focused on three inherently connected and complimentary components: (1) preventing IoT devices from being turned into DDoS bots by inspecting traffic towards IoT networks at an upstream ISP/IXP, (2) detecting DDoS traffic leaving an IoT network by inspecting traffic at its gateway, and (3) mitigating attacks as close to the devices in an IoT network originating DDoS traffic. To this end, we present three security solutions to address the three aforementioned components to defend against IoT-enabled DDoS attacks.