This technical paper was recently presented at the 2022 Cable & Connectivity Industry Forum. Do you have research to share? Submit a Technical Paper Abstract before April 7, 2023 to be considered for the 72nd annual IWCS Cable & Connectivity Industry Forum, taking place September 17–20 in Orlando, Florida, USA! Review the IWCS Call for Papers to discover suggested topics and submission criteria.
Abstract: This paper delves into the efficacy of various external noise mitigation strategies and how they affect a network’s susceptibility to packet-in-packet (PIP) cybersecurity attacks. This style of attack is performed by embedding a malicious, but valid, Internet Protocol (IP) packet into the payload of a benign IP packet to traverse network security measures. Properly-timed bit-flip errors can cause network devices to interpret embedded packets as valid datagrams—potentially allowing malicious packets to infiltrate networks via faulty Ethernet cables. Because this style of attack relies heavily on externally induced noise, a cable’s ability to insulate itself from environmental stressors is closely tied to its susceptibility. Four common cable constructions were used in combination with both shielded and unshielded plugs and jacks to evaluate numerous short channels’ ability to mitigate external noise and prevent PIP attacks. These channels were evaluated passively with a waveform generator and an oscilloscope to determine the amount of coupling that could be induced by a variety of disturber types. Some noise sources were constructed such that coupling between them and the cable core of the channel under test were optimized. The channels under test were then subjected to PIP attacks over a 10GBASE-T link in the presence of the noise sources used for passive testing. The rate at which the embedded packets were detected was then measured.
After the passive and active testing was completed, the efficacy of each external noise mitigation strategy was evaluated. The fully shielded cable was the most effective at mitigating external noise and was, therefore, the least susceptible to packet-in-packet attacks. The cable with a continuous metallic isolation wrap enclosed in nonconductive materials was the second most effective at mitigating external noise. The cable with a discontinuous, or segmented, metallic isolation wrap was the third most effective at mitigating external noise. The completely unshielded cable was the least effective at mitigating external noise and was, therefore, the most susceptible to PIP attacks.
About the Presenter: Michael Dodds received his Master of Science in Electrical Engineering and Biomedical Engineering from Drexel University in 2017. He is currently a Test Engineer at Leviton’s New Holland facility.
