Match-Prevention Technique Against Denial-of-Service Attack on Address Resolution and Duplicate Address Detection Processes in IPv6 Link-Local Network

摘要

Address Resolution (AR) and Duplicate Address Detection (DAD) are considered the most important processes in Neighbour Discovery Protocol (NDP), which occurs frequently from each Internet Protocol version 6 (IPv6) host communicating with other neighbouring hosts. Two NDP messages are used during AR and DAD to communicate with one another in the same IPv6 link-local network, namely Neighbour Solicitation (NS) and Neighbour Advertisement (NA) messages. However, NDP messages have non-secure designs and lack verification mechanisms for authenticating whether incoming messages originate from a legitimate or illegitimate node. Therefore, any node in the same link can manipulate NS or NA messages and then launch a Denial-of-Service (DoS) attack. Techniques proposed to secure AR and DAD include Secure NDP (SeND) and Trust-NDP (Trust-ND); however, these techniques either entail high processing time and bandwidth consumption or are vulnerable to DoS attacks because of their designs. Therefore, to secure AR and DAD, this study aims to introduce a prevention technique called Match-Prevention, which secures target IP addresses and exchange messages (i.e. NS and NA). The processing time, bandwidth consumption and DoS prevention success rate of Match-Prevention in different scenarios are evaluated, and its performance is compared with those of existing techniques, including Standard-Process (i.e., Standard-AR and Standard-DAD), SeND and Trust-ND. Results show that Match-Prevention requires less processing time during AR and DAD processes and less bandwidth consumption compared with other existing techniques. In terms of DoS prevention success rate, the experiments show that Standard-Process and Trust-ND are unable to secure AR and DAD from DoS attacks, whilst SeND is vulnerable to flooding attacks. By contrast, Match-Prevention allows IPv6 nodes to verify the incoming message, discard the fake message before further processing and prevent a DoS attack during AR and DAD in an IPv6 link-local network.