Intrusion Detection Evasive Techniques
Most attackers are aware of IDSs and use evasive techniques to dodge them. These evasive techniques include flooding, fragmentation, encryption, and obfuscation.
IDSs depend on resources such as memory and processor power to effectively capture packets, analyze traffic, and report malicious attacks. By flooding a network with noise traffic, an attacker can cause the IDS to exhaust its resources examining harmless traffic. In the meantime, while the IDS is distracted and occupied by the volume of noise traffic, the attacker can target its system with little or no intervention from the IDS.
Because different network media allow variable maximum transmission units (MTUs), you must allow for the fragmentation of these transmission units into differently sized packets or cells. Hackers can take advantage of this fragmentation by dividing attacking packets into smaller and smaller portions that evade the IDS but cause an attack when reassembled by a target host.
Network-based intrusion detection (covered later in this chapter) relies on the analysis of traffic that is captured as it traverses the network from a source to its destination. If a hacker can establish an encrypted session with its target host using Secure Shell (SSH), Secure Socket Layer (SSL), or a virtual private network (VPN) tunnel, the IDS cannot analyze the packets and the malicious traffic will be allowed to pass. Obviously, this technique requires that the attacker establish a secure encrypted session with its target host.
Obfuscation, an increasingly popular evasive technique, involves concealing an attack with special characters. It can use control characters such as the space, tab, backspace, and Delete. Also, the technique might represent characters in hex format to elude the IDS. Using Unicode representation, where each character has a unique value regardless of the platform, program, or language, is also an effective way to evade IDSs. For example, an attacker might evade an IDS by using the Unicode character c1 to represent a slash for a Web page request.
Disguising an attack using Unicode, hex, or control characters to conceal an attack from an IDS is commonly referred to as obfuscation.