An intrusion prevention system is a network security device that monitors network and/or system activities for malicious or unwanted behavior and can react, in real-time, to block or prevent those activities. Network-based IPS, for example, will operate in-line to monitor all network traffic for malicious code or attacks. When an attack is detected, it can drop the offending packets while still allowing all other traffic to pass. Intrusion prevention technology is considered by some to be an extension of intrusion detection (IDS) technology. The term "Intrusion Prevention System" was coined by Andrew Plato who was a technical writer and consultant for * NetworkICE

Intrusion prevention systems (IPS) evolved in the late 1990s to resolve ambiguities in passive network monitoring by placing detection systems in-line. Early IPS were IDS that were able to implement prevention commands to firewalls and access control changes to routers. This technique fell short operationally for it created a race condition between the IDS and the exploit as it passed through the control mechanism. Inline IPS can be seen as an improvement upon firewall technologies (snort inline is integrated into one), IPS can make access control decisions based on application content, rather than IP address or ports as traditional firewalls had done. However, in order to improve performance and accuracy of classification mapping, most IPS use destination port in their signature format. As IPS systems were originally a literal extension of intrusion detection systems, they continue to be related.

Intrusion prevention systems may also serve secondarily at the host level to deny potentially malicious activity. There are advantages and disadvantages to host-based IPS compared with network-based IPS. In many cases, the technologies are thought to be complementary.

An Intrusion Prevention system must also be a very good Intrusion Detection system to enable a low rate of false positives. Some IPS systems can also prevent yet to be discovered attacks, such as those caused by a Buffer overflow.

IPS, Application Firewalls, Unified Threat Management & Access Control

The role of an IPS in a network is often confused with access control and application-layer firewalls. There are some notable differences in these technologies. While all share similarities, how they approach network or system security is fundamentally different.

An IPS is typically designed to operate completely invisibly on a network. IPS products do not typically claim an IP address on the protected network but may respond directly to any traffic in a variety of ways. (Common IPS responses include dropping packets, reseting connections, generating alerts, and even quarantining intruders.) While some IPS products have the ability to implement firewall rules, this is often a mere convenience and not a core function of the product. Moreover, IPS technology offers deeper insight into network operations providing information on overly active hosts, bad logons, inappropriate content and many other network and application layer functions.

Application firewalls are a very different type of technology. An application firewall uses proxies to perform firewall access control for network and application-layer traffic. Some application-layer firewalls have the ability to do some IPS-like functions, such as enforcing RFC specifications on network traffic. Also, some application layer firewalls have also integrated IPS-style signatures into their products to provide real-time analysis and blocking of traffic. Application firewalls do have IP addresses on their ports and are directly addressable. Moreover, they use full proxy features to decode and reassemble packets. Not all IPS perform full proxy-like processing. Also, application-layer firewalls tend to focus on firewall capabilities, with IPS capabilities as add-on. While there are numerous similarities between the two technologies, they are not identical and are not interchangeable.

Unified Threat Management (UTM), or sometimes called "Next Generation Firewalls" are also a different breed of products entirely. UTM products bring together multiple security capabilities on to a single platform. A typical UTM platform will provide firewall, VPN, anti-virus, web filtering, intrusion prevention and anti-spam capabilities. Some UTM appliances are derived from IPS products such as 3Com's X-series products. Others are derived from a combination with firewall products, such as Juniper's SSG or Cisco's Adaptive Security Appliances (ASA). And still others were derived from the ground up as a UTM appliance such as Fortinet or Astero. The main feature of a UTM is that it includes multiple security features on one appliance. IPS is merely one feature.

Access Control is also an entirely different security concept. Access control refers to general rules allowing hosts, users or applications access to specific parts of a network. Typically, access control helps organizations segment networks and limit access. While an IPS has the ability to block access to users, hosts or applications, it does so only when malicious code has been discovered. As such, IPS does not necessarily serve as an access control device. While it has some access control abilities, firewalls and network access control (NAC) technologies are better suited to provide these features.

Contrast with Intrusion Detection Systems (IDS)

IPS systems have some advantages over intrusion detection systems (IDS). One advantage is they are designed to sit inline with traffic flows and prevent attacks in real-time. In addition, most IPS solutions have the ability to look at (decode) layer 7 protocols like HTTP, FTP, and SMTP which provides greater awareness. However, when deploying network-based IPS (NIPS), consideration should be given to whether the network segment is encrypted since not as many products are able to support inspection of such traffic.

IPS can do more than just drop packets. Because an IPS is inline, it does not have to interpret the network stack. An IPS can correct CRC, unfragment packet streams, prevent TCP sequencing issues, and clean up unwanted transport and network layer options. Intrusion detection system evasion techniques were made famous by Insertion, Evasion, and Denial of Service: Eluding Network Intrusion Detection and can be addressed with IPS. IPS that have evolved from IDS tend to still have these issues for the software was designed with detection and not the concept of correction in mind.

Types

Host-based

A host-based IPS (HIPS) is one where the intrusion-prevention application is resident on that specific IP address, usually on a single computer. HIPS compliments traditional finger-print-based and heuristic antivirus detection methods, since it does not need continuous updates to stay ahead of new malware. As ill-intended code needs to modify the system or other software residing on the machine to achieve its evil aims, a truly comprehensive HIPS system will notice some of the resulting changes and prevent the action by default or notify the user for permission.

Extensive use of system resources can be a drawback of existing HIPS systems, which integrate firewall, system-level action control and sandboxing into a coordinated detection net, on top of a traditional AV product. This extensive protection scheme may be warranted for a laptop computer frequently operating in untrusted environments (e.g. on cafe or airport Wi-Fi networks), but the heavy defenses may take their toll on battery life and noticeably impair the generic responsiveness of the computer as the HIPS protective component and the traditional AV product check each file on a PC to see if it is malware against a huge blacklist. Alternatively if HIPS is combined with an AV product utilising whitelisting technology then there is far less use of system resources as many applications on the PC are trusted (whitelisted). HIPS as an application then becomes a real alternative to traditional antivirus products.

Network

A network-based IPS is one where the IPS application/hardware and any actions taken to prevent an intrusion on a specific network host(s) is done from a host with another IP address on the network (This could be on a front-end firewall appliance.)

Network intrusion prevention systems (NIPS) are purpose-built hardware/software platforms that are designed to analyze, detect, and report on security related events. NIPS are designed to inspect traffic and based on their configuration or security policy, they can drop malicious traffic.

Content-based

A content-based IPS (CBIPS) inspects the content of network packets for unique sequences, called signatures, to detect and hopefully prevent known types of attack such as worm infections and hacks.

Protocol Analysis

A key development in IDS/IPS technologies was the use of protocol analyzers. Protocol analyzers can natively decode application-layer network protocols, like HTTP or FTP. Once the protocols are fully decoded, the IPS analysis engine can evaluate different parts of the protocol for anomalous behavior or exploits. For example, the existence of a large binary file in the User-Agent field of an HTTP request would be very unusual and likely an intrusion. A protocol analyzer could detect this anomalous behavior and instruct the IPS engine to drop the offending packets.

Not all IPS/IDS engines are full protocol analyzers. Some products rely on simple pattern recognition techniques to look for known attack patterns. While this can be sufficient in many cases, it creates an overall weakness in the detection capabilities. Since many vulnerabilities have dozens or even hundreds of exploit variants, pattern recognition-based IPS/IDS engines can be evaded. For example, some pattern recognition engines require hundreds of different signatures (or patterns) to protect against a single vulnerability. This is because they must have a different pattern for each exploit variant. Protocol analysis-based products can often block exploits with a single signature that monitors for the specific vulnerability in the network communications.

Rate-based

Rate-based IPS (RBIPS) are primarily intended to prevent Denial of Service and Distributed Denial of Service attacks. They work by monitoring and learning normal network behaviors. Through real-time traffic monitoring and comparison with stored statistics, RBIPS can identify abnormal rates for certain types of traffic e.g. TCP, UDP or ARP packets, connections per second, packets per connection, packets to specific ports etc. Attacks are detected when thresholds are exceeded. The thresholds are dynamically adjusted based on time of day, day of the week etc., drawing on stored traffic statistics.

Unusual but legitimate network traffic patterns may create false alarms. The system's effectiveness is related to the granularity of the RBIPS rulebase and the quality of the stored statistics.

Once an attack is detected, various prevention techniques may be used such as rate-limiting specific attack-related traffic types, source or connection tracking, and source-address, port or protocol filtering (black-listing) or validation (white-listing).

Host-based vs. network

• HIPS can handle encrypted and unencrypted traffic equally, because it can analyze the data after it has been decrypted on the host.
• NIPS does not use processor and memory on computer hosts but uses its own CPU and memory.
• NIPS is a single point of failure, which is considered a disadvantage; however, this property also makes it simpler to maintain. However, this attribute applies to all network devices like routers and switches and can be overcome by implementing the network accordingly (failover path, etc.). A Bypass Switch from a vendor like Net Optics can be deployed to alleviate the single point of failure disadvantage though. Multi-segment Bypass Switches have recently become more popular as IPS vendors have rolled out high density solutions. This also allows the NIPS appliance to be moved and be taken off-line for maintenance when needed.
• NIPS can detect events scattered over the network (e.g. low level event targeting many different hosts, like hostscan, worm) and can react, whereas with a HIPS, only the hosts data itself is available to take a decision, respectively it would take too much time to report it to a central decision making engine and report back to block.
* * *

References

Testing Methodology http://www.breakingpointsystems.com/resources/testmethodologies/ips

Common Vulnerabilities and Exposures (CVE) http://www.cve.mitre.org/compatible/product.html

NIST SP 800-83, Guide to Malware Incident Prevention and Handling http://csrc.nist.gov/publications/nistpubs/index.html

NIST SP 800-31, Intrusion Detection Systems http://csrc.nist.gov/publications/nistpubs/index.html

Study by Gartner "Host-Based Intrusion Prevention Systems (HIPS) Update: Why Antivirus and Personal Firewall Technologies Aren't Enough" http://www.gartner.com/teleconferences/attributes/attr_165281_115.pdf

Open Source Intrusion Prevention Systems

• Snort NIPS
• Untangle NIPS
• GNOME Lokkit

External Resources

• Intrusion Prevention without False Positives - Blue Lane Technologies
• Intrusion Detection Resources - Philip Chan; CS Florida Institute of Technology

See also

• Intrusion-detection system
• Network intrusion detection system
• denial-of-service
• Host-based Intrusion Detection