Title: Intrusion Detection Systems IDS
1Intrusion Detection Systems (IDS)
2What is IDS?
A security system designed to identify intrusive
or malicious behavior via monitoring of
activity. An IDS identifies suspicious patterns
that may indicate an attempt to attack, break in
to, or otherwise compromise a system. An IDS can
issue an alert or take a predefined action once a
suspicious pattern has been detected. For
instance, if a port scan is detected, then an
alert is issued to an administrator. This alert
could be in the form of an email, popup window,
log entry, or page.
3Why use an IDS system?
- To increase the chance of discovering a
compromised system. - To detect the threats of which your system may be
vulnerable. - To provide auditability in the event that an
intrusion takes place. - To serve as a deterrent against break in
attempts. - To provide documentation to justify the expenses
associated with new security technologies.
4Types of IDS systems.
- Network Based
- Host Based
- Application Based
5Network Based IDS
A network based IDS system is installed on a
computer or appliance and monitors the traffic of
a network segment. Typically an NIDS system is
attached to a network via a span port. Eg Snort
6Network Based IDS
SPAN (Switched Port Analyzer) - The SPAN on a
switch mirrors the traffic from a switched
segment onto a predefined SPAN port. A monitoring
device attached to the SPAN port can see all the
traffic from any of the other switched ports.
7Network Based IDS
- NIDS systems use signature matching to
determine if an attack has occurred or is
underway. NIDS systems try to match patterns in
network traffic to known signatures. - There are several verification methods that
can be used to detect attacks as well.
- Protocol Stack (IP, TCP, UDP)
- DOS, DDOS Use malformed packets
8Network Based IDS
- Application Protocol (Http, Ftp, Telnet)
- DNS Cache poisoning.
9Network Based IDS
- Advantages
- If placed appropriately and on a well designed
network, it can monitor a large number of hosts. - NIDS systems are usually passive. They can be
implemented with little to no disruptions. - NIDS systems are usually not detectable or
susceptible to attacks.
10Network Based IDS
- Disadvantages
- Can be overwhelmed by network volume and fail to
detect attacks. - Requires access to all network traffic to be
monitored, which may not be reliably obtained. - NIDS systems can not analyze encrypted traffic.
- Can not reliably determine if an attack was
successful. - Some form of attacks such as those involving
fragmented packets are not easily identified by
NIDSs.
11Host Based IDS
- A host based IDS system resides on a computer and
- monitors only activity on that machine. This type
of system - can be used to detect when a user creates,
deletes, or - modifies a monitored file.
- Eg Tripwire, CFengine
12Host Based IDS
- Advantages
- Can detect events local to a computer and some
attacks not detectable by NIDS. - Can process traffic before it is encrypted.
- Switched networks do not affect HIDSs.
- Adds audibility locally for detecting attacks or
for determining how an attack took place.
13Host Based IDS
- Disadvantages
- More management, since they are host based.
- Vulnerable to direct attacks and attacks against
the host operating system. - Not aware of multi-host attacks. Eg port
scanning - Requires a large amount of disk space for log
files. - Can cause performance overhead on the host system.
14Application Based IDS
- An application based IDS system is used to detect
- the abnormal behavior of an application.
- Eg Functions in MS Word used to detect malicious
VB scripts
15Application Based IDS
- Advantages
- It is aware of specific users. It can track
unauthorized activity of a user. - Can process data before it is encrypted and
leaves the application or after it has been
decrypted by the program.
16Application Based IDS
- Disadvantages
- Most susceptible IDS system to attacks.
- Less capable of detecting software tampering.
17Signature Based IDS
- Signature based IDS systems examine data for
patterns - and attempts to match them to know attacks.
- Eg Virus protection software
18Signature Based IDS
- Attacks typically have clear and distinct
patterns. - Footprinting and fingerprinting activities.
- Exploits involving a specific attack sequence.
- Port scans, DOS, DDOS attacks
- The problem with signature based IDS systems is
that they - must continually be updated so that they can
detect new - types of attacks. Another problem is that slow
attacks can - often go unnoticed.
19Statistical Anomaly Based IDS
- Statistical anomaly based IDS systems create a
normal - network activity baseline. Network activity is
then - periodically sampled. If activity is found that
is outside of - the baseline, then an alert is triggered.
- Disk, CPU, memory, and network usage can all be
used as - a baseline for comparison.
20Statistical Anomaly Based IDS
- Advantage
- A statistical anomaly based IDS system is that it
can - detect new types of attacks without requiring
constant - updates.
- Disadvantage
- Requires more overhead and processing then a
signature - based system.
21Log File Monitor
- A log file monitor examines logs from servers,
network - devices, and other IDSs for abnormal activity.
- Advantage
- It can scan activity across multiple hosts.
- Disadvantage
- Requires a lot of disk space for log files and
overhead for - processing.
22Extras
- NIDS System
- Snort
- http//www.snort.org/
- NIDS .avi demo
- Bro Intrusion Detection Monitor
- http//www.cse.ohio-state.edu/rwagner/cube.avi
- HIDS System
- Tripwire
- http//www.tripwire.org/
23Snort
- Rule
- alert tcp EXTERNAL_NET any -gt HOME_NET 111
(msg"RPC portmap proxy attempt TCP"
flowto_server,established content"00 01 86
A0" depth4 offset16 content"00 00 00
05" within4 distance4 content"00 00 00
00" depth4 offset8 classtyperpc-portmap-dec
ode sid1922 rev6) - Alert
- 119236 RPC portmap proxy attempt UDP
- Classification Decode of an RPC Query
Priority 2 - 02/25-104456.627405 164.107.112.3332846 -gt
164.107.112.255111 - UDP TTL64 TOS0x0 ID0 IpLen20 DgmLen144 DF
- Len 116