KERBEROS IMPLEMENTATION IN MANETS
Atta-ur-Rahman, Mureed Hussain
SZAB Institute of Science and Technology, Islamabad, Pakistan
Kahina Kabri, Dominique Seret
Department of Mathematics et Informatics, University of Paris 5, France
Keywords: Kerberos, MANETs, Authentication, Ah-hoc networks.
Abstract: In this paper implementation of Kerberos is proposed for Mobile Ad-hoc Networks (MANETS) for user
authentication and authorization. Kerberos uses symmetric cryptography with a trusted server to enable
secure authentication and key exchange between client nodes. Kerberos protocol is designed to provide
reliable authentication over open and insecure networks where communications between the hosts
belonging to it may be intercepted. So simply Kerberos is an authentication protocol for trusted hosts on un-
trusted networks. There are two approaches used in MANETS: proactive approach & reactive approach. In
proactive approach protocols are also known as traditional distributed shortest-path protocols which are
used to maintain the routes at all times based on periodic updates with high routing overhead. We have
implemented Kerberos concept with proactive approach using Optimized Link State Routing Protocol
(OLSR).
1 INTRODUCTION
There are two basic wireless network topologies
such as infrastructure based networks and
infrastructure less networks also known as ad-hoc
networks. The combination of both networks is
called Heterogeneous Wireless Network or
Heterogeneous Mobile Ad-hoc Networks. Mobile
nodes move in cells in mobile and ad-hoc networks.
Structure of mobile ad-hoc networks is same as
general mobile networks, but difference lies in their
ability to self-configuring network of mobile nodes.
Ad-hoc networks do not need any backbone
infrastructure support and are easy to deploy.
MANETS are useful when infrastructure is absent,
destroyed or impractical. Ad hoc networks such as
Bluetooth are networks designed to dynamically
connect remote devices such as cell phones, laptops,
and PDAs which communicate over radio and do not
need any pre installed communication infrastructure.
Communication can be performed if two nodes are
close enough to exchange packets.(Security in
Adhoc Networks)
In context of security in MANETs, apart from
traditional to these classical threats, there are various
special threats in MANETS, e.g. denial of service
attacks against the energy resource can be performed
by using any of the services a node is
offering.(Karygiannis, 2002). In MANETS security
and user authentication are still an unresolved and
challenging research area. In case of disaster areas it
is especially important to restrain unauthorized
nodes accessing the networks by implementation of
an authorization scheme (Security in Adhoc
Networks).
2 REQUIREMENTS AND
ARCHITECTURE
To fulfil the authentication mechanisms of MANETs
there are various considerations which need to be
taken into account:
1. The MANETs topology is table driven and
dynamic.
2. MANETs may be globally connected with other
networks.
3. The MANETs has an addressing scheme with
Duplicate Addressing Detection (DAD)
mechanisms.
161
ur-Rahman A., Hussain M., Kabri K. and Seret D. (2008).
KERBEROS IMPLEMENTATION IN MANETS.
In Proceedings of the International Conference on Security and Cryptography, pages 161-166
DOI: 10.5220/0001930401610166
Copyright
c
SciTePress
4. In OLSR MPR may also function as KDC
server.
5. The mobile nodes support security protocols
like IPSec.
6. Access to network resources is restricted to
only for authorized nodes.
7. The mobile nodes are preconfigured with
security credentials to perform their
authentication procedures.
For a best viewing experience the used font should
be Times New Roman, on a Macintosh use the font
named times, except on special occasions, such as
program code (Section 2.3.7).
3 OLSR IN ADHOC NETWORKS
OLSR is a proactive routing protocol for MANETS.
This protocol inherits the stability of link state
algorithm and has the advantage of having routes
immediately available when needed due to its
proactive nature. OLSR is an optimization over the
classical link state protocol, tailored for mobile ad
hoc networks.
The OLSR is a table driven and proactive
protocol, as it is used to exchange topology
information with other nodes of the network
regularly and proactively. The nodes which are
selected as an intermediate nodes call multipoint
relay (MPR) (2-Level Authentication Mechanisms,
2006).
Neighbour nodes announce this information
periodically in their control messages. The protocol
uses the MPR to facilitate efficient flooding of
control messages in the network. The advantage of
this approach is that connections are established
quickly. Multipoint relays reduce the size of the
control messages. This technique signicantly
reduces the number of retransmissions of broadcast
control messages. OLSR is characterized by two
types of control messages: neighborhood messages
and topology messages, called respectively Hello
messages and Topology Control (TC) messages.
MPRs have been shown below in solid black circles:
Figure 1: Multipoint Relays Selection (MPRs).
3.1 Neighbour Discovery
Each node must detect its neighbour nodes with
which it has a direct link. Due to the uncertainties in
radio propagation, a link between neighbouring
nodes may enable the transmission of data in either
one or both directions over the link. For this, each
node periodically broadcasts Hello messages,
containing the list of neighbours known to the node
and their link status, as shown in Fig 3.
Figure 2: Obtaining TGT.
Figure 3: Ticket Granting Mechanisms.
The MPR set is chosen so that a minimum of
one-hop symmetric neighbors are able to reach all
the symmetric two-hop neighbors. In order to
calculate the MPR set, the node must have link state
information about all one-hop and two-hop
neighbors. This information is, as already
mentioned, gathered from HELLO messages
Thus, Hello messages enable each node to dis-
cover its one-hop neighbors, as well as its two-hop
neighbors (the neighbors of its neighbors). Each
node of the network independently selects its
multipoint relays (Karygiannis, 2002)
Each node keeps a table of routes to all known
destinations, through its MPR nodes. Every node
periodically broadcasts list of its MPR Selectors
(instead of the whole list of neighbors). Upon receipt
of MPR information each node recalculates and
updates routes to each known destination.
SECRYPT 2008 - International Conference on Security and Cryptography
162
4 RELATED WORK
In context of secure protocols for sensor networks,
Wireless Encryption Protocol (WEP) is playing a
vital role to secure link-level data during wireless
transmission between clients and intermediate
devices. WEP protocol is not designed for end to
end data. It only works from node to intermediate
device (Karygiannis, 2002).
There are two types of authentication systems;
open system authentication and shared key
authentication system based on cryptography. The
open system intermediate devices accept the nodes
without verifying the identity. So there is an only
one way authentication mechanism in which only
intermediate devices are authenticated by the nodes
(Karygiannis, 2002). In shared key scenario a node
is allowed to join network with WEP shared key.
So the open system is highly vulnerable to
attacks and openly invites unauthorized users and
nodes. But in case of shared key authentication
which is also known as cryptographic approach
based on the fact that client has knowledge about the
shared secret.
802.11 standard supports privacy through the use
of cryptographic techniques for the wireless
networks. The WEP (Karygiannis, 2002) also uses
the RC4 symmetric key stream cipher algorithms to
generate the pseudo-random data sequences, and this
key stream simply adds modulo 2 to the data to be
transmitted. WEP protocol is applicable all over the
802.11 layers to protect the traffic such as TCP/IP,
IPX and HTTP (Karygiannis, 2002).
There are various problems in WEP protocol
reported by various group of computer security
specialists. These includes the passive attacks based
on the statistical data analysis for which integrity
can be compromised because of static WEP Key
which is shared for long time of period with plain
text frame transmission in WEP (Karygiannis,
2002). There is no user authentication in the WEP
protocols mean only Service Set Identifier (SSID)
identification occurs and nodes authentication is
simple and based on shared key.
Another authentication technique proposed by
Zhangyan (Security in Adhoc Networks) with the
help of implementation of external Certificate
Authority (CA) and tamper-resistant chip to support
ubiquitous security in the MANETS. This technique
uses broadcast blacklist and shared password to
normal nodes using broadcast encryption.
The external CA used for this purpose which can
issue public key pair and its certificates to every
node and publish public key, so there is trust model
based on CA between nodes. In case some nodes are
compromised, the external CA joins network to
broadcast the blacklist (compromised node list) and
new password to the legitimate devices. Another key
is issued by the CA through broadcast called
encryption root key and child key issued to the
legitimated nodes broadcast encryption root key is
used for encryption and child key is used for
decryption.
Another technique which is being employed in
(Security in Adhoc Networks) is tamper resistant
adhoc chip which can be embedded into any adhoc
node or device to support the external CA based
security solutions.
In this technique there is a problem that existing
nodes cannot be used for adhoc network services
because they have not any chip which will recognize
by external CA. So this approach is not appropriate
for existing devices and there should be special
nodes with tamper resistance adhoc chips.
Another approach is proposed by Andreas
Hafslund and Jon (2-Level Authentication
Mechanisms, 2006). In this approach a 2-level
authentication mechanism in an internet connected
MANET was proposed. In this approach they
proposed in level-1 authentication all the nodes will
be authenticated to access the local MANET service
or MANET network resources and in level-2
authentication nodes will be authenticated by the
external gateway to access the global internet. So in
this approach there are two levels of authentication
and there is big overhead. Therefore there are
chances of DOS and DDOS attacks on gateway
nodes or other attacks like IP spoofing. It will also
create some problems related to QOS.
5 PROPOSED WORK
The proposed technical measures involve the use of
trust model for secure and authorized user access
over the networks. The proposed system is based on
Kerberos protocol for MANETS. Kerberos is widely
used in windows system and is very helpful for user
authentication for windows operating system. In an
open network environment, a workstation cannot be
trusted to identify its users correctly to network
services. Kerberos provides an alternative approach
whereby a trusted third-party authentication service
is used to verify users' identities.
Kerberos is based on secret key distribution
model developed by Needham & Schroeder at
Massachusetts Institute of Technology (MIT) based
on symmetric cryptography. It is based on trusted
KERBEROS IMPLEMENTATION IN MANETS
163
server model for secure authentication and Key
Exchange between the clients (Zhang).
It is possible to run Kerberos authentication
software on more than one machine. However, there
is always only one master copy of the Kerberos
database. The machine where database is hosted is
called master machine or just the master while other
machines may possess read-only copies of the
Kerberos database, and are called slaves.
In t In this approach basic idea is long lived
memorized passwords with short lived session keys
(dominate the WEP static key) (Karygiannis, 2002).
So it must be kept highly secure.
Figure 4: Logging operations.
6 BASICS OF KERBEROS
Kerberos keeps a database of its clients and their
private keys. The private key is a large number
known only to Kerberos and the client it belongs to.
In case that the client is a user, it is an encrypted
password. Network services requiring
authentication, register with Kerberos, as do clients
wishing to use those services. The private keys are
negotiated at registration.
Since Kerberos knows these private keys, it can
create messages which convince one client that the
other is really who it claims to be. Kerberos also
generates temporary private keys, called session
keys, which are given to two clients and no one else.
A session key can be used to encrypt messages
between two parties.
AKDC: IDA || IDB || N1
KDCA:EKa[Ks||IDB||N1||EKb[Ks||IDA]]
(1)
AB: EKb[Ks||IDA]
BA: EKs[N2]
AB: EKs[f(N2)]
In the above scenario secure distribution of a
new session key for communications between A &
B is being taking place.
It becomes vulnerable to replay attack if an old
session key has been compromised “AB:
EKb[Ks||IDA]” can be resent convincing B that it is
communicating with A. BA: EKs[N2] & AB:
EKs[f(N2)] is used for modification to addresses
(time stamps and using an extra nonce)
In ordered to avoid the requirements of
synchronization (Woo and Lam, 1992) proposed a
structure defined as under.
Figure 5: KDC Implementation.
To avoid impersonation, severs must be able to
confirm identities, so there should be authentication
approach for user authentication so Authentication
Server (AS) is used to store the passwords of all
users. Authentication Server shares unique secret
keys with each server.
Another approach which is used is known as
Ticket Granting Service. Kerberos Server shares a
secret key with each user which is knows as master
key. Kerberos server invents a session key KAB
which encrypts master keys of both communicating
parties, known as ticket. Therefore master keys
derived from the user’s passwords and session keys
will be used for single session. And the ticket
granting ticket is sent to KDC to acquire the session
key for communication between two parties, as
shown in figure 5.
7 KERBEROS IN MANETS
In MANET MPR or master node will act as
Kerberos node that is authentication server (AS) and
SECRYPT 2008 - International Conference on Security and Cryptography
164
will share unique key with each server and Ticket
Granting Server (TGS) as well, which will share a
secrete key with each user known as Master Key
(MK).
MPR or master node which is now working as
kerberos server invents a session key K
AB
when a
user A informs to MPR it wants to talk B, then K
AB
is encrypted with A’s Master Key(MK) & with B’s
Master Key (MK) as well known as ticket. User B
can decrypt the K
AB
and user A’s name. Master Key
(MK) is derived from user’s password. Session Key
SA is used by user A for a single session, which will
be valid for a small time. Node on behalf of user A
asks the Authentication Server (AS) for a session
key SA. SA is transmitted & encrypted with user
A’s MK (Master Key). The MPR (AS) also sends
TGT, encrypted with Kerberos Server Master Key
(MK). SA use A’s name and TGT expiration time.
TGT is sent to KDC to acquire the session key for
communication between two parties. TGT Server
and AS are collocated, means both need to use the
same information. Generally Kerberos uses the DES
algorithms. Users passwords converts into DES Key
& decrypts the information. Once getting the key,
Master key (MK) is discarded and only retains the
TGT & Session Key (S). The scenario is discussed
as under. All ticket generating mechanisms are show
in figures 2, 3 and 4.
TGS decrypts the TGT and checks the expiration
time then generates KAB with name of user A and
expiration time and encrypts with user B’s master
key KB and all this will be encrypted with SA.
Ticket grating scenario is defined as under.
User B keeps track of the recent time stamps.
The following scenario shows the login operation of
user B.
Figure 6: Realms replication.
7.1 Secure Authentication
TGS issues tickets to users who have been
authenticated to AS thus only the correct user with
the password can acquire ticket. Replication is
necessary to avoid a single point of failure in case of
single Kerberos Server so there is need of multiple
Kerberos Servers. Those will share same Master
Key and identical databases. Kerberos Server
maintains the master copy, and all other sites will be
updated periodically.
7.2 Concepts of Realms in MANETS
Each Kerberos realm has a master Kerberos
machine, which houses the master copy of the
authentication database. It is possible (although not
necessary) to have additional, read-only copies of
the database on slave machines elsewhere in the
network. The advantages of having multiple copies
of the database are those usually cited for
replication: higher availability and better
performance. If the master machine is down,
authentication can still be achieved on one of the
slave machines. The ability to perform
authentication on any one of several machines
reduces the probability of a bottleneck at the master
machine (CISCO).
Keeping multiple copies of the database
introduces the problem of data consistency. We have
found that very simple methods suffice for dealing
with inconsistency. The master database is dumped
every hour. The database is sent, in its entirety, to
the slave machines, which then update their own
databases. A program on the master host, called
kprop, sends the update to a peer program, called
kpropd, running on each of the slave machines. First
kprop sends a checksum of the new database it is
about to send. The checksum is encrypted in the
Kerberos master database key, which both the
master and slave Kerberos machines possess. The
data is then transferred over the network to the
kpropd on the slave machine, s shown in figure 6.
The slave propagation server calculates a checksum
of the data it has received, and if it matches the
checksum sent by the master, the new information is
used to update the slave's database (CISCO). All
passwords in the Kerberos database are encrypted in
the master database key Therefore, the information
passed from master to slave over the network is not
useful to an eavesdropper. However, it is essential
that only information from the master host be
accepted by the slaves, and that tampering of data be
detected, thus th checksum (CISCO).
Kerberos introduced RC4-HMAC support, which
is also present in Windows and is more secure than
DES. Among the supported encryptions (but not by
Windows) the triple DES (3DES) and newer
KERBEROS IMPLEMENTATION IN MANETS
165
AES128 and AES256 are worth mentioning. The
Kerberos protocol is to prevent the user's password
from being stored in its unencrypted form, even in
the authentication server database. Considering that
each encryption algorithm uses its own key length, it
is clear that, if the user is not to be forced to use a
different password of a fixed size for each
encryption method supported, the encryption keys
cannot be the passwords. For these reasons the
string2key function has been introduced, which
transforms an unencrypted password into an
encryption key suitable for the type of encryption to
be used. This function is called each time a user
changes password or enters it for authentication. The
string2key is called a hash function, meaning that it
is irreversible: given that an encryption key cannot
determine the password which generated it. Famous
hashing algorithms are MD5 and CRC32 is used.
8 CONCLUSIONS
We have proposed a trust model to solve the user
authentication problems in the MANET. This
mechanism is designed to provide reliable
authentication over open and insecure networks
where communications between the hosts belonging
to it may be intercepted in the MANET. Our idea
has the advantages of high security, flexibility and
practicality. It can be treated as base-bone to
implement secure authentication in the ad hoc
networks.
REFERENCES
Tom Karygiannis, Les Owens “CSD, NISTE
Gaithersburg”, MD 2098, 8930 (2002).
Security in Adhoc Networks by Zhang Yan, networking
laboratory Helsinki University of Technology.
2-Level Authentication Mechanisms in an Internet
Connected MANET 6
th
Scandinarian Workshop on
Wireless Adhoc Networks (2006).
Yanchoa Zhang, New Jersey Institute of Technology.
www.indengines.com Sunnyvale, California 94089
USA. J. Broch, D. Johnson,
Jennifer G. Steiner, Project Athena, Massachusetts
Institute of Technology, Cambridge, MA 02139
www.cisco.com/warp/public/106/1.html#intro
Z. J. Haas, M. R. Pearlman, P. Samar, “The Zone Routing
Protocol (ZRP) for Ad Hoc Networks,” draft-ietf-
manetzone-zrp-02.txt, IETF, (2000).
P. Jacquet, P. Muhletaler, A. Qayyum, A. Laouiti, T.
Clausen, L. Viennot, “Optimized Link Stat e Routing
Protocol”, in IEEE INMIC, Pakistan ( 2001).
A. Qayyum, L. Viennot, A. Laouiti, “Multipoint Relaying:
An Efcient Technique for ooding in Mobile Wireless
Networks” Tech. Rep. 3898, INRIA, http://
www.inria.fr, (2000).
Maltz, “The Dynamic Source Routing Protocol for Mobile
Ad Hoc Networks”, draft-ietfmanet-dsr-01.txt, IETF
1998.
J. Broch, D. A. Maltz, D. B. Johnson, “Supporting Hiear-
archy and Heterogeneous Interfaces in Multi-hop
Wireless Ad Hoc Networks”, Workshop on Mobile
Computing, IEEE International Symposium on
Parallel Architectures, algorithms and Networks,
Australia pp. 75–85, (1999).
Mounir Benzaid, Pascale Minet, Khaldoun AlAgha,
Cedric Adjih and Geraud Allard Integration of
Mobile-IP with OLSR for Universal Mobility.
T. Y. C. Woo and S. S. Lam. Authentication for
distributed Systems Computer, 25(1):39–52, January
1992.
SECRYPT 2008 - International Conference on Security and Cryptography
166