Download SECURITY TECHNIQUE COMBINED WITH CONGESTION

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

AppleTalk wikipedia , lookup

Zero-configuration networking wikipedia , lookup

Computer security wikipedia , lookup

IEEE 1355 wikipedia , lookup

Wake-on-LAN wikipedia , lookup

Deep packet inspection wikipedia , lookup

Distributed operating system wikipedia , lookup

Distributed firewall wikipedia , lookup

Network tap wikipedia , lookup

Backpressure routing wikipedia , lookup

Computer network wikipedia , lookup

Wireless security wikipedia , lookup

Piggybacking (Internet access) wikipedia , lookup

IEEE 802.1aq wikipedia , lookup

Cracking of wireless networks wikipedia , lookup

Recursive InterNetwork Architecture (RINA) wikipedia , lookup

CAN bus wikipedia , lookup

Airborne Networking wikipedia , lookup

Peer-to-peer wikipedia , lookup

Routing wikipedia , lookup

Kademlia wikipedia , lookup

Routing in delay-tolerant networking wikipedia , lookup

Transcript
SECURITY TECHNIQUE COMBINED WITH
CONGESTION AVOIDANCE IN WIRELESS MESH
NETWORKS
Mankiran Kaur
Assistant Professor
QUEST, Jhanjehri Mohali
[email protected]
Jatinder Singh Saini
Assistant professor
BBSBEC, Fategarh sahib
[email protected]
ABSTRACT
Security in wireless network is one of the prime concern
in today’s Information Age, where information is an asset
not only to an organisation but also to an individual.
Security to a great extent is able to protect the network
from various unauthorized attacks. On the other side
implementation of security mechanisms also causes an
overhead in terms of increased load in the network.
Further the increased load in the network paves path to
congestion which degrades the performance of the
wireless network. In this paper we try to highlight various
challenges pertaining to security in mesh networks and
the ways of reducing security threats. We propose an
improved version of AODV which has a congestion
avoidance mechanism. We also use a security technique
called SBPGP for enhanced security of Mesh network.
Keywords – AODV, PGP, SBPGP
1. INTRODUCTION
Wireless mesh networks (WMNs) consist of wireless
nodes and wireless access points. WMN’s nodes are
composed of mesh routers and mesh clients. Each node
works as a host and router as well. Nodes in mesh
network consist of in built routing functionality which
forward packets to other nodes that are within their range
.WMN is self-organized network in which nodes
automatically establish connectivity among each other.
Nodes in the mesh network co-operate with each other in
the network to forward packets. The other protocols for
example MAC layer and network layer protocols usually
assume that the nodes that are taking part in
communication are honest and well-behaved with no
malicious or intention to make any harm to the network.
Usually in network some nodes may behave selfishly and
consists of unauthorized users. The nodes in the network
assumes that the node is trusted so and they start
communicating with them .In this way the malicious
nodes enter the network and can harm the network by
sending forge and fake messages thus initiating attack.
Attacks hampers the routing process of the network and
creates problem in route establishment process. So
network security is very important in wireless network.
2. SECURITY IN MESH NETWORK
WMN is exposed to various kinds of threats and risks
messages can be changed, there can be delay in
transmission, fake messages can be inserted. Resources
of the network can be accessed without authorized access
and can lead to complete denial of service (DOS) .So
Security in Mesh Network is very important The
Authentication and Key Management in Wireless Mesh
Networks is a crucial aspect of any security solution.
2.1 Security Protocols
Since various protocols available to tackle wireless
network security, but there is a chance of intrusion in the
network. Possible ways for securing network could be to
secure all wireless LAN devices Network can be secured
by implementing strong security in the network. Their
must be some mechanism in each node so that they can
identify malicious nodes that have the intention to harm
the network. Some of the techniques are explained below:
2.1.1 Mac Address Filtering
In the network there are various access points .These
access points contains some inbuilt technique that allows
the administrator to only give access to the node which
have a valid MAC ID’s. This method is good for
identification of the node but it has a drawback that id
can be hacked. Their are various cracking methods
available like SMAC .
2.1.2 Static Ip Addressing
The method of static IP addresses for the network devices
and end clients manually makes it tough for a attacker to
log on in the network. But this is not a robust mode of
securing the network. There are lot of tools available to
spoof the IP addresses. Using these spoofed IP intruder
can easily harm the network security.
2.1.3 Wepi
WEP (Wired Equivalent Privacy)[18][19] is a encryption
protocol which is based on the RC4 encryption algorithm,
with a secret key of 104 bits or it can be 40 bits combined
with a 24-bit Initialization Vector (IV) to encrypt the
plaintext message .Various users have observed many
drawbacks in WEP the use of RC4 is accepted as a strong
cryptographic cipher. However, Attackers can attack not
only on whole cryptographic algorithm, they can also
attack any weak point in the cryptographic system. But
technique of violating WEP has came into existence.
3. WMN SPECIFIC SECURITY
CHALLENGES
Due to the shared nature of wireless transmission
medium, the absence of globally central authority, and
the lack of security of mesh routers lead to main
challenges for securing WMNs. Firstly the Correct
routing messages are difficult to achieve in multihop
routing scenario, the most harmful kind of malicious
information is due to the fabricated routing
messages[23].Secondly, an authentication mechanism is
implemented with the help of Public Key Infrastructure
(PKI), which requires a globally trusted authority to issue
certificates. Having a globally trusted authority is not
possible all the time in Wireless Mesh network Thirdly,
the mesh devices are usually placed openly without any
protection so they are not physically protected So it
becomes very much easy for the intruder to have full
control over the device, thereafter launching attack from
the router and the data sent by router will be considered
as authenticated by other nodes. And if the device is
equipped with some cryptographic technique it is easily
broken by the attacker. Therefore some authentication is
required. Pretty good privacy is a cryptographic
technique which can applied in wireless networks. PGP is
explained as follow:
3.1 Pgp Based Security
Pretty good Privacy (PGP) is a security model that is used
for encryption and decryption of data which sends from
source to destination. It responsible for securely transmit
data from source to destination. It activates the central
authority (CA) for providing certificate to each node
when it is used in wireless ad-hoc network. It provides
security to ad-hoc network and improves the performance
of network. They use public key infrastructure (PKI) to
resist the collisions intentionally caused by malicious
nodes. It using certificates, digital signature and key
issuing can authenticate message, identify valid nodes or
malicious nodes. It is symmetric and asymmetric
cryptographic, web of trust model. This model fulfils
various security requirements such as authentication,
privacy, confidentiality and Non-repudiation of Mesh
network.. It protects the message (or its contents) from
being altered or destroyed. PGP algorithm (128 bits) is
implemented for security which is faster as well as secure
than previously implemented algorithms. It entails the
state level registration authority acting as certificate
authority (CA).
PGP security provides the same session key for
encryption and decryption between source and
destination node. In this each node request for the session
key from central Authority to communicate with
neighbouring node. Source sends the session key request
for neighbouring node then CA checks the authorization
of node. CA provides the session key for communication
between Source and neighbouring node. When source
sends the message to neighbouring node then it encrypt
the message with private key of source node and after
that whole message are encrypted with that session key.
So this process is continued until the destination node is
reached.
3.1.1 How PGP works
PGP combines the aspects of conventional authentication
and public key cryptography .In PGP user encrypts
plaintext with PGP .The PGP compresses the plaintext
.Compressing data saves transmission time of modem and
diskspace and enhances the cryptographic security .after
compressing the plaintext PGP creates a session key
,which is one time secret key .This one time secret key
works very securely with encryption algorithm to encrypt
plain text and results in ciphertext. After encrypting data
the one time secret key is encrypted with receivers public
key. Then ciphertext and public key encrypted secret key
is transmitted to the receiver. Decryption is reverse of this
Process.
3.2 Sbpgp Security Based On Pgp
Security Technique
In this work, we have applied the SBPGP model for
issuing PGP type certificate. Let us consider a MESH
network, where there are 4 access points forming a mesh
network and 3 clients are attached to each access point
forming a stationary network.
In our case there is one central authority for issuance of
certificates to access point and they further issue
certificate to the node nearest to access point than that
node further issue certificate to node immediate next
node . This process continues the node issuing the
certificate is called senior node and other junior node.
This junior node can be senior node to immediate next
node. Each node becoming senior node to next node in
the network for issuance of certificate. Each node in the
network maintains the information of senior node in its
routing table. Each nodes checks whether the node is
certified from senior nodes after verifying each others
certificate communication starts . Before entering the
network, each node obtains a public key certificate from
senior nodes .
Each node is equipped with detection mechanism to
identify Misbehaving nodes . SBPGP is based on PGP
type certification design.
3.2.1 Various
Certificate
Attributes
Of
Pgp
1)Certificate serial number
2)PGP Version
3)certificate holder Public key
4)Holder information
5)Digital sign [26] of certificate owner Key.
6)Verifying Signature
Type
In this technique authentication is done by verifying
certificate of certification authority its immediate senior
node from seniority table as seniority table is maintained
by each node.
3.3. Applying Security Technique
It has been observed that applying security technique in a
network overload the network with lot of route request
and buffer gets full due certificate issuance and
revocation procedure in network. Suppose one node is
performing certificate revocation process and at the same
time it gets route request from the other node this can
lead to buffer overflow and can cause congestion in the
network.
This can lead to packet loss ,and decreased throughput to
overcome this we have Improved AODV protocol .In this
new protocol when buffer gets full the node will follow
the alternative path through the node which is having free
buffer space.
4. Different Routing Protocols in
Mesh Network
Ad hoc routing protocols are usually
1) Reactive
2) Proactive
3) Hybrid.
A. Reactive Protocols
Reactive protocols are also known as on demand driven
routing protocols. They are called reactive protocols
because they Start route discovery by not by themselves,
route discover is done on demand when requested by
other nodes, when a source node send the request to
create a route. Route setup is done when demanded.
B. Proactive protocols:
In Proactive protocols, every node in the network
maintains routing table of itself and routing table of other
nodes in the network .Each node maintains most recent
routing information by sending control messages after
small interval among the nodes . The proactive routing
protocols uses link state routing information which
frequently links the information about neighbors . Some
of the existing routing proactive protocols are DSDV and
OLSR.
C. Hybrid Routing Protocol:
Hybrid routing protocol is a combination of proactive and
reactive routing protocols. In the beginning routing is
done with proactively prospected routes or predefined
routes and when any node wants to make connection
route establishment is done through reactive flooding
Some of the examples of hybrid protocols are TORA
protocol and ZRP protocol.
4.1 Aodv protocol
Ad-Hoc on Demand Distance Vector Protocol (AODV):
AODV is an reactive routing protocol, when a node in
the network wants to communicate with other node in the
network it sends the route request to other .Each node has
the topology information which is provided by AODV.
Control messages are used in the network to find a route
to the destination in the network.
Route Discovery Mechanism in AODV:
Suppose node “A” wants to communicate with another
node “B” , the first will send a Route Request message
(RREQ). This message is circulated through flooding
technique to the other nodes. The route request message
is forwarded to the neighbors, and the neighboring node
forward the route request message to their neighboring
nodes. This process continues until the node finds fresh
route to the destination with minimum possible hops
from itself .Once the route to the destination is found by
intermediate node it sends the route reply message to the
source node .As soon as RREP message reaches the
source node, a route establishment is done between
source node and destination node they start
communicating with each other. So there are four phases
of AODV protocol route request, route reply, route
establishment and route maintenance.
.
4.2 Iaodv Protocol
IAODV is enhancement of AODV protocol. In AODV
Congestion avoidance is not done when the load on the
one node increase its buffer starts overflowing and
congestion occurs. As we are applying security technique
so congestion in the network increases which results in
large packet loss. So congestion avoidance mechanism is
required. In IAODV we have added the congestion
avoidance technique to reduce packet loss and improve
overall performance of the network. Applying security
mechanism in a network results in increase load on the
network as node gets overwhelmed with data its buffer
gets full and starts overflowing So a mechanism is
required to monitor congestion level at each node to
overcome this problem we have applied congestion
detecting mechanism at each node in the network. In case
of congestion in the network we need to balance load to
reduce packet loss in the network. In the proposed work
we have applied congestion avoidance mechanism to
overcome this problem.
4.2.1 Two Main Changes we have proposed
a) Original AODV routing protocol is not resetting a
new shortest routing path during buffer overflow.
b) Congestion avoidance is done which is not done in
AODV .It is performed by creating a Cycle on a node
where the congestion probability is high . Each node
contains the routing table including information about its
own I.P. address, I.P. address of nearer neighbor nodes,
distance between the nodes, & queue length of each node
.
4.2.2.Congestion Avoidance
Description
Congestion is created when the capacity of the link or node
exceeds beyond its limit. when the rate of sending increased
and receiver is unable to receives the data as nodes
threshold limit has reached, than buffer starts overflowing
which results in long queuing delays and packet loss to the
large extent. So there is need to monitor incoming and
outgoing traffic across the link. In this work we have
considered downlink and uplink nodes of each node in the
network . Uplink nodes are those nodes through which
traffic is incoming and Downlink nodes are those nodes
with outgoing traffic .If number. of uplink and downlink
nodes are equal than the traffic is balanced as the node has
many options to route traffic through downlink nodes. But if
uplink nodes are more and there are less number of
downlink nodes, then congestion is created .Because than
the node have less options available nodes to route traffic
To overcome this problem .We have taken the ratio of
downlink and uplink nodes.
• We have propose the term Ratio of downlink and uplink
node (RDU) as the ratio of downlink and uplink nodes for
a particular node. Uplink nodes are depicting incoming
traffic and downlink node depicts outgoing traffic .
RDU=Total of downlink nodes/Total of uplink nodes
The individual RDU values at each node are used to
make forwarding decisions. When a node wants to
transmit packet it calculates its RDU before transmitting
Data.
1) If RDU is greater than one, it means that the node has a
greater number of downlink nodes in comparison with
uplink nodes. If so, it can implement any Fair Queuing (FQ)
mechanism and forward packet to appropriate downlink
node normal routing process is carried out.
.
• We propose a congestion avoidance scheme which
works well with multipath routing for WMN using the
RDU value. Queue monitoring is done to monitor
congestion level at each node.
The main steps in the proposed algorithm are:
1. Neighbour list maintenance : Each node in the
network maintain a list of uplink and downlink nodes.
This count is used to calculate the ratio of downlink
nodes and uplink nodes at each node. This neighbour list
and RDU values are stored at the network layer of a node.
Current queue sizes of each node in the list should also be
maintained in the neighbour list. Essentially the
neighbour list should comprise of three fields: (node id,
uplink/downlink, last updated queue size).
2. Queue length updation: Nodes need to regularly
advertise their current queue sizes to enable neighbour
nodes to update the queue sizes in their respective
neighbour tables.
3. Congestion avoidance at each node: The basic
mechanism of congestion avoidance is to use the RDU
values at each node along with the candidate node’s
queue length and forward packets to appropriate
candidate nodes. Instead of blindly forwarding the
packets to downlink nodes on the route, each node should
make an informed decision in this regard.
5. PROPOSED ALGORITHM
We will be implementing congestion avoidance on
congestion affected node as it will balance load on the
node and will check the free buffer space of neighbouring
node.
Let us consider a network where ni with 1 ≤ i ≤ N where
N is the total number of nodes, Li is the neighbour list
maintained at each node which holds the set of uplink and
downlink nodes followed by queue sizes of these nodes.
Li is the initial neighbour list
Bi is the buffer
2) If RDU is less than one, it depicts that there are more
uplink nodes as compared to downlink nodes. So rate
reduction is required to prevent congestion.
As the node’s queue fills up, it needs to inform
neighbouring uplink nodes to start sending fewer packets.
[24]Rate reduction scheme can be used such as the
congestion bit or using backpressure messages .
3)If RDU value is one, then the node checks queue sizes of
the downstream nodes and route packets fairly through them
it helps in avoiding congestion
Threshold = Qth
Ratio of uplink and downlink node = RDU
5.1 CONGESTION AVOIDANCE
ALGORITHM
Step1. Set the neighbour list of each node Li.
[Li=Downlink nodes, uplink nodes, Queue size of uplink
and downlink nodes
Step 2. Initially the buffer of each node is empty.
Step 3. When any node receives packet which require
forwarding than it stores the packet in its buffer which
increase its queue size and updation is send to the
neighbour list.
7. RESULTS
7.1 PACKET DELIVERY FRACTION
Step 4. Node checks its neighbourhood list Li arranged in
ascending order for making forwarding decision
Step 5. Node checks its current buffer size and compares
it with queue threshold
Step 6. If current buffer size is =Queue Threshold Limit
than calculate RDU which is the ratio of downlink and
uplink nodes in the network is calculated
a)if (RDU>1)
than the normal routing procedure is initiated end
if
b)if (RDU<1)
Than reduction technique is used to reduce sending rate
end if
c)if(RDU = 1)
Than the queue size of the downlink node is checked and
routing is done with least queue occupancy
Step 7. If current buffer size >Queue Threshold than
reduction technique is used to reduce sending rate
Step 8.If the current buffer size< Queue threshold than
the continue forwarding packets
Figure 1.
---------AODV
--------- IAODV
7.2 PACKET LOSS
6.SIMULATION
In this section, we simulate the seniority-based
authentication scheme using NS2 environment. The
simulation is conducted in 2200*2200 sq units..we are
having 4 access point each allocated with 3 client nodes
.every node is equal distance to other node .
Parameter
Simulation Time
No .of nodes
Traffic type
Pause time
Area
Data Packet size
Transmission range
Buffer size
MAC Protocol
Mobility model
Routing Protocol
Observation Parameters
Throughput, Packet loss
value
50 Sec
16
CBR
10 sec
2200*2200 sq.units
512 bytes
250m/sec
20 packets
802.11
Random waypoint
AODV,IAODV
Packet delivery ratio,
Figure 2.
---------AODV
--------- IAODV
7.3 THROUGHPUT
(WICOM 2007),21-25 September 2007, Shanghai,
China.
[2]. Nathan Lewis, Noria Foukia, and Donovan G.
Govan. “Using Trust for Key Distribution and Route
Selection in “Wireless Sensor Network” Proc. Of
IEEE/IFIP Network Operations and Management
Symposium(NOMS 2008), April 2008, Salvador, Brazil.
[3]. QIU Xiu-feng, LIU Jian-wei, Abdur Rashid Sangi.
“MTSR: Wormhole Attack Resistant Secure Routing for
Ad Hoc Network” Proc. of IEEE Youth Conference on
Information Computing and Telecommunications (YCICT), pp 419-422, 28-30 November 2010.
[4] Shi-Hong Chou, Chi-Chun Lo, Chun-Ch ieh Huang.
“Mitigating Routing Misbehavior in Dynamic Source
Routing Protocol Using Trust-Based Reputation
Mechanism for Wireless Ad-Hoc Networks” Proc. of 8th
Annual IEEE Consumer Communications and
Networking Conference - Security and Content
Protection
Figure 3.
-------AODV
-------IAODV
8. CONCLUSION
In this thesis, we have used the concept of trusted model
and authentication architecture in Wireless Mesh
Networks. The wireless Mesh network is more
susceptible to various kinds of attacks if proper security
mechanisms are not implemented. On the other side
adding security feature in the protocol has a decremental
affect on the network performance due to increased load
in the network. Keeping this in view we have
implemented the concept of buffer in AODV. Simulation
results show that buffering helps in balancing the load
across the network. It improves the performance metrics
like Packet loss, throughput and packet delivery ratio.
9. FUTURE SCOPE
In this paper we have evaluated the performance of
standard mesh-based architecture along with security
technique SBPGP and compared it with IAODV. It was
found that the proposed routing algorithm can be
efficiently work to avoid congestion in the network by
routing data through nodes with less queue occupancy.
We can apply different security technique in the Mesh
Network to enhance security in the network. It has been
observed end to end delay is increased applying
congestion avoidance technique. Dynamic path selection
routing mechanism can be applied to decrease end to end
delay and enhance overall network performance.
5.REFERENCES
[1]. Ding Xuyang, Luo Huiqiong. “Trust Evaluation
Based Reliable Routing in Wireless Mesh Network”
Proc. of IEEE International Conference of Wireless
Communication, Networking & Mobile Computing
[5].Shantanu Konwar, Amrita Bose Paul, Sukumar
Nandi, Santosh Biswas. “MCDM based Trust Model for
Secure Routing in Wireless Mesh Networks” Proc. of
IEEE World Congress on Information and
Communication Technologies (WICT), pp 910-915, 1114 December 2011, Mu mbai, India.
[6] U.Venkanna, R.Leela Velusami. “Black Hole
Attack and their counter measure based on trust
management in MANET: A Survey” In Proc. of
International Conference on Advances in Recent
Technologies in Communication and Computing (IET
2011)” ,
[7]. Naveen Kumar Gupta Kavita Pandey. “Trust Based
Ad-hoc On Demand Routing Protocol for MANET”
Proc. of IEEE Conference 2013.
[8]. Meenakshi Mehla, Himani Mann. “Sbpgp Security
Model Using Iodmrp” International Journal Of
Computational Engineering Research, Vol. 2,Issue No.3,
pp-823-828, May-June 2011.
[9]. Jashanvir Kaur and Er. Sukhwinder Singh Sran.
“SBPGP Security based Model in Large Scale Manets”
International Journal of Wireless Networks and
Communications, Volume 4, Number 1 (2012).
[10] Ranjeet Singh Harwant Singh Arri. “Analysis
of QOS Parameters of AAMRP and IODMRP using
SBPGP Security Model” International Journal of
Computer Applications, Volume 69– No.15, May 2013.
[11] Nidh Mittal
, Jan ish.
“Performance
Evaluation of
AODV and
DSDV under
Seniority,
Based Pretty Good Privacy Model
(SBPGP)” International Journal of Scientific &
Engineering Research, Volume 4, Issue 6, June-2013.
[12] Tanupreet Singh, Shivani Dua, Vikrant Das,
“Energy-efficient routing protocols in Mobile ad-hoc
Networks”, International journal of Advanced research in
Computer Science and Software Engineering, Vol. 2,
Issue 1, January 2012.
[13] P. Kuppusamy, K. Thirunavukkarsu and B.
Kalavathi, “A study and Comparison of OLSR, AODV
and T ORA Routing Protocols
in
Ad
hoc
Networks”, Proceedings of 3rd IEEE Conference on
Electronics Computer Technology (ICECT 2011), 8-10
April 2011.
[14] F. Maan , Y. Abbas and N. Mazhar,
“Vulnerability Assessment of AODV and
SAODV Routing Protocols against Network Routing
Attacks and Performance Comparison”, in IEEE Wireless
Advanced (2011).
[15]. Yih-Chun Hu, Adrian Perrig, and Dav id B.
Johnson, Member, “Wormhole Attacks in Wireless
Networks” IEEE Journal on selected areas in
Communications,.
[16] Majid Khabbazian, Hugues Mercier and Vijay K.
Bhargava, Wormhole Attack in Wireless Ad Hoc
Networks: Analysis and Countermeasure, Proc. of IEEE
Global Telecommunications Conference (GLOBECOM
'06), PP 1-6, San Francisco, CA, USA, 27 November - 1
December 2006.
[17] Stewart S. Miller “Wi-Fi Security” The McGrawHill Companies, 2003.
[18] Sultan Weatherspoon, “Overview of IEEE 802.11b
Security” Network Communications Group, Intel
Corporation.
[19] F.Maa Y.Abbas and N. Mazhar, “Vulnerability
Assessment of AODV and SAODV Routing Protocols
against Network Routing Attacks and Performance
Comparison”, in IEEE Wireless Advanced(2011).
[20] Yih-Chun Hu, Adrian Perrig, and Dav id B. Johnson,
Member, Wormhole Attacks in Wireless Networks, IEEE
Journal on selected areas in Communications.
[21] Majid Khabbazian, Hugues Mercier and Vijay K.
Bhargava, Wormhole Attack in Wireless Ad Hoc
Networks: Analysis and Countermeasure, Proc. of IEEE
Global Telecommunications Conference (GLOBECOM
'06), PP 1-6, San Francisco, CA, USA, 27 November - 1
December 2006
[22]W. Zhang. "Security Issues in Wireless Mesh
Networks", Wireless Mesh Networks, 2007
[23] Mohammad Z. Ahmad. "Congestion Avoidance and
Fairness in Wireless Sensor Networks", IEEE
GLOBECOM 2008 – 2008 IEEE Global
Telecommunications Conference, 11/2008
[25] Chou, Shi-Hong, Chi-Chun Lo, and Chun- Chieh
Huang. "Mitigating routing misbehaviour in Dynamic
Source Routing protocol using trust-based reputation
mechanism for wireless ad-hoc networks", 2011 IEEE
Consumer Communications and Networking
Conference (CCNC), 2011.
[26]. Jawaid Quamar. "A hybrid cryptography model for
managing security in dynamic topology of MANET",
2008 International Symposium on Biometrics and
Security Technologies, 04/2008