.3 Protocol overview
In this section, represent the overview of SARP; the Discovery phase, the usage
of the Node Information, the Transactions, the Trust Reporting, and the Defense against
attacks.
4.3.1 Discovery Phase
Once nodes are deployed in the field of interest, the nodes start the discovery phase
in which each node learns which nodes are its neighbours, and how far it is from the other
nodes in network.
4.3.2 Node Information
Each node stores certain information of neighbors in its own information tables.
Each node has two kinds of tables. First, Neighbor Table stores the node identifier, the
hop count, the parent and the trust for the neighbor. Second, Transaction Table contains
records that include information about the transaction identifier, the data source that iniJyothi Engineering College, Cheruthuruthy Dept. of CSE, May 2016
On-Off Attack Management Based on Trust 21
tiated the transaction, the number of packets in the transaction and other information that
is useful for trust reporting.
4.3.3 Transactions
A transaction is the unit of data transmission and of trust reporting [1] . When a
node has a stream of data to report to a sender node, and it starts transmitting a specific
number of packets. If the source node is not available to reach to the destination node
directly, the node transmits the data to a transaction parent node. This node is called as a
Forwarder, and this Forwarder also finds a transaction parent node until the data reaches
to the correct destination.There are two criteria for choosing the transaction parent node.
First, the node should be closer to the sender. The distance of each node is recognized
in the Discovery Phase, and the information is used to find the closer node. Second, the
node should be trusted. The Trust Framework is used for finding the most trusted node
among the neighbours of the Forwarder [10] . When there is no trusted node in the higher
hop count group, the transaction parent node can be a peer or child node to detour around
the untrusted node.
Therefore, there are four kinds of possible transaction parent nodes; closer and
trusted node, farther and trusted node, closer but untrusted node, and farther but untrusted
node [18] , [20]. First, the closer and trusted node is the best case that can transmit the
data to the destination fast. Second, the farther and trusted node is a possible transaction
parent node that can transmit the data to the destination, but bypass the shortest route.
Third, the closer but untrusted node is not a good choice as a transaction parent node
because the data can be lost. Last, the farther but untrusted is the worst case that can
contaminate the data and even uses longer route. SARP aims to transmit the stream of
data and trust reporting by using the closer and trusted node, which can transmit faster
than other possible transaction parent nodes. If there is no possible closer and trusted
node, however, SARP uses the farther and trusted node, and avoids the untrusted nodes
for the uncontaminated data and trust reporting [19].
4.3.4 Trust Reporting
SARP relies on updating trust of nodes based on performance of the nodes in transactions. When a transaction is completed that means a stream of data is delivered from
a source node to destination node. The trust reporting process begins in two steps; ReJyothi Engineering College, Cheruthuruthy Dept. of CSE, May 2016
On-Off Attack Management Based on Trust 22
Figure 4.4: Level Zero
questing Trust Reports and Processing Trust Reports. First, Requesting Trust Reports
begins when a transaction finished. The sender or the Forwarder sends this requesting
message to its transaction parent node, and the transaction parent node sends back its
one-hop report with its own report to the requester [17]. As a result of this, the requester
will have both one hop and two hop report, when it receives the full report from its transaction parent node. Second, Processing Trust Reports begins when the requester receives
the report from its transaction parent node. The requester, which is the sender or the
Forwarder, calculates the trust value of its one-hop and two hop neighbors depending on
the report.
4.3.5 Defense against attacks
SARP provides multiple defense schemes with its own features.
4.4 Design
Design of the system described with the help of data flow diagram which explain
the entire working flow consideration of the system. It consist of three levels. In zeroth
level, the basic structure of the system explained. In first level, the entire system flow
is explained in which trust module is the major component. In the second level, the
operation in the trust module is further explained.
Jyothi Engineering College, Cheruthuruthy Dept. of CSE, May 2016
On-Off Attack Management Based on Trust 23
Figure 4.5: Level One
Figure 4.6: Level Two
Jyothi Engineering College, Cheruthuruthy Dept. of CSE, May 2016
On-Off Attack Management Based on Trust 24
4.5 System Specification
– Front End : Java
– Back end : Mysql
– Jdk versions : Jdk 1.7
– IDE used Netbeans
4.5.1 Language and Tools Used
For implementing the system, Java language is used. Java used due to its easiness
and simplicity nature. Java is a general purpose high level language. It is an object
oriented language. And the system is implemented using Netbeans. It is a software
development platform written in Java.
Jyothi Engineering College, Cheruthuruthy Dept. of CSE, May 2016
On-Off Attack Management Based on Trust 25
CHAPTER 5
RESULTS
Trust of the nodes in the system is analysed based on its nature and behaviour in the
system. If any malicious nature detected trust of the node is reduced and data transfer is
not possible if the trust is very low. Thus the security of the system can be increased. The
system help to detect and prevent the On- Off attack in an effective manner by calculating
trust of each node in the system. The system consist of multiple node and there nodes are
initialized first into the network, then only the data transfer and other operation with other
nodes can be initiated. All nodes in the system is unique and specified with a particular
port number and ip address. In order to enter to a network, firstly the node should be
register to the network by providing its port and ip address. Then the details of the node
is recorded in the system. In next time, the node should login to the system by providing
the details that the node already given. According to the behaviour of the node trust value
is updated and accordingly the data transfer within each node is possible.