International Journal of Computer Networks and Applications

47 - 61

Authors: Sanjeev Kumar Sharma, Sanjay Sharma

MANETs are very advantageous and fruitful when there is no infrastructure or has been destroyed due to some reasons like earthquake, floods and so on. Data communication in MANETs are very challenging issue due to threats like limited battery power, limited bandwidth, mobility, routing and QoS and so on. Due to these challenges and scarcity of resources in MANETs, it’s very difficult to achieve high quality of service. QAODV, AODV, OLSR, DSDV and ZRP and so on, are some of very popular routing protocols for MANETs. Minimum no. of hops is the route selection criteria used by most of the routing protocols. This makes it necessary to consider QoS parameters to the routing protocols. Above mentioned protocols are not pure QoS routing protocol and insufficient to achieve high Qos because they do not consider parameters which will affect the QoS, Instead they only consider hop count as a route selection criteria. We present a new QoS based routing algorithm AODV-QSRP (AODV based QoS Routing Protocol) for Mobile Adhoc Netwok. AODV-QSRP is based on existing AODV and is reactive (On-Demand) in nature. AODV-QSRP attempts to provide high QoS for the real time applications while considering the various important QoS parameters like Bandwidth, Delay, Link Quality and Battery power for route selection. We have simulated and compared the performance of AODV-QSRP with QAODV, AODV, OLSR, DSDV and ZRP. To implement and simulating the result Network Simulator-2 (NS-2.32) on Fedora platform is used, which is an event driven and real-time simulator.

15 - 26

Authors: Vincent O. Nyangaresi, Solomon. O. Ogara, Silvance O. Abeka

In mesh networks, the sender machine is connected to the receiver machine via multiple paths. Efficient transmissions along these paths require proper link choice so as to quickly deliver the packets the destination. Poor link selection can lead to overutilization of some links while the other redundant links remain underutilized. Over-utilized links experience heavy congestions under peak hours. The transmission control protocol (TCP) employs congestion control algorithms to prevent transmitters from overloading the network with data. These algorithms include slow start, congestion avoidance, fast retransmit and fast recovery. The slow start algorithm is utilized during the initial communication phase while congestion avoidance, fast retransmit and fast recovery are reactionary algorithms one packet loss is detected. This paper aimed to analyze the behavior of TCP under these congestion control algorithms in wired mesh networks. The dimensions that were used for this analysis included three way handshake, packet loss, duplicate acknowledgements, segment retransmissions, recovery, I/O plots and time-sequence plots. The objective of this study was to practically understand how the TCP protocol detects and handles network congestions in mesh networks. To achieve this objective, an experimental research design was employed. It involved the practical design of experimental setups that were used to collect data that was analyzed to provide an explanation of the TCP congestion control mechanisms. The results obtained indicate that the TCP first carries out a three handshake before data transmission can take place. It was also observed that the receipt of three duplicate acknowledgements is interpreted by TCP to be packet loss caused by network congestion. Moreover, it was established that TCP initiates fast retransmit and fast recovery when packet loss is detected. The contribution of this paper lies in the fact that it provided a practical understanding of how TCP detects and reacts to mesh network congestion, a concept that is critical to network administrators in their quest for packet loss prevention over the TCP architecture. Towards the end of the paper, suggestions for developing better ways of congestion handling in mesh networks by use of round trip times as a basis for adaptive congestion detection and control are elaborated.

139-147

Authors: Prachi

Increase in volume and intensity of network attacks, forcing the business systems to revamp their network security solutions in order to avoid huge financial losses. Intrusion Detection Systems are one of the most essential security solutions in order to ensure the security of any network. Considering huge volumes of network data and complex nature of intrusions, the performance optimization of Network Intrusion Detection System became an open problem that is gaining more and more attention from the researchers nowadays. The objective of this paper is to identify a machine learning algorithm that provides high accuracy and real-time system application. This paper evaluates the performance of 15 different machine learning algorithms using NSL-KDD dataset on the basis of false discovery rate, average accuracy, root mean squared error and model building time. Firstly, 5 machine learning algorithms out of 15 are chosen on the basis of maximum accuracy and minimum error in WEKA. Simulation of these machine learning algorithms is performed using 10-fold cross validation. Thereafter, the best machine learning algorithm is selected on the basis of maximum accuracy and minimum model building time so that it can be readily implemented in real-time Intrusion Detection Systems.

108 - 117

Authors: Sukhkirandeep Kaur, Roohie Naaz Mir

Advancement in MEMS technology, networking and embedded microprocessors have led to the development of a new generation of a Wireless Sensor Network (WSN) that can operate in unattended and harsh environment depending upon application. WSN consist of large number of power-conscious devices called sensor nodes that detect and observe any physical phenomenon and can be used in a wide range of applications. Underlying topology plays an important role in the performance of the Wireless Sensor Network. Depending upon the application, deployment is performed either deterministically or randomly. WSN suffers from a lot of issues that includes energy conservation, scalability, latency, computational resources and communication capabilities. Energy efficiency is critical issue in Wireless Sensor Networks as nodes are equipped with limited power supply and nodes that bypass most of the traffic deplete their energy faster that leads to decreased network lifetime. Incorporating clustering in the network improves scalability, increase energy efficiency and reduces redundancy. A new clustering approach for WSN has been discussed that includes load balancing and improves energy efficiency by precise selection of CH's. Analysis and simulation results demonstrate the effectiveness of the proposed approach.

57-62

Authors: Sonal Telang Chandel, Sanjay Sharma

A routing protocol is used to determine the optimal path to forward the packets from source to destination. A separate set of protocols is designed for IPv4 & IPv6 network such as RIP, OSPF-v2, RIPng, OSPF-v3, and AODV. IPv4 is replaced by IPv6 (next generation internet protocol), but both the protocols are not suitable for each other, due to this they need to coexist for a long time. In this paper, the performance is evaluated for different routing protocols for IPv4 and IPv6 over wired, wireless and the hybrid network. Simulations are carried out on EXata/Cyber 2.1 Simulator/Emulator. The performance of routing protocols is measured on the basis of throughput, jitter, end-to-end delay and packet delivery ratio (PDR). By the simulation, on the basis of throughput and PDR performance of AODV (IPv4) is very good as compared with other routing protocols. It has been observed that due to the size of the IPv6 header, which is larger than IPv4 header the performance of routing protocols downgrades.