With the advent of new applications, different service needs come up. These needs could be in the form of reliability in delivering data, capacity amount in a particular range and certain amount of permissible delay. In order to provide high Quality of service to Networks, it is essential to provide a path between a given source and multiple destinations which satisfy certain constraints. For a domain catering to high QoS, there is a request of resources with certain constraints by all the applications. Speed and Scalability which are not flexible in terms of Network size and Topology are the basic issues to be considered here. Multimedia applications in general make use of k shortest paths whenever communication is to be carried out between a single source and one or more than one destination. In this paper, a genetic algorithm is used, which helps in determination of k shortest paths from a source node to more than one destination node, with bandwidth constraint. The algorithm makes use of the connection matrix as well as link bandwidth for determination of k shortest paths. The significance of using K shortest paths in a network is to increase Throughput and Packet delivery ratio.

In today’s world tremendous amount of unstructured data, especially text, is being generated through various sources. This massive amount of data has lead the researchers to focus on employing data mining techniques to analyse and cluster them for an efficient browsing and searching mechanisms. The clustering methods like k-means algorithm perform through measuring the relationship between the data objects. Accurate clustering is based on the similarity or dissimilarity measure that is defined to evaluate the homogeneity of the documents. A variety of measures have been proposed up to this date. However, all of them are not suitable to be used in the k-means algorithm. In this paper, an extensive study is done to compare and analyse the performance of eight well-known similarity and dissimilarity measures that are applicable to the k-means clustering approach. For experiment purpose, four text document data sets are used and the results are reported.

Mobile Ad-Hoc networks (MANETs) can be classified as Decentralized, Independent and Self- Organizing dynamic networks of intellectual movable nodes. In such networks, devices are connected by provisional wireless links. Dynamic topology imposes challenges in developing an efficient routing protocol, for enabling successful communication between mobile devices. Based on the nature of working, proactive & reactive protocols are the two broadly classified categories of routing protocols. This paper presents, relative experimental analysis of proactive routing protocols viz., Optimized Link State Routing Protocol (OLSR) and its variant (Kenji Yamada et al., 2010) with Cooperative Multi-Point Relay (MPR) Selection. To compare OLSR and its variant protocol, the Network Simulator- 2.35 is used to carry out numerous simulations, on arbitrary scenarios, by varying the number of network nodes & mobility of nodes. As per the simulation outcomes, the OLSR with a cooperative MPR selection has outperformed the traditional OLSR protocol in static scenarios or when the network load has been varied. On the contrary, the traditional OLSR protocol has performed better in mobile scenarios. But, as demonstrated from various experimentations, it exhibits higher Routing Overheads as compared to its variant protocol. Further, on the basis of simulation results, efforts can be made in the direction of performance optimization of OLSR and its variant protocol, to improve its performance in highly mobile scenarios as well, keeping in view other performance metrics.

Internet reliability and performance is based mostly on the underlying routing protocols. The current traffic load has to be taken into account for computation of paths in routing protocols. Addressing the selection of path, from a known source to destination is the basic aim of this paper. Making use of multipoint crossover and mutation is done for optimum and when required alternate path determination. Network scenario which consists of nodes that are fixed and limited to the known size of topology, comprises the population size. This paper proposes a simple method of calculating the shortest path for a network using Genetic Algorithm (GA), which is capable of giving an efficient, dynamic and consistent solution in spite of, what topology, changes in link and node happen and volume of the network. GA is used in this paper for optimization of routing. It helps us in enhancing the performance of the routers.

Internet plays a vital role in communication. Determination of internet capability is done by Routing protocol. After a network fails, routing protocols have very slow convergence rate, which is a grave problem and needs to be tackled. Multiple Routing Configuration (MRC) is a technique which helps IP networks to recover very quickly from link and node failures. In MRC, packet forwarding persists on an optional link as soon as a failure is detected and additional information is always contained in the routers. This paper discusses the effect of packet size on throughput, packet delivery ratio, packet loss and delay for various routing protocols like OSPF, OSPF with 1 and 2 link breakage and MRC.

K-Modes is an eminent algorithm for clustering data set with categorical attributes. This algorithm is famous for its simplicity and speed. The K-Modes is an extension of the K-Means algorithm for categorical data. Since K-Modes is used for categorical data so 'Simple Matching Dissimilarity' measure is used instead of Euclidean distance and the 'Modes' of clusters are used instead of 'Means'. However, one major limitation of this algorithm is dependency on prior input of number of clusters K, and sometimes it becomes practically impossible to correctly estimate the optimum number of clusters in advance. In this paper we have proposed an algorithm which will overcome this limitation while maintaining the simplicity of K-Modes algorithm.

Mobile ad-hoc networks are autonomous networks of intelligent mobile nodes. Designing an efficient routing protocol with goal efficient route establishment is still a burning research issue. Routing protocols are broadly classified into proactive and reactive protocols on the basis of their nature of working. In this paper, we present comparative simulation analysis of two proactive protocols namely, Optimized Link State Routing (OLSR) and Destination-Sequenced Distance Vector (DSDV). Simulations have been carried out multiple times using Network Simulator-2.35 (NS-2.35), on random scenario patterns and compared in two environments (OLSR and DSDV) by varying the network size and mobility of nodes. We have patched and installed OLSR protocol on NS-2.35 as it’s not available as a part of NS-2.35 installation. The simulation results indicate that, OLSR perform better than DSDV protocol for application oriented metrices such as packet delivery fraction, packet loss and end-to-end delay. But, Routing Overheads is significantly consistently higher for OLSR protocol in all the cases considered, which is proven experimentally. Furthermore, based upon the simulation results, work can be done in the direction of improving the performance of the OLSR protocol to make it a choice of proactive protocol for large and denser network.