With the rapid development of high-speed wide-area networks and powerful yet low-cost computational resources, grid computing has emerged as an attractive computing paradigm. It provides resources for solving large scientific applications. It is typically composed of heterogeneous resources such as clusters or sites at different administrative domains connected by networks with widely varying performance characteristics. The service level of the grid software infrastructure provides two essential functions for workload and resource management. To efficiently utilize the resources at these environments, effective load balancing and resource management policies are fundamentally important. This paper addresses the problem of load balancing and task migration in grid computing environments. We propose a fully decentralized two-level load balancing policy for computationally intensive tasks on a heterogeneous multi-cluster grid environment. It resolves the single point of failure problem which many of the current policies suffer from. In this policy, any site manager receives two kinds of tasks namely, remote tasks arriving from its associated local grid manager, and local tasks submitted directly to the site manager by local users in its domain, which makes this policy closer to reality and distinguishes it from any other similar policy. It distributes the grid workload based on the resources occupation ratio and the communication cost. The grid overall mean task response time is considered as the main performance metric that need to be minimized. The simulation results show that the proposed load balancing policy improves the grid overall mean task response time.

The keyloggers are covert security threat to the privacy and identity of users. The attackers are exploring different techniques of keylogging using hardware keyloggers, software keyloggers and screen capturing software to steal the user sensitive data. The Incognizance of the user is imposing greater risk. To overcome this problem, we have proposed a model. In this model solution to Keylogger and Screen Recording Software has been proposed by using the concept of fabricated password on untrusted machine. It deceives the untrusted system's key logging and video capturing software. The concept used here is "WYSINT What You See Is Not True". The main feature of this model is that it has a hardware recognition to retrieve the key .This key is required by the Temporary Filter layer (TFL) as an intermediary to change into the trusted password after bypassing all the capturing techniques and returning the original password to the required website.

Wireless sensor networks in terms of energy sources are limited. Furthermore due to this type of network infrastructure wireless communications and channel errors not possible to reach the correct packet to the destination exists; hence the proposing algorithms to improve the quality of service in these networks and sending packets are very important. In this paper we proposed a reliable and energy aware packet delivery mechanism to ensure quality of service in wireless sensor networks. In our proposed algorithm to ensure that a packet of information sent to the destination, the multi-path Forwarding method is used; So that several copies of an information packet via separate routes are sent to the destination, also routing decisions in this way occurs by considering the remaining energy in the neighborhood of nodes that are located in two hop of sender node. Simulation results show that the rate of release of data packets reduced in this way and thus the reliability of packet is increased, also the energy efficiency of sensor nodes effectively improved. Therefore this algorithm increase overall lifetime in wireless sensor networks.

This paper presents a methodology for providing high availability to the demands of cloud's clients. To attain this objective, failover stratagems for cloud computing using integrated checkpointing algorithms are purposed in this paper. Purposed strategy integrate checkpointing feature with load balancing algorithms and also make multilevel checkpoint to decrease checkpointing overheads. For implementation of purposed failover strategies, a cloud simulation environment is developed, which has the ability to provide high availability to clients in case of failure/recovery of service nodes. \The primary objective of this research work is to improve the checkpoint efficiency and prevent checkpointing from becoming the bottleneck of cloud data centers. In order to find an efficient checkpoint interval, checkpointing overheads has also considered in this paper. By varying rerun time of checkpoints comparison tables are made which can be used to find optimal checkpoint interval.
The purposed failover strategy will work on application layer and provide highly availability for Platform as a Service (PaaS) feature of cloud computing.

In 4G-domain multi home mobile host (MH) will have access to different wireless networks. In this 4G wireless networks real time applications are initiated to support services like VOIP, videoconference. These applications use RTP over UDP. These protocols are unresponsive to congestion events. Our proposed congestion management scheme minimizes congestion level in 4G domains using coalition game. Its main objective is to decide target networks for handoff with minimum congestion level. Our mechanism is based on the coalition game formation. It aims at maximizing the utilization of the resources available and meeting QoS requirement of users as much as possible. This will reduce congestion level with vertical handoff initiation.

In Hill Cipher, the plain text is divided into equal sized blocks. The blocks are encrypted one at a time. Cipher text only Crypto analysis of Hill Cipher is difficult. But it is susceptible to known plain text attack. In this work, Hill Cipher is improvised to make it more secure. The output of hill cipher is randomized to generate multiple cipher texts for one plain text. Any one cipher text is used for transmission of data. This approach thwarts any known plain text attacks and also chosen cipher text attacks.

Clustering is related to data mining for information retrieval. Relevant information is retrieved quickly while doing the clustering of documents. It organizes the documents into groups; each group contains the documents of similar type content. Document clustering is an unsupervised approach of data mining. Different clustering algorithms are used for clustering the documents such as partitioned clustering (K-means Clustering) and Hierarchical Clustering (Agglomerative Hierarchical Clustering (AHC)). This paper presents analysis of Suffix Tree Clustering (STC) Algorithm and other clustering techniques (K-means, AHC) that are being done in literature survey. The paper also focuses on traditional Vector Space Model (VSM) for similarity measures, which is used for clustering the documents. This paper also focuses on the comparison of different clustering algorithms. STC algorithm improves the searching performance as compare to other clustering algorithms as the papers studied in literature survey. The paper presents STC algorithm applied on the search result documents, which is stored in the dataset. This paper articulates the key requirements for web document clustering and clusters would be created on the full text of the web documents. STC perform the clustering and make the clusters based on phrases shared between the documents. STC is faster clustering algorithm for document clustering.