R&D

Increasing technology space has pressurized the orgainsational enviroment to safegraurd its network from outside as well as inside attack. Any malicious intrusion can dragdown a highly reputed organisation to the floors of defamation and even insolvency. Henceforth network security is one of the biggest challenge for organisation. Although traditional concepts of firewall and intrusion detction system is prevailing yet there is a need to secure the enviroment from the so called zero day attacks. Attackers every now and then generate exploits to penetrate the secure network enviroment and the existing known signatures are not able to detect such an attack. This paper brings forth the concept of deployment of honeypot so as to make the IDS system more effective and efficient for the zero day attack leading to least penetration to the network enviroment. This paper lays down a new network architecture so that the IDS system can be updated with the latest attack scenarios. The aforesaid objective is achived with the help of establishment of honeypot. Later tcpdump is used for he further analysis of payload so as to generate alert signautre in IDS(Snort v2.0).

SVM and ISVM are the most prominent technique for excellent learning and classification in the field of machine learning such as face detection and recognition, handwriting automatic identification and automatic text categorization, video tracking, Number plate recognition, Traffic Control etc. Nowadays, Face recognition is a challenging task in the field of computer vision. Motive of face recognition is to compare between the Given face database with the input face image and then declare a decision that identifies to whom the input image class belongs to or doesn't belong to the face database.In this analysis we not only study and also compare the sophisticated classification technique i.e. SVM and ISVM for face recognition along with its pros and cons.

Increasing technology space has pressurized the orgainsational enviroment to safegraurd its network from outside as well as inside attack. Any malicious intrusion can dragdown a highly reputed organisation to the floors of defamation and even insolvency. Henceforth network security is one of the biggest challenge for organisation. Although traditional concepts of firewall and intrusion detction system is prevailing yet there is a need to secure the enviroment from the so called zero day attacks. Attackers every now and then generate exploits to penetrate the secure network enviroment and the existing known signatures are not able to detect such an attack. This paper brings forth the concept of deployment of honeypot so as to make the IDS system more effective and efficient for the zero day attack leading to least penetration to the network enviroment. This paper lays down a new network architecture so that the IDS system can be updated with the latest attack scenarios. The aforesaid objective is achived with the help of establishment of honeypot. Later tcpdump is used for he further analysis of payload so as to generate alert signautre in IDS(Snort v2.0).

Classification technique is useful to categorize data into classes from the current dataset on the basis of a training set of data containing observations (or instances) whose category membership is known. The decision Tree classification algorithm doesn't work well separately for high dimensional data. To improve the efficiency, in our work, we apply principal component analysis and linear transformation together on the original data set for dimensionality reduction after that classification algorithm is applied for a better solution.

Image segmentation is the first step in image analysis and pattern recognition. It is a critical and essential component of image analysis system, is one of the most difficult tasks in image processing, and determines the quality of the final result of analysis The Segmented image require feature vector on the source of which pixel classification can be performed. If we depend only intensity values of image. Then it will not give the better segmentation result. Therefore, it requires some other features which are inherent in the image. The moment is an invariant feature used in the object recognition. The main aim is to classify each pixel of an image by novel feature. The idea of novel feature is to compute each pixel using window based method and then assign class labels using k-means classification technique. If we want to get colour map then colour map is obtained by performing two steps. First quantize the image colors second is labelling the pixels in image. Objective of color quantization is to obtain the image without significant perceptual degradation in restricted color range. Now label the pixels belonging to same quantized range with same label. In this way a color-map is obtained for all pixels in image. Aim is to Comparison between different window sizes with different image size. The basic difference between window images is 2*2 Gives not better result the other one like 3*3 and 4*4 but it is also depend on image size.It is varied with size of image each image of different window size give. As window size increases, smoothness of area considered by that window increases, the noise tolerance of proposed algorithm increases, the computational time of proposed algorithm increases and sharpness of area decreases. It also affects the values of moment computed for feature vector due to averaging of pixels of a window. All the experiments have been performed on both gray and color scale images in MATLAB version 7.10.

A very basic problem in reliability modeling is to obtain information about the form of the population from which the sample is drawn. Goodness of fit test is employed on different real data sets to determine how well the observed sample data "fits" on proposed model for reliability analysis. In this paper, we obtain the tables and graphs of critical values of Kolmogorov-Smirnov (KS) test, and Q-Q test for Gompertz model with two unknown parameters. These plots are used to investigate whether an assumed model adequately fits a set of data and we present power comparison between Computation of MLE using Newton-Raphson method and p-value with its corresponding D-value obtaining by KS-Test Q-Q test for model validation to obtain feasible real data sets which are most suitable for parameter estimation of Gompertz model. For this analysis, we used different tools which are developed in R language and environment for model analysis, model validation and estimation of parameters using method of maximum likelihood.

The Cloud computing is the way of using a network of remote servers hosted on the Internet to store, process and manage data, rather than on to the local server. Security is the most important concern in the cloud computing because the user data is transferring through the insecure medium i.e. Internet. In this paper, we present an overview of existing issues related to the cloud security and algorithms for finegrained access control and data security in the cloud environment. All these algorithms are described more or less on their own. Cloud security is a very popular task. We also explain the fundamentals of sequential rule mining. We describe today's approaches for cloud security. From the broad variety of efficient algorithms that have been developed we will compare the most important ones. We will systematize the algorithms and analyze their performance based on both their run time performance and theoretical considerations. Their strengths and weaknesses are also investigated. It turns out that the behavior of the algorithms is much more similar as to be expected.

SVM and ISVM are the most prominent technique for excellent learning and classification in the field of machine learning such as face detection and recognition, handwriting automatic identification and automatic text categorization, video tracking, Number plate recognition, Traffic Control etc. Nowadays, Face recognition is a challenging task in the field of computer vision. Motive of face recognition is to compare between the Given face database with the input face image and then declare a decision that identifies to whom the input image class belongs to or doesn't belong to the face database.In this analysis we not only study and also compare the sophisticated classification technique i.e. SVM and ISVM for face recognition along with its pros and cons.

A web service is defined as a software system designed to support interoperable machine-to-machine interaction over a network. Put in another way, Web services provide a framework for system integration, independent of programming language and operating system. Web services are widely deployed in current distributed systems and have become the technology of choice. The suitability of Web services for integrating heterogeneous systems is largely facilitated through its extensive use of the Extensible Markup Language (XML). Thus, the security of a Web services based system depends not only on the security of the services themselves, but also on the confidentiality and integrity of the XML based SOAP messages used for communication. Recently, Web services have generated great interests in both vendors and researchers. A web service, based on existing Internet protocols and open standards, and provides a flexible solution to the problem of application integration. This paper provides an overview of the web services, web service security and the various algorithms used for encryption of the SOAP messages.

Image fusion is one of the important embranchments of data fusion. Its purpose is to synthesis multi-image information in one scene to one image which is more suitable to human vision and computer vision or more adapt to further image processing such as target identification. This paper mainly compares the Scilab inbuilt module and novel method for image fusion. By using scilab as experimental platform, we approved the feasibility and validity of method. The result indicate that the fused image quality would be very effective and clear.

Cloud computing Environment provides the platform to share computing resources and provide different services like SaaS, PaaS and IaaS that will used by organization as private, public or Hybrid. Cloud computing basically known as Everythingas-a-service .Since cloud computing is used shared distributed resources over the world wide network (e.g., internet) in the open environment, thus it will create various security issues in cloud environment and their application. On our personal computer we have full control on data and processes but in cloud environment we used data and application services which are provided by some of other Cloud Service Providers (CSP) (e.g., AWS, Google docs, Go Grid etc.).We overcome these issues of security and privacy by using Radius server which is known as AAA server (Authentication, Authorization and Accounting). Using this we will secure our CSP main server that further link with Data Center which store database of various cloud users. We use client password to encrypt and decrypt their data when they access or store data in Cloud Data Center (CDC).

Attackers are using the malicious websites to launch the cyber attacks against the client's user applications which is one of the growing threat to the internet community. The growing threat to spread the internet attacks through malicious websites led to new technologies to defend against these kind of attacks and client honeypots is one of the these technologies. Client honeypots crawl the internet websites to find and to identify the malicious web servers. The technology called Active Honeypots (Honeyclients) is designed to act as security devices in search of malicious servers that attack client side applications. The main objective of Active Honeypot is to actively browse the cyber space to identify and enlist the malicious websites. This Paper presents the usefulness of client honey pots in terms of collection of malwares which are embedded into malicious websites , the comparative analysis of state based detection of malwares with signature based detection mechanisms , then the design and implementation of the system by incorporating the hybrid model of detection known as signature & state based malware infection detection. Both the detection mechanism is running in parallel mode and most of software tools used in this implementation is open source.

Recently, Vehicular ad hoc network (VANET) can offer various services and benefits to VANET users and thus deserves deployment effort. Intelligent Transportation Systems (ITS) are aimed at addressing critical issues like passenger safety and traffic congestion when integrates ICT with transportation infrastructure. This infrastructure is called Vehicular Ad hoc Networks (VANET), Vehicular communication systems facilitate communication devices for exchange of information among vehicles and between vehicles and roadside equipment. Working in tandem with the fielded Intelligent Transportation Systems (ITS) infrastructure, VANET will enhance the awareness of traffic by aggregating, propagating and disseminating up-to -the minute information about existing or impending traffic-related events. In support of their mission, Here we consider that every study should chose the appropriate simulator based on its requirements.