Information Technology

This paper presents a queuing model to evaluate the latency associated with file transfers or replications in peer-to-peer (P2P) computer systems. The main contribution of this paper is a modelling framework for the peers that accounts for the file size distribution, the search time, load distribution at peers, and number of concurrent downloads allowed by a peer. We propose a queuing model that models the nodes or peers in such systems as M/G/1/K processor sharing queues. The model is extended to account for peers which alternate between online and offline states. The proposed queuing model for the peers is combined with a single class open queuing network for the routers interconnecting the peers to obtain the overall file transfer latency. We also show that in scenarios with multipart downloads from different peers, a rate proportional allocation strategy minimizes the download times.

Object detection, tracking and recognition currently generated much attention in computer vision community due to their real time applications. The main goal is to detect and track specific instances of an object or broader object classes, such as cars or pedestrians in images sequence. Due The potency of processing video database lies on the search techniques employed in the video processing system. Usage of improper search techniques will make the processing system feeble. This paper proposes unique object detection and tracking system where Local Binary pattern are employed to detect moving object and track them in successive frames. Detecting a specific object from video suing Local Binary Pattern is more suitable when, objects information is not available. Initially, the database video clips are segmented into diverse shots before the feature extraction process. The proposed system comprises two stages: (i) feature extraction and (ii) tracking of object in the video clips. In the feature extraction stage, the color feature is extracted first based on the color quantization. Next, the edge density feature is extracted for the objects present in the query video. Subsequently, the texture feature is extracted using LGXP (local Gabor XOR patterns) technique. Eventually, the object is detected based on those features extracted and the detected object is tracked by utilizing both forward and backward tracking method. The proposed methodology is proved to be more effectual and precise in object detection and tracking.

SHARON based Furtive packet dropping is a suite of four afflictions misrouting the distance, clone node, and colluding collision that can be easily established against multi-hop wireless ad hoc networks. Furtive packet falling interrupts the packet from reaching the target through despiteful behavior at an intermediate node. Nevertheless, the entire vindictive node gives the impression to its neighbors that it performs the morganatic forwarding action. Furthermore a legitimate node comes under intuition. We introduce a protocol called FAWAN that can observe and isolate furtive packet dropping affliction efficiently. FAWAN presents two techniques that can be overlaid on baseline local monitoring: having the neighbors maintain additional information about the routing path, and adding some FAWAN provides an innovative mechanism to better a commute local monitoring by considerably increasing the number of nodes in a neighborhood that can do monitoring.

Software protection is an area of active research in which the software industry is encountering a number of threats. Software piracy is one such threat, which proves detrimental in protecting the intellectual property rights. There have been a variety of techniques developed to address the issue like software watermarking, code obfuscation, and tamper-proofing. In the current research we address the issue of software piracy through a prevention technique known as software watermarking which aims at providing copyright protection and authorized access of commercial software. In this paper fragile software watermark is used to embed personal information into the software. Then this personal information is merged with the hardware parameters of the client machine extracted during the process of installation and License key provided by the vendor. This combined string (Watermark + Hardware parameters + key) is send to the server for registration. This process is implemented and tested on different machines and the accuracy of the proposed model is found to be 99%. The proposed model will be beneficial in combating software piracy and securing the software from redistribution.

E-commerce has become one of the most commodious methods of shopping because of the technological revolution. Research shows internet shopping is much preferred by people rather than the traditional mode of shopping. Millions of user-generated comments are posted daily on the web, and analysis of these opinions could be more directive towards the customers and manufacturers. That makes the Sentiment analysis of online reviews one of the most sought-after research topic. This paper portrays our experimental work on domain-specific feature-based sentiment analysis of product review. In this paper, we worked with some fixed predefined core features of a product for presenting the customer’s acceptance of the principal attributes of a product so that the manufacturer can improve the quality of the basic features. We have proposed a feature-oriented sentiment prediction scheme. That analyses the generated expressions from the textual reviews of a product for predicting sentiment and assigns scores for our predefined features to present a net sentiment profile of a product of all parameters. With 92% accuracy, our sentiment detection scheme is proved to be an effective way to highlight the core attributes that seem to be really beneficial to the customer and manufacturer.

Android has the biggest market share among all Smartphone operating system. The number of devices running with the Android operating system has been on the rise. By the end of 2012, it will account for nearly half of the world's Smartphone market. Along with its growth, the importance of security has also risen. Security is one of the main concerns for Smartphone users today. As the power and features of Smartphone’s increase, so has their vulnerability for attacks by viruses etc. Perhaps android is more secured operating system than any other Smartphone operating system today. Android has very few restrictions for developer, increases the security risk for end users. Antivirus Android apps remain one of the most popular types of applications on Android. Some people either like having that extra security or just want to be extra cautious just in case and there is nothing wrong with that. In this list, we’ll check out the best antivirus Android apps and anti-malware apps on Android! The efficiencies of these applications have not been empirically established. This paper analyzes some of the security tools written for the Android platform to gauge their effectiveness at mitigating spyware and malware.

A method to improve the bandwidth for L and C band of backfire antenna is proposed. The performance of backfire antenna is investigated by performing numerical calculation by using various mathematical formulas to determine necessary dimension of the antenna and simulation by using IE3D software. Here we design proposed geometry for 3GHz. For proposed geometry we got 37.5% bandwidth for VSWR < 1.5 return loss -45 dB and directivity is 8 dBi .

The development of intelligent data analysis techniques for fraud detection can be well motivated from an economic point of view. Following the definition of fraud, it is easy to state the losses caused by fraud as primary motivation for fraud detection mechanism. Fraud in communication networks can be described and characterized as determined unobserved intentions to illegitimately use the communication networks in order to avoid service charges or gain unjust advantage. Efforts in this work are directed at fraud detection in post-paid organizational mobile communications networks. Two different complementary approaches are used: (differential and absolute, user profiling and classification approaches). It is observed that fraudulent intentions are reflected in the observed call data, which was subsequently used in describing behavioural patterns of users. Relevant user groups based on call data were identified and users are assigned to a relevant group to model the fraud detection mechanism. In the task the call data was used to learn models of calling behaviour so that these models make inferences about users’ intentions. From the analysis and model detectability experiment carried out in this scientific research work. It was discovered that the model detects over 89% of the fraudsters in the testing set (i.e fraud with certainty factor of 0.89). With the bias proportion of 0.0 and Mean Absolute Error (MAE) of (2.71) generated in the fraud detection. The model of course shows a good performance.

In our paper , we have suggested a variant of the RSA public cryptosystem, it is more effecient than the traditional RSA algorithm which uses the Euler’s Totient Function .The discussed RSA variant is very much similar to the traditional RSA. It also is a block cipher like the traditional one but it uses a different Ø(n) function. In the suggested scheme the range of Ø(n) is so high that it increases the security manifold. This makes cryptanalysis very difficult. Methods like brute force attack will fail as the range is very high and such attacks consume a lot of time. We represent Ø(n) as f(n,h) where n can be a Positive Integer(Z). As the value of h keeps increasing the range of values of Ø (n) also keeps increasing exponentially which in turn increases the security exponentially. This is the reason why the proposed scheme can be assumed to be more efficient, reliable and scalable.

In modern world, decision support systems are gaining increased popularity in various domains, including medical informatics and engineering. Digital access to the patient case data in would help the novice nurses in making faster and accurate decisions. The discussions in this paper are based on N-CODES (Nursing Computer Decision Support), an interactive handheld device which takes in patient data and delivers clinical knowledge i.e. it provides a diagnostic for decision support by the filtered and collated data. The focus of this paper is mainly on one aspect, the indexing of patient specific cases, and individual rule fragments. ‘Improve’ data library has been used, which contains instances of each possible patient class in the clinical decision support system (CDS). These classes are organized into frequent case (FC) trees by developing an algorithm which will find the frequently occurring rules. This algorithm will support the patient case indexing by organizing cases into rule vectors consisting of rule instances for a specific case ending with an intervention and also will aid in determining similarity of new cases with stored exemplar cases.