NCRACS 2016 - National Conference on Recent Advances in Computer Sciences

"NCRACS 2016 Conference Papers "

HISTORY BASED ROUTING FOR RESCUE OPERATIONS IN VEHICULAR SENSOR NETWORKS[ ]


In vehicular sensor networks (VSNs), an increase inthe density of the vehicles on road and route jamming in the network causes delay in receiving the emergency alerts, which results in overall system performance degradation. In order to address this issue in VSNs deployed in dense urban regions, in this paper, we propose collaborative learning automata-based routing algorithm for sending information to the intended destination with minimum delay and maximum throughput. The learning automata (LA) stationed at the nearest access points (APs) in the network learn from their past experience and make routing decisions quickly. The proposed strategy consists of dividing the whole region into different clusters, based on which an optimized path is selected using collaborative LA having input parameters as vehicle density, distance from the nearest service unit, and delay. A theoretical expression for density estimation is derived, which is used for the selection of the “best” path by LA. The performance of the proposed scheme is evaluated with respect to metrics such as packet delivery delay (network delay), packet delivery ratio with varying node (vehicle) speed, density of vehicle.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
A SURVEY ON MULTI-PACKET RECEPTION WITH RANDOM ACCESS IN WIRELESS NETWORKS[ ]


This survey provides a comprehensive review of existing carrier sensing enhancements for IEEE 802.11 wireless networks. Medium Access Control (MAC) Layer plays a significant role in Wireless Local Area Network (WLAN). The original physical carrier sensing mechanism, used by wireless stations to gain access to the medium is limited. It faces lot of challenges in terms of packet loss, fairness, collisions, reception and vulnerable to near-far problem. Multi-Packet Reception (MPR) technology addresses the above problem and provides the capability for a wireless receiver to parallely decode multiple packets from concurrent transmissions. New research advances are leading to increase in the reception capability of a single centrally receiving node called Access Point (AP) in WLAN. This article presents an in-depth survey of the existing literature in this area, detailing the various approaches and their efficacy in addressing the near-far problem and consequently increasing performance. It offers a comparison of the techniques, by evaluating the models, limitations, assumptions, and performance gains. The benefits of MPR technology in terms of throughput gain and reduction in packet collision and its various technologies that enable MPR at the datalink layer of a wireless network stack are highlighted. Finally, this article bestows an intuition to achieve sustained throughput and to eliminate near-far problem.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Diffusion speed based face liveliness detection to expose spoofing attack[ ]


Prevailing face biometric systems are open to spoofing attacks. A spoofing attack occurs when a person tries to masquerade as someone else by misrepresent data and thereby gaining illegitimate access. Spoofing with the help of photographs or videos is one of the most collective approaches of attacking face recognition and highly secured systems. In this a real-time method is proposed based on the diffusion speed the diffusion speed of a single image is obtained to address this problem. Particularly the difference in surface properties between a live face and a fake one (spoofed image) is capably revealed in the diffusion speed, the antispoofing features are exploited by employing the total variation flow scheme. More precisely, defining the local patterns of the diffusion speed is proposed and that is so-called as local speed patterns, as the features, which is the input into the linear SVM classifier to define whether the given face is fake or not. The significant advantage of this method is that, in contrast to previous approaches, it accurately identifies varied malevolent attacks regardless of the medium of the image, e.g., paper or screen. Moreover, this method does not require any specific user cooperation with the system.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
An Animal Health Monitoring System Using Zigbee Device[ ]


Animal detection plays a vital role in day to day life. It is important to detect the presence of animals entering into the human living areas near the forest, since it causes damage to life of people living near by the forest areas. It is important to safe guard the life of human by detecting the presence of animal and take necessary actions to safeguard human life. It is also equally important to save the animals. In order to overcome the above drawback a warning system must be developed . This paper involves the use of PIR(Passive Infra Red Sensor) which senses the presence of animal. The microcontroller is the heart of the system. It controls every component of the system. The LCD monitor displays the if the animal has been detected. Buzzer is used for alerting. With the help of GSM the warning message is sent to five persons.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Suggestion Synthesizer for Profit Maximization Using Preferences[ ]


In this gloabalized market,it is a challenging task for business people to sustain their customers.Every customer has more number of choices to pick a product or a service what suits for them.Business people spend lakhs and Lakhs of Money for getting feedback from their customers about a product or a service through surveys to improve their business.Since it is a highly competitive environment ,comparing a person’s product or service with every other product or a service vendor is a toughest task.Only when we compare we will be able to get a clear idea for offering new services or a product to the customer satisfying their needs.In addition,every business will be having a target customer set.For Example.A restaurant suited in a city’s center near by to the bus stop will definitely have soft drinks and chat items.There are several different businesses which will also be available like mobile recharge centers,petty shops with food items like chocolates and etc..But,how they compete for earning profit is the question in the gloabalized market.This paper,deals with the above problem and provides a model that provides solution by Comparing the preferences that user gives and finds out the features that are wanted but are not provided by the business men as a suggestion.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Data Mining with Big Bang Data[ ]


Big Data is used to identify the datasets that are due to their large size and complexity. Big Data concerns large-volume, complex, growing data sets with multiple, autonomous sources. Big Data are now rapidly expanding in all science and engineering domains, including physical, biological and biomedical sciences. Big Data mining is the capability of extracting useful information from these large datasets or streams of data, that due to its volume, variability, and velocity, it was not possible before to do it. The Big Data challenge is becoming one of the most exciting opportunities for the next years. This survey paper includes the information about what is big data, a HACE theorem that characterizes the features of the Big Data revolution from the Data mining perspective, Data mining with big data, features and Challenging issues.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Exploration of Association Rules by ap-plying Automated Cogency[ ]


A new EARAC ( Exploration of Association Rules by applying Automated Cogency) algorithm is presented to explore association rules. Cogency is estimated using pairwise item conditional probability. EARAC algorithm explores association rules by only one pass through the dataset.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Personalized Health Information Agent[ ]


In the future, artificial intelligence machines will replace or enhance human capabilities in many areas. One such a field is Heath care, where in Electronic Health Records (EHR) has been widely used with the help social media to share information about medicines which will cure diseases with others. Artificial intelligence is one rapidly growing technology which helps the researchers to find medical data which are available in the social media and suggest better medicines for the disease queried by the users. Information retrieval is performed by C4.5 algorithm. The system is designed in such a way that the search result will be more optimized with good quality and more efficient to assist the people searching for medical information for particular disease in the internet. The collected information are stored and categorized by using Support Vector Algorithm. Finally, Apriori algorithm is used to make the decision according to the categorized data about the medicines according to the user search.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
A Forced Caching Scheme to Improve Data Access in Disruption Tolerant Networks[ ]


Data access is a major issue in Disruption Tolerant Networks (DTN’s) because of its opportunistic contacts, irregular network connectivity, long delay and low node density. In this peculiar DTN environment determining the data location to query a request involves more time and high cost. To resolve this issue, the paper proposes a Cooperative Caching scheme based on opportunistic path forwarding in which the data is redundantly cached at multiple nodes. The underlying idea is to select a set of nodes called as Network Central Locations (NCL) to cache data thereby reducing data access delay. Additionally coordination among all caching nodes is also achieved to reduce transmission of redundant information in network. The simulation results obtained proves the efficiency of the system in terms of delivery rate and delay.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
An Efficient Clustering-Guided Fuzzy Roughset Genetic Learning for Unsupervised Feature Selection[ ]


Clustering is the application of data mining techniques to discover patterns from the datasets. Here a Fuzzy based kernel mappings clustering (FKMC) in high dimensional data is proposed which incorporates genetic roughset based feature selection concept- the process of deriving the similarity information from the unsupervised dataset. A frequent change in similarity information makes cluster aggregation a difficult task. Process of finding the optimal feature data points that are similar to a training data is challenging task which intricate linking of raw data points to one another and elimination of anomaly information .Finally, extensive experiments are governed on both synthetic and real world datasets.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Dynamic Cluster Head Selection to Minimize the Energy Consumption in Wireless Sensor Networks[ ]


The real world application is mainly depends on Wireless Sensor networks (WSNs).The Data Collection Scheme is the major task used in WSN. The Velocity Energy-efficient and Link-aware Cluster-Tree (VELCT) scheme for data collection in WSNs which failed to mitigate the problems of coverage distance, mobility, delay, traffic, tree intensity and end-to-end connection. The VELCT select the cluster head statically. The proposed dynamic clustering used to select the cluster head dynamically. At a certain amount of time, the clustering changes the cluster head dynamically. The Data Collection Tree (DCT) used to collect the information and tree formation is initiated. The dynamic cluster head is formed based on the tree formation formed in the cluster. The Proposed scheme minimizes the energy exploitation, reduces the end-to-end delay and traffic in cluster head in WSNs by effective usage of the DCT. The strength of this above proposed method used to reduce the energy consumption of the cluster head. The simulation results should increase the QoS parameters such as minimize the energy consumption, Packet delay, traffic in the network lifetime for wireless based network.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
GENERATING FREQUENT ITEMSET IN BIG DATA USINGFIN ALGORITHM[ ]


Emerging research area in Big data are handling issues like storing, searching, sorting, retrieving, securing, analyzing and visualizing are of immense importance. Customer behavior can be analyzed using Association rule mining which improves the organization thereby excelling business intelligence. To mine frequent itemset in big data we propose a novel FIN algorithm with map-reduce concept which helps in increase the performance by parallel processing. The parallel execution on identifying the frequent item set can performed by map-reducer function and using FIN algorithm. The performance of FIN is faster when compared with traditional algorithm such as Apriori, FP growth. This will also be applicable when it is perform in big data. The parallelization of the algorithm is increases the effectiveness. This proposed strategy can recommend the closely related products to the customers.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Design and Implementation of Context Aware Security by Hierarchical Multilevel Architectures using Secure Internet Services[ ]


Data gathering management in distributed Internet services is usually in light of username and password, explicit logouts and components of user session termination utilizing incredible timeouts. This paper gives a framework for how to leverage Lightweight Directory Access Protocol (LDAP) to implement Role-based Access Control (RBAC) on the Web in the server-pull architecture. LDAP-based directory services have recently received much attention because they can support object-oriented hierarchies of entries in which we can easily search and modify attributes over TCP/IP. To implement RBAC on the Web, we use an LDAP directory server as a role server that contains users' role information. The role information in the role server is referred to by Web servers for access control purposes through LDAP in a secure manner (over SSL). We provide a comparison of this work to our previous work, RBAC on the Web in the user-pull architecture.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
COLOR CO-OCCURRENCE AND BIT PATTERN FEATURE BASED CBIR[ ]


Content Based Image Retrieval is most recently used technique for image retrieval from large image database. Content based image retrieval technique can be found in a number of different domains such as Data Mining, Education, Medical Imaging, Crime Prevention, Weather forecasting, Remote Sensing and Management of Earth Resources The reason behind using CBIR is to get perfect and fast result. There are many technique of CBIR used for image retrieval. This paper presents a technique for CBIR by exploiting the advantage of lowcomplexity ordered-dither block truncation coding (ODBTC) for the generation of image content descriptor. The color co-occurrence feature (CCF) and bit pattern features (BPF) are used to index an image. The CCF use CMY Color Model. The CCF and BPF of an image are simply derived from the two ODBTC quantizers and bitmap, respectively, by involving the visual codebook. Experimental results shows that the proposed method efficiently retrieve images compared to other methods.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Hybridized Level Transformation Technique for Related Time Series Data[ ]


Sequence pattern matching technique is to find that related time series data but the data has high dimensionality, so it is critical. To improve the accuracy of the sequence pattern matching. DCT and DFT based MBR transformation is introduced to construct a low dimensional MBR which would convert the high dimensional MBR into low dimensional MBR. The DCT based approach is proved as better approach than the DFT based approach. DCT approach is based on the energy compression technique which might lead to a more computational complexity. This problem can be resolved in the proposed methodology by introducing the hybridized DCT-SVD approach where the SVD would select the most optimal energy efficient component for every block. It will take advantages of DCT first, and use SVD only for the blocks that DCT does not compact energy well. The DCT-SVD approach provides better result than the existing approach in terms of improved accuracy and performance evaluation.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Secure Binary Image Steganography Based on Huffman Coding[ ]


In insecure communication, data hiding techniques have an important role to protect secret information from unauthorized access. Steganography is a hiding technique that hides the secret information inside the digital medium in undetectable manner. In this paper, an secure binary steganography method based on huffman coding is proposed. First extract the complement, rotation, and mirroring-invariant local texture patterns (crmiLTPs) from the binary image first. The weighted sum of crmiLTP changes when flipping one pixel is then employed to measure the flipping distortion corresponding to that pixel.The steganographic scheme generates the cover vector by dividing the scrambled image into superpixels. The secret message is hided in cover image using Huffman coding. Experimental results have demostated that the proposed steganographic scheme can achieve statistical security without degrading the image quality or the embedding capacity.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
DETECTION AND REMOVAL OF BLACK HOLE ATTACK IN MOBILE AD-HOC NETWORKS USING COOPERATIVE BAIT DETECTION METHOD SCHEME[ ]


In Mobile Ad-Hoc Network (MANET) due to dynamic topology the nodes are free to move in and out of the network at any point of time. MANET is widely used in military based applications due to their infrastructure less property. Ad-hoc network are vulnerable to various types of attacks such as eavesdropping, denial of service, etc. Protecting the network from malicious attacks such as black hole attack, grey-hole attack which is very demanding in case of reactive routing protocols. This paper mainly focuses on designing Ad-hoc On Demand based routing (AODV) to protect the network from black hole/grey-hole attack by using Cooperative Bait Detection method Scheme. CBDS method implements a reverse tracing technique to track the malicious nodes in the Network.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Emotion Recognition Using Human Appearance Model[ ]


Facial Expression gives important instruction about emotion of a person. Facial expression recognition has become one of the major problem in computer vision. In this Project, a personalized pre-processing method is proposed to increase the accuracies of neutral and emotion classification. Supervised methods cannot be accommodate to all appearance lightness across the faces like nose, eye, mouth, in the limited amount of training data. To overcome the shortcomings in the traditional supervised Emotion Recognition(ER) methods in terms of accuracy and speed a light-weight method to classify neutral and emotion classification is proposed. Key Emotion (KE) points over the face will be extracted based on the sensitivity to identify the emotion from the image. This proposed method provides a secure and robust framework to classify the emotions through various images.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
A Multi-Level Adaptable Components Framework for Unified Networked Environment[ ]


For service providing systems, incoming service request rate at a server is a crucial parameter. Given that Internet traffic is bursty in nature and it is not usually well shaped by the time it arrives at a server, it is possible to encounter extremely heavy traffic load at a server particularly at a peak time. Coping with such traffic loads may imply dropping request packets at the server buffer outright, undoubtedly an egregious solution for both the service provider and its clientele, or alternatively, a “reduced” quality service may be proffered if service requesters are willing to accept such a compromise. Here, the system expected/target behavior is assumed to be predicated by a high level policy, and in order to serve that specific policy well, the system is expected to adapt its behavior to the monitored change in local network traffic intensity at the server site. If reduced quality of service is available to be dispensed to client set without necessarily dropping them out altogether, system functionality may be maintained, if not improved.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
Efficient Detection Technique and Algorithm for Health Care Improvement in Social Networks[ ]


Recently, the large scale use of expertise has a significant impact on the worldwide health care sector. In social network, the number of users is spending their time for sharing communication and distributes content simply along with other users. When social networking sites distribute the essential reason of online communication and message, detailed objectives and outline of usage differ extensively across diverse services. By using data mining techniques the most important knowledge is extracted from the social network which is attracted tremendously in the field of biomedical and health informatics community. The several methods are suggested such as link prediction, hierarchical clustering and subgraph detection techniques to handle the health care data in social networks. However, the preceding research methods are issue with long time computation, inaccurate results and inefficiency. To overcome the above mentioned problems, we enhance a framework to analyze the forum reviews more effectively. This framework is focused on the analysis sentiments of positive, negative, neutral, side effects of treatment and recognizes posts, likes and comments of user’s forum. The proposed work is identifying the user communities and influential users based on their opinions of cancer treatment. We introduced a self organizing map (SOP) for analyzing the word frequency from the user’s forum posts, likes and comments. This research work is used network based modelling method to model the user’s forum communications and improved the stability quality measure. It is used to discover the consumer opinions and recognize influential users within the retrieved modules by using information derived from both word-frequency data and network-based properties. The experimental results concluded that the proposed system is better than the existing system in terms of effective performances.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
CLOTHING PATTERN RECOGNITION BASED ON LOCAL AND GLOBAL FEATURES[ ]


Choosing clothes with appropriate color and pattern is very challenging for amaurotic people. Amaurotic people are those who have partial sight due to medical reasons or lost their sight in an injury. Although, there are some automatic system for identifying clothes and patterns, still it is challenging by reason of large intraclass pattern variation. To accord with such obstacles, a Computerized Clothing pattern and Color Recognition system is introduced. The proposed system consists of 3 components a camera, data analysis unit and speaker. The camera captures the user’s cloth, data analysis unit will identify the complex pattern and finally the results are described to amaurotic people verbally. This system is capable of recognizing 4 major patterns (plaid, striped, irregular and patternless) and 11 colors. The system is proficient of analyzing both local and global features of the pattern. Radon signature detects the principal orientation of the image and to distillate global features of clothing patterns, Statistical properties from wavelet subbands are used. Finally to extract local features, SIFT detectors are used. Both local and global features are integrated to recognize complex clothing patterns. Clothing color identification is done using HSI color space. This system is found to be active and simple for amaurotic people.

------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------