2016
Browse
19 results
Search Results
Item Traffic Prioritization Using Packet Aggregation in Saturated WSN(IUT, MCE, 2016-11-20) Hassan, Mehedi; Ibne Alam, ZubairIn the sensor networks, network tra c prioritization is gaining attention in the WSN community, as more and more features are being integrated these networks. Real-world deployment experience suggests that WSN brings new challenges to existing problems, such as resource constraints, low data-rate radios, and diverse application scenarios. But most of tra c prioritization problems deals with how to handle the HP (high-priority) data packet ignoring the LP (Low-Priority) data packet. Whenever any HP data packet arrives, LP data packet transmissions are suspended. This results in a loss of LP packets due to network congestion in a saturated network, weak radio signal, multi-path fading or cache over ow. We propose a framework that that will not suspend the LP data packet completely when any HP data packet has arrived. The framework will cache and aggregate the LP data packets and transmits the LP data packet after certain interval while transmitting the HP data packets. We di erentiate between the HP and LP data by leveraging transmission power di erence and radio-capture e ect. We classify the LP data tra c by using a hierarchical aggregation algorithm for data reduction of similar data. 5Item Teacher Student Communication System "LetStalk"(IUT, CSE, 2016-11-20) Nayeer, Mashroor; Shuvo, Md. Aminul HoqueThe project is about communication between teachers and students and to make it smooth some facilities are provided like a social networking site. Actually it’s more likely a social networking software where students can post anything and other students and teachers can like and comment on it or even can download it. Chatting facility is also provided to maintain personal communication between teacher and student. Everyonecan update their personal information on their profile and upload pictures too. The motivation behind making this software was to increase the bond among teachers and studentsItem Strain Estimation and Detection of Cancerous Breast Lesion through ultrasound image(IUT, CSE, 2016-11-20) Faiyaz, Abrar; Anan, Md. Samiul BasharCurrently in the domain of breast cancer detection, biopsy and ultrasound imaging technique is being used. Between these two techniques biopsy is costly, time consuming and tissue damaging for patients. Ultrasound imaging technique is an easy alternative which needs experts to accurately detect breast lesion and it's type. In the BIRAD method of identi cation, Sonologists look for certain features of the malignant lesions. These malignant lesions in static image has some signi cant features which with 15-30% cases result in contradictory results. Most important feature that a breast lesion posses is it's tissue property ie. tissue destiny and the deforming mechanism. These properties are associated with the strain value when pressure is applied to the tissue. There are di erent ways the strain value can be calculated by using multiple images with di erent strain values. Generally,the malignant lesions exhibits more rigidity than the benign ones. From this motivation, optical ow algorithm has been implemented to calculate the strain value and compare it's measure of mean related to accuracy with cross correlation method used to calculate the strain values. From this motivation, we have gured out strain with higher accuracy adopting our \M- Filtered B-Mode Image Tissue Elasticity" estimation method. Higher the accuracy, higher will be the mean of the calculated strain. Satisfactory SNR (dB) results in our implemented method of optical ow which is promising for a classi cation with higher accuracy.Item Personal Expense Assistant Management: An Android Based Application(IUT, CSE, 2016-11-20) Sali, Moussa; Abbo, Abdel SalamPersonal Expense Assistant Management is an application aiming to manage our daily expenses in a more efficient and manageable way. The application attempts to free the user with as much as possible the burden of manual calculation and to keep the track of his expenditure. Instead of keeping a dairy or a log of the expenses on the smartphones or laptops, it enables the user to not just keep the tab on the expenses but also to plan ahead keeping the past budget in mind. With the help of this application, a user may be able to add, delete or change the current entered bill entry efficiently. The graphical representation of the budget is the lucrative part of the system as it appeals the user more and is easy to understand and incorporate for future planning. The user interface of the system ticks the boxes of consistency, easy readable dialogue boxes, easy exit and easy to get used to requirements for any ideal user interface.Item Online Shopping System with Prediction Analysis(IUT, CSE, 2016-11-20) Hasan, Md. Nasid; Aziz, JysanOnline shopping is the process whereby consumers directly buy goods, services etc. from a seller interactively in real-time without an intermediary service over the internet. Our system has a special feature that is predictive analysis of each product based on past data analysis. There is two di erent views in this system one is for user and another is for admin. In user view user can see all the products and choose among them. Here we try to implement a clothing system. The system has some categories like Man.Woman,Kids etc. Each category has also some subcategories. In admin view admin can update or delete any data. Admin can see the predictive analysis of each product. The unique feature of our system is the future prediction of each product. Here the prediction works based on the past and present data analysis of each product. Generally in a big system Big Data,Cloud computing etc. used for analyzing the huge number of data. But we have made a small protocol for less amount of data.In the KNN learning algorithm,the decision classi er will detect by itself according to the event and data. But in our small protocol we will build a system with static classi er,we will take the decision according to the static range.Item Lightweight AES Algorithm for Resource Constraint Devices(IUT, CSE, 2016-11-20) Chowdhury, Arnab Rahman; Mahmud, JunayedOur world is slowly approaching brink of mankind's next technological revolution since the conception of Internet of Things(IoT) emerges. The realization of IoT requires an enormous amount of sensor nodes which have limited battery power, memory, compu tational latency and communication bandwidth to acquire inputs from the connected objects. In the current developments of the resource constraint environments, the trend is shifted towards lightweight cryptographic algorithm. Many lightweight cryptograpic algorithms have been developed and also existing algorithms are modied in terms of resource constraint environment. In this paper we analyzed a very popular block cipher AES-128 and tried to make it lightweight regarding energy consumption. We also studied some other lightweight block cipher which are optimized in terms of energy, power and memory.Item KAJ Secondary School(IUT, CSE, 2016-11-20) Surajo, Jabir; Jaiteh, KarambaWe are in the age of big changes, particularly the technology revolution and this has affected all aspect of our life. This modern-day revolution, at the global level, has manifested itself in the form of many innovations and breakthroughs and giant leaps in internetworking technology. With these changes, we have new opportunities, people can now transcend the barriers of time and distance with the internet’s speed. This has completely changed our living style, and every sector got changed and most of the traditional activities are going on now online like jobs, study. The aim of the project is to develop a document management system that is able to deliver access to anyone authorized anytime, anyplace on any device. The system developed makes use of rich internet technology to replace desktop application with rich internet application. The system shares the advantage of both web application and desktop application, and removes the most disadvantages of both. The main purpose of this project is designed for better interaction between students, teachers, parents & management. It has also student`s Forum where students can post and discuss any topic. Thus this project has four major parts, the admin part for managing the all system, the teachers’ part using to perform various teaching related activities, the students part to get and view their attendance as well as to communicate each other, the parents part to view their children’s details, performance and to communicate with the school.Item Gene Co-expression Network analysis of Lung Adenocarcinoma Cell Carcinoma data(IUT, CSE, 2016-11-20) Uddin, Md. Saif; Ahamed, Md. TanvirA gene co-expression analysis on Lung adenocarcinoma data was done to find modules of genes that might highly impact the growth of this type of tumor. Along with that, cancer survival data was used to relate modules to prognostic significance for survival time. Analysis on microarray data revealed modules that were significant in gene enrichment analysis and 4 genes - TTk, C6orf173, CENPE, DCC1 were found that were significant in terms of survival time. A second analysis was done on a second set of RNAseq data and the significant genes in modules was found there, were also found in the RNAseq data implying that these genes might indeed play a crucial role in Lung adenocarcinoma.Item Friend Recommendation System in Social Network using Personality Analysis and User Behavior(IUT, CSE, 2016-11-20) Neehal, Nafis; Noor, Shoaib BinSocial networking is a tool used by people all around the world. Its purpose is to promote and aid communication. Social networks, such as Facebook, were created for the sole purpose of helping individuals communicate. These networks are becoming the modern way to make friends. These new friends communicate through these networks. There exist recommendation systems in all the social networks which help users to nd new friends and connect to more peoples. With friends, there comes a strong friend recommendation system also. The existing social networks do have their own friend recommendation system which is based on the friends of friends' methodology. This graph based friend recommendation system is not very accurate most of the time and drive users to wrong direction. We tried to make this recommendation system more accurate adding some extra layers of personality analysis and user behavior. With the vast amount of user data, our system will gure out each user's personality traits and behavior which will be used to help him/her nding out new users with same nature.Item Enterprise Social Network (ESN)(IUT, CSE, 2016-11-20) Hossain, Md. Kamal; Parvez, Md. SabitThe enterprise social networking website is an online community designed to make your working life more active and stimulating. The social network can help you maintain collaboration with your colleagues and share les and messages, and establish working relationship by sharing their views and knowledge. Enterprise Social Network is a Social Networking Website, which a web based service providing its users to construct a public pro le within a bounded system. This website provides the features of keeping intact with one and all at one place. Let's keep in touch with your peers and pals, without boring you at the same time. Enterprise Social Network has various other attractive features like chatting, post, comment, upload les.