https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/4555 Fri, 17 Apr 2026 19:02:26 GMT 2026-04-17T19:02:26Z https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/2820 Title: Connectivity Model Based Routing for Opportunistic Networks Authors: Thabotharan, K. Abstract: Opportunistic networks are uniquely characterized by the intermittent connectiv- ity experienced by the constituent nodes in a highly mobile wireless environment. The protocol design for such opportunistic networks raises new challenges to the conventional ad-hoc networking community, as some of the assumptions held for the latter, such as the existence of a multi-hop end-to-end path between a con- tent originator and a recipient are not generally valid for opportunistic networks. Also, the limited buffer space of a node prevents it from comfortably participating in the opportunistic content exchanges with its peer nodes. As such, any forth- coming new routing protocol for opportunistic networks must be designed and evaluated with these limitations in mind. In this work we have laid the founda- tion towards the design of such protocols by the careful analysis of opportunistic connectivity traces, characterization of network connectivity properties, and the identification of underlying probabilistic distributions in order to fully capture the dynamics of this new type of networking paradigm. In this thesis, the following original contributions are made to the field of opportunistic networking research: We have developed and simulated a novel connectivity model to regenerate the connectivity traces by reproducing the prop- erties gleaned from field connectivity traces for the testing and validation of new protocols and architectures. We have modeled the underlying dynamic structure of a network system as having two higher level properties of predictability and connectedness of nodes in predicting their contacts with their peer nodes with certain level of confidence, which in turn, can determine the dynamic connec- tivity of the network. We have validated that the proposed connectivity model can generate synthetic connectivity traces with such properties. We have pro- posed an adaptive opportunistic network routing protocol and using simulation based tests show that it outperforms three well known routing protocols in mobile peer-to-peer ad-hoc networking, in metrics such as message delivery ratio and the quantum of energy expended for message delivery. We have further enhanced our adaptive routing protocol to minimize congestion by considering node popularity, and show, with empirical simulation based tests, that our proposed congestion aware adaptive routing protocol outperforms all the three routing protocols and the proposed adaptive routing protocol in terms of number of messages delivered per unit of consumed power, for larger buffer sizes. Thu, 14 Sep 0030 00:00:00 GMT https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/2820 0030-09-14T00:00:00Z https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/2819 Title: An AI Approach to Investigate Sociological Impact on Education in Puttalam District Sri Lanaka Authors: Jayathilaka, Fr.P.R. (Rev) Abstract: This research is on an application of neural network and fuzzy models to evaluate the sociological factors, which affect the educational performance of the students in Sri Lanka. One of its major goals is to prepare the grounds to device a counseling tool, which helps these students for a better performance at their examinations, especially at their G.C.E O/L examination. Closely related sociological factors are collected as raw data and the noise of these data are filtered through the fuzzy interface and the supervised neural network is being utilized to recognize performance patterns against the chosen social factors. The geographical area of the research is Puttalam District. It is a district in the North West Province of Sri Lanka and it is an area of people of different religions, ethnicities, and etc. It also covers two Educational Zones namely Puttalam and Chilaw. The System is trained with the data obtained from a questionnaire given to the sample students before their G.C.E. Ordinary Level Examination and Examination results of the particular group of students. These data are fed into the system and the ANN is trained. Well trained ANN would be able to utilize as a counseling tool to help the children. With the questionnaire tool the selected social environmental data of a student could be obtained. With the well trained ANN, it is possible to predict the G.C.E. O/L results and identify the socially weak area which needs more attention. If possible the social environment could be adjusted to encourage student to perform better at the Examination. Thu, 14 Sep 0030 00:00:00 GMT https://dl.ucsc.cmb.ac.lk/jspui/handle/123456789/2819 0030-09-14T00:00:00Z