Thiruthlall Nepal - Academia.edu (original) (raw)

Uploads

Papers by Thiruthlall Nepal

Africa research journal, Dec 1, 2015

In this paper, we investigate the performance of an alternative received signal filtering techniq... more In this paper, we investigate the performance of an alternative received signal filtering technique based on local averaging to improve the quality of handover decisions in Long Term Evolution (LTE) networks. The focus of LTE-Advance (LTE-A) networks is to provide enhanced capacity and reliability of radio access as well as broadband demand for mobile users. The necessity to maintain quality of service, especially for the delay sensitive data services and applications, has made mobility and handover decisions between the base stations in the LTE networks critical. Unfortunately, several handover decision algorithms in the LTE networks are based on the Reference Signal Received Power (RSRP) obtained as a linear averaging over the reference signals. The critical challenge with the linear averaging technique is that the limited reference signal available in the downlink packet introduces an estimation error. This estimation error is a result of the effects of linear averaging on propagation loss components in eliminating fast-fading from the received signals. Moreover, prompt and precise handover decisions cannot be based on inaccurate measurement. The standardized LTE layer 3 filtering technique is applied to the local averaged layer 1 signal to render it suitable for LTE handover decisions. The local averaging technique produces better handover than the linear averaging technique in terms of the reduced number of handover failures, improved high spectral efficiency and increased throughput, especially for cell-edge users with high speeds. The findings of this study suggest that the local averaging technique enhances mobility performance of LTE-Advance networks.

With the introduction of the degree programs in technikons the call for research outputs sounds i... more With the introduction of the degree programs in technikons the call for research outputs sounds increasingly urgent. Technikons have now realised that they are well positioned to also undertake applied and experimental research. The need for a sustainable research programme in the area of Al and telecommunications was identified at M L Sultan Technikon about three years ago. Positive spin-offs of such a programme include improved co-operation in research among the staff and students in the departments of Electronic Engineering and Computer Studies, partnerships with academic institutions such as University of Durban Westville, University of Natal and other research organisations. This provides also an opportunity for staff to continually update their skills and knowledge base in terms of rapidly changing technology, and an improved ability for the Technikon to continue to offer relevant and appropriate qualifications. The paper will present the Technikon's focus areas of research in Al and its application to Telecommunications. It is funded by the Foundation of Research and Development, Telkom and Ericson. Its outputs are envisaged to be in the form of practical and theoretical results, useful to both the of IT field and to the telecommunications industry. The results of the research will have implications for the local, regional and international communities. Another effect is improving the qualifications of least one third of the staff members within the departments of Computer Studies and Electronic Engineering, Development of telecommunications infrastructure and in particular rural telecommunications is a complex process that involves both technical and socio-economic factors. It requires the building the technical base for rural telecommunications, assessment procedures for the evaluation of opportunities and strategic planning, improvement of maintenance and repair activities. A crucial element in all these activities is the need for improvement of decision support. The latter is provided at different levels depending on the nature of the tasks. Thus a service engineer needs a decision support system for better identification of technical faults that affects mainly the work of a single person at a time. On the other hand decisions that need to be taken on the development of telecommunications infrastructure in larger rural areas are not only technical decisions but also socio-economic decisions involving groups of many stakeholders. This paper examines various· research topics undertaken at ML Sultan Technikon related to the following aspects of decision making or technological issues associated with the development of rural telecommunications: • Investigation of Problem Structuring in the Analytic Hierarchy Process and Soft Systems Methodology for the Evaluation of the Socio Economic Impact of Telecommunications Services in Rural Areas. • Investigation of the Role of Different Measurement Scales in the Analytic Hierarchy Process when applied to the evaluation of research projects in the area of telecommunications. • Handling of Uncertainty in Judgements in the Analytic Hierarchy Process when applied to the selection of equipment for rural telecommunications development. • Investigation into Case Based Reasoning for fault identification and repairs in telecommunications. • Expert systems approach to equipment maintenance for the telecommunications industry. • The application of Artificial Intelligence in Digital Signal Processing. This project focuses on speaker identification using Artificial Neural Networks. The above topics are initiated at the Centre of Excellence in Rural Telecommunications uniting the human resources of the departments of Computer Studies and Electronic Engineering. The projects are varied but at the same time sufficiently coherent. This will ensure that the Technikon will build a critical mass in the area of Artificial Intelligence and Telecommunications and will be able to offer the local industry wider expertise and assistance

SAIEE Africa Research Journal, 2015

2015 3rd International Conference on Artificial Intelligence, Modelling and Simulation (AIMS), 2015

Support Vector Machine (SVM) is widely recognized as a potent data mining technique for solving s... more Support Vector Machine (SVM) is widely recognized as a potent data mining technique for solving supervised learning problems. The technique has practical applications in many domains such as e-commerce product classification. However, data sets of large sizes in this application domain often present a negative repercussion for SVM coverage because its training complexity is highly dependent on input size. Moreover, a single kernel may not adequately produce an optimal division between product classes, thereby inhibiting its performance. The literature recommends using multiple kernels to achieve flexibility in the applications of SVM. In addition, color features of product images have been found to improve classification performance of a learning technique, but choosing the right color model is particularly challenging because different color models have varying properties. In this paper, we propose color image classification framework that integrates linear and radial basis function (LaRBF) kernels for SVM. Experiments were performed in five different color models to validate the performance of SVM based LaRBF in classifying 100 classes of e-commerce product images obtained from the PI 100 Microsoft corpus. Classification accuracy of 83.5% was realized with the LaRBF in RGB color model, which is an improvement over an existing method.

submitted in partial fulfillment of the requirements for the Degree of Doctor of Technology: Info... more submitted in partial fulfillment of the requirements for the Degree of Doctor of Technology: Information Technology, 2002.

The evaluation of telecommunications infrastructure in rural areas is a complex and messy problem... more The evaluation of telecommunications infrastructure in rural areas is a complex and messy problem that involves many tangible and intangible factors. Some of them are technical in nature while others are soft, involving social, cultural and political aspects of the problem. The evaluation requires, inter alia, societal intervention, and since societies reflect a multiplicity and diversity of values and goals, the intervention should confront these realities. One such reality that can be highlighted is the fact that we are dealing with rural communities that are disadvantaged in terms of telecommunication and other service provision. This underlies the need to consider the evaluation process from multiple perspectives, explore methodologies that will facilitate participation and engagement, and include emancipation of the disadvantaged. This research proposes a framework for the evaluation of rural telecommunications in a regional context. As such, it is viewed not just as an enginee...

Thesis (B.Com.(Hons.)-Accounting)-University of Natal, 1988.

Rural Telecommunications planning in a developing country is a complex process that has to consid... more Rural Telecommunications planning in a developing country is a complex process that has to consider the various interrelationships in a rural sociotechnical system. A primary function of telecommunications planning is the selection or design of the most appropriate technologies for the particular rural application. However, the issues involved in the selection and deployment of rural communication technologies are not only technological, but are often complex and 'messy', cutting across various aspects of the rural society. Uncertainty and multiple conflicting objectives usually characterise the decision-making process in such an environment. This paper is concerned with the use of the Analytical Hierarchy Process as a suitable method of dealing with the qualitative and quantitative data in relation to conflicting objectives, in arriving at a consensus decision in a typical rural telecommunications planning situation, for the selection of the most appropriate access technology.

… and Theory of Automated Timetabling II, 1998

This paper describes a study of existing tertiary timetabling techniques in Southern Africa, base... more This paper describes a study of existing tertiary timetabling techniques in Southern Africa, based on initial questionnaires and follow-up interviews. Heuristics obtained from this study were used as the basis for a prototype automated timetabling package. The prototype's ...

Journal of Organisational Transformation & Social Change, 2008

ABSTRACT The article explores some issues related to the process of a systemic intervention using... more ABSTRACT The article explores some issues related to the process of a systemic intervention using our experience in three projects in the Information and Communications Technologies (ICT) sector carried out in South Africa between 1997 and 2002. The projects served as a testing ground for exploring some theoretical and practical aspects associated with the process of facilitating the intervention. The paper presents reflections on some of the lessons learned regarding the process in each of the three cases associated with complex developmental decision-making issues.

The paper reviews paradigm incommensurability issues and possible ways of overcoming it. One of t... more The paper reviews paradigm incommensurability issues and possible ways of overcoming it. One of the open problems of pluralist systemic intervention theory, on which there is no uniform understanding in the literature, is whether work across paradigms represents some kind of a meta paradigm or whether it belongs to a completely new paradigm itself or should it be grounded within the theory of existing paradigms. Another issue that requires additional investigation is the use of methods within different paradigms from those for which they were created originally. Our understanding of those issues was informed by the developments in Critical Systems thinking in the past decade. It was further developed through by our involvement in three interventions in the fields of rural telecommunications and software development. They involved mixing of techniques from several soft systems approaches and multi-criteria decision analysis. We justified mixing of methods using the theory of the thre...

Decision Support Systems, 2007

This paper shares experiences of the authors with combining Multiple Criteria Decision Making (MC... more This paper shares experiences of the authors with combining Multiple Criteria Decision Making (MCDM) and techniques from soft systems approaches for decision support at particular stages of complex problem solving. These are based on their involvement in three projects within the Information and Communications Technology sector. The justification for the combined use of MCDM and separate techniques from systems thinking is discussed. The contribution of this paper is in the demonstration of how combinations of methods that are parts of soft systems methodologies and MCDM may support multiple perspective representations of complex managerial problems, and in the lessons learned from the three cases. Published by Elsevier B.V.

Support Vector Machine (SVM) is widely recognized as a potent data mining technique for solving s... more Support Vector Machine (SVM) is widely
recognized as a potent data mining technique for solving
supervised learning problems. The technique has practical
applications in many domains such as e-commerce product
classification. However, data sets of large sizes in this
application domain often present a negative repercussion for
SVM coverage because its training complexity is highly
dependent on input size. Moreover, a single kernel may not
adequately produce an optimal division between product
classes, thereby inhibiting its performance. The literature
recommends using multiple kernels to achieve flexibility in the
applications of SVM. In addition, color features of product
images have been found to improve classification performance
of a learning technique, but choosing the right color model is
particularly challenging because different color models have
varying properties. In this paper, we propose color image
classification framework that integrates linear and radial basis
function (LaRBF) kernels for SVM. Experiments were
performed in five different color models to validate the
performance of SVM based LaRBF in classifying 100 classes of
e-commerce product images obtained from the PI 100
Microsoft corpus. Classification accuracy of 83.5% was
realized with the LaRBF in RGB color model, which is an
improvement over an existing method.

Africa research journal, Dec 1, 2015

SAIEE Africa Research Journal, 2015

2015 3rd International Conference on Artificial Intelligence, Modelling and Simulation (AIMS), 2015

Thesis (B.Com.(Hons.)-Accounting)-University of Natal, 1988.

… and Theory of Automated Timetabling II, 1998

Journal of Organisational Transformation & Social Change, 2008

Decision Support Systems, 2007

Support Vector Machine (SVM) is widely recognized as a potent data mining technique for solving s... more Support Vector Machine (SVM) is widely
recognized as a potent data mining technique for solving
supervised learning problems. The technique has practical
applications in many domains such as e-commerce product
classification. However, data sets of large sizes in this
application domain often present a negative repercussion for
SVM coverage because its training complexity is highly
dependent on input size. Moreover, a single kernel may not
adequately produce an optimal division between product
classes, thereby inhibiting its performance. The literature
recommends using multiple kernels to achieve flexibility in the
applications of SVM. In addition, color features of product
images have been found to improve classification performance
of a learning technique, but choosing the right color model is
particularly challenging because different color models have
varying properties. In this paper, we propose color image
classification framework that integrates linear and radial basis
function (LaRBF) kernels for SVM. Experiments were
performed in five different color models to validate the
performance of SVM based LaRBF in classifying 100 classes of
e-commerce product images obtained from the PI 100
Microsoft corpus. Classification accuracy of 83.5% was
realized with the LaRBF in RGB color model, which is an
improvement over an existing method.