Walter Kihuya - Academia.edu (original) (raw)

Papers by Walter Kihuya

International Journal of Computer Applications Technology and Research, 2019

The estimation of the QoE provides valuable input in order to measure the user satisfaction of a ... more The estimation of the QoE provides valuable input in order to measure the user satisfaction of a particular service/application. Network QoE estimation is challenging as it tries to measure a subjective metric where the user experience depends on a number of factors that cannot easily be measured. All the Network analysis models can be divided into two major groups: qualitative and quantitative. In recent years many quantitative models have been developed in terms of quantitative measures i.e. use of scale of numbers between 1 to 5 to represent user perception of QoS. The challenge with this model is where user perception is subjective and not precise thus cannot be clearly measured using quantitative methods. On the other side qualitative models are in early stages of exploration. Little has been done on qualitative methods. Basing on previous studies, few models exists that measure qualitative analysis of computer network quality of experience. However none incorporated all the four parameters of integrity of service; throughput, delay, packet loss and jitter as parameters of network QoE. In this work, a literature survey is done on qualitative performance by use of a variety of variables, input and output linguistic terms. After a broad survey of the literature, we tend to propose a fuzzy logic model for analysis of computer network QoE. Likewise, the model combines all the four parameters of network integrity of service parameters since they are the primary factor for QoS quantification of any network [1]. Moreover, by using the fuzzy logic concept, the output linguistic terms shows the user perception about a product or a service (QoE) to certain levels by use of membership functions, in this case triangular membership function which shows the mapping of each linguistic term to certain range of values rather being precise to a particular value. By such means, the developed fuzzy logic model tends to accommodate some degree of uncertainty and vague network values to be used for analysis purposes. The concern is to advance the analysis and evaluation of quality of experience in computer networks by use of fuzzy logic concept. The target population for this model is the ISPs' clients. This will enable ISPs to have the best responsive measures to deal with clients' QOE parameters so as to meet the QOS as per SLAs.

The estimation of the QoE provides valuable input in order to measure the user satisfaction of a ... more The estimation of the QoE provides valuable input in order to measure the user satisfaction of a particular service/application. Network QoE estimation is challenging as it tries to measure a subjective metric where the user experience depends on a number of factors that cannot easily be measured. All the Network analysis models can be divided into two major groups: qualitative and quantitative. In recent years many quantitative models have been developed in terms of quantitative measures i.e. use of scale of numbers between 1 to 5 to represent user perception of QoS. The challenge with this model is where user perception is subjective and not precise thus cannot be clearly measured using quantitative methods. On the other side qualitative models are in early stages of exploration. Little has been done on qualitative methods. Basing on previous studies, few models exists that measure qualitative analysis of computer network quality of experience. However none incorporated all the four parameters of integrity of service; throughput, delay, packet loss and jitter as parameters of network QoE. The study's objective is to address this gap by proposing a fuzzy logic model for analysis of computer network QoE. The tools used in the study are Linux MTR tool for data extraction, Ms. Excel for data cleaning and presentation, Visual paradigm for constructing of Unified Modeling language diagrams, mat lab software for plotting of functions/data, implementation of algorithms and creation of user interfaces. Experimental research design and sampling mechanisms is applied for 15 samples. The methodology in use is fuzzy logic. In order to deal with fuzziness associated with linguistic variables, inference rules are introduced. Five input linguistic terms are identified: Very High, High, Medium, Low and Very Low. Five output linguistic terms are defined to describe the opinion scores: Excellent, Good, Fair, Poor and Bad. Four variables are used: delay, jitter, packet loss and throughput. This results to a total of 625 rules (5^4). The rules are further condensed to 240 logical rules basing on expert knowledge. The collected data was used for simulation in matlab environment basing on the 240 rules. The results shows, analysis of Computer network QoE is subjective in nature rather than objective thus requires a resilient mechanism like fuzzy logic in order to capture clear-cut results to be used for decision making. The target population for this model is the ISPs' clients. This will enable ISPs to have the best responsive measures to deal with clients' QOE parameters so as to meet the QOS as per SLAs.

Two important Quality of Experience (QoE) parameters in computer networks service-related perform... more Two important Quality of Experience (QoE) parameters in computer networks service-related performance metrics are throughput and end-to-end delay. A slow delay indicates high network efficiency thus high throughput. Similarly, the entire four parameters linked to the integrity of service (Delay, Jitter, Packet Loss, and Throughput) are deliberated to be primary factors affecting almost all computer networks. The study's objective is to identify the vital primary parameters with preeminent results to be kept into consideration when analyzing Network QoE by using a fuzzy logic methodology. The tools used in the study were Linux MTR (my trace-route) tool for data extraction, Ms. Excel for data cleaning and presentation, MATLAB(matrix laboratory) software for plotting of functions/data, and creation of interfaces. A Fuzzy logic methodology was implemented using an experimental research design. The analysis of results using independent sample T-Test technique compared four primary pa...

IJCSIS Vol 17 No 4 April Issue, 2019

International Journal of Computer Applications Technology and Research

The estimation of the QoE provides valuable input in order to measure the user satisfaction of a ... more The estimation of the QoE provides valuable input in order to measure the user satisfaction of a particular service/application. Network QoE estimation is challenging as it tries to measure a subjective metric where the user experience depends on a number of factors that cannot easily be measured. All the Network analysis models can be divided into two major groups: qualitative and quantitative. In recent years many quantitative models have been developed in terms of quantitative measures i.e. use of scale of numbers between 1 to 5 to represent user perception of QoS. The challenge with this model is where user perception is subjective and not precise thus cannot be clearly measured using quantitative methods. On the other side qualitative models are in early stages of exploration. Little has been done on qualitative methods. Basing on previous studies, few models exists that measure qualitative analysis of computer network quality of experience. However none incorporated all the four parameters of integrity of service; throughput, delay, packet loss and jitter as parameters of network QoE. The study's objective is to address this gap by proposing a fuzzy logic model for analysis of computer network QoE. The tools used in the study are Linux MTR tool for data extraction, Ms. Excel for data cleaning and presentation, Visual paradigm for constructing of Unified Modeling language diagrams, mat lab software for plotting of functions/data, implementation of algorithms and creation of user interfaces. Experimental research design and sampling mechanisms is applied for 15 samples. The methodology in use is fuzzy logic. In order to deal with fuzziness associated with linguistic variables, inference rules are introduced. Five input linguistic terms are identified: Very High, High, Medium, Low and Very Low. Five output linguistic terms are defined to describe the opinion scores: Excellent, Good, Fair, Poor and Bad. Four variables are used: delay, jitter, packet loss and throughput. This results to a total of 625 rules (5^4). The rules are further condensed to 240 logical rules basing on expert knowledge. The collected data was used for simulation in matlab environment basing on the 240 rules. The results shows, analysis of Computer network QoE is subjective in nature rather than objective thus requires a resilient mechanism like fuzzy logic in order to capture clear-cut results to be used for decision making. The target population for this model is the ISPs' clients. This will enable ISPs to have the best responsive measures to deal with clients' QOE parameters so as to meet the QOS as per SLAs.

IJCSIS Vol 17 No 4 April Issue, 2019

International Journal of Computer Applications Technology and Research, 2019

The estimation of the QoE provides valuable input in order to measure the user satisfaction of a ... more The estimation of the QoE provides valuable input in order to measure the user satisfaction of a particular service/application. Network QoE estimation is challenging as it tries to measure a subjective metric where the user experience depends on a number of factors that cannot easily be measured. All the Network analysis models can be divided into two major groups: qualitative and quantitative. In recent years many quantitative models have been developed in terms of quantitative measures i.e. use of scale of numbers between 1 to 5 to represent user perception of QoS. The challenge with this model is where user perception is subjective and not precise thus cannot be clearly measured using quantitative methods. On the other side qualitative models are in early stages of exploration. Little has been done on qualitative methods. Basing on previous studies, few models exists that measure qualitative analysis of computer network quality of experience. However none incorporated all the four parameters of integrity of service; throughput, delay, packet loss and jitter as parameters of network QoE. The study's objective is to address this gap by proposing a fuzzy logic model for analysis of computer network QoE. The tools used in the study are Linux MTR tool for data extraction, Ms. Excel for data cleaning and presentation, Visual paradigm for constructing of Unified Modeling language diagrams, mat lab software for plotting of functions/data, implementation of algorithms and creation of user interfaces. Experimental research design and sampling mechanisms is applied for 15 samples. The methodology in use is fuzzy logic. In order to deal with fuzziness associated with linguistic variables, inference rules are introduced. Five input linguistic terms are identified: Very High, High, Medium, Low and Very Low. Five output linguistic terms are defined to describe the opinion scores: Excellent, Good, Fair, Poor and Bad. Four variables are used: delay, jitter, packet loss and throughput. This results to a total of 625 rules (5^4). The rules are further condensed to 240 logical rules basing on expert knowledge. The collected data was used for simulation in matlab environment basing on the 240 rules. The results shows, analysis of Computer network QoE is subjective in nature rather than objective thus requires a resilient mechanism like fuzzy logic in order to capture clear-cut results to be used for decision making. The target population for this model is the ISPs' clients. This will enable ISPs to have the best responsive measures to deal with clients' QOE parameters so as to meet the QOS as per SLAs.

IJCSIS Vol 17 No 4 April Issue, 2019

International Journal of Computer Applications Technology and Research

The estimation of the QoE provides valuable input in order to measure the user satisfaction of a ... more The estimation of the QoE provides valuable input in order to measure the user satisfaction of a particular service/application. Network QoE estimation is challenging as it tries to measure a subjective metric where the user experience depends on a number of factors that cannot easily be measured. All the Network analysis models can be divided into two major groups: qualitative and quantitative. In recent years many quantitative models have been developed in terms of quantitative measures i.e. use of scale of numbers between 1 to 5 to represent user perception of QoS. The challenge with this model is where user perception is subjective and not precise thus cannot be clearly measured using quantitative methods. On the other side qualitative models are in early stages of exploration. Little has been done on qualitative methods. Basing on previous studies, few models exists that measure qualitative analysis of computer network quality of experience. However none incorporated all the four parameters of integrity of service; throughput, delay, packet loss and jitter as parameters of network QoE. The study's objective is to address this gap by proposing a fuzzy logic model for analysis of computer network QoE. The tools used in the study are Linux MTR tool for data extraction, Ms. Excel for data cleaning and presentation, Visual paradigm for constructing of Unified Modeling language diagrams, mat lab software for plotting of functions/data, implementation of algorithms and creation of user interfaces. Experimental research design and sampling mechanisms is applied for 15 samples. The methodology in use is fuzzy logic. In order to deal with fuzziness associated with linguistic variables, inference rules are introduced. Five input linguistic terms are identified: Very High, High, Medium, Low and Very Low. Five output linguistic terms are defined to describe the opinion scores: Excellent, Good, Fair, Poor and Bad. Four variables are used: delay, jitter, packet loss and throughput. This results to a total of 625 rules (5^4). The rules are further condensed to 240 logical rules basing on expert knowledge. The collected data was used for simulation in matlab environment basing on the 240 rules. The results shows, analysis of Computer network QoE is subjective in nature rather than objective thus requires a resilient mechanism like fuzzy logic in order to capture clear-cut results to be used for decision making. The target population for this model is the ISPs' clients. This will enable ISPs to have the best responsive measures to deal with clients' QOE parameters so as to meet the QOS as per SLAs.

IJCSIS Vol 17 No 4 April Issue, 2019