School of Computing & Informatics
https://repository.maseno.ac.ke/handle/123456789/1325
2024-03-29T10:56:47ZComparative analysis of the performance of open shortest path first and openflow on quality-of-service metrics in a large simulated mobile core network
https://repository.maseno.ac.ke/handle/123456789/5607
Comparative analysis of the performance of open shortest path first and openflow on quality-of-service metrics in a large simulated mobile core network
ODHIAMBO, Joseph. Nicholas Omumbo
Mobile application systems are increasingly being defined using Internet Protocol (IP). Therefore, one of the challenging tasks that mobile IP phone service companies have to deal with as their networks expand is an upsurge of user equipment (UE) and routers, which places excessive demand on the routing infrastructure and results in degraded quality of service (QoS) values at the mobile core network. Studies show that most mobile IP phone subscriber companies struggle to meet acceptable QoS values due to their continued use of classical routing methods. Classical routing protocols such as Open Shortes Path First (OSPF) lose efficiency as the network size increases, which leads to poor QoS values. An attempt to explore emerging Software Defined Network (SDN) technologies to enhance QoS and promote a centralized controller-basedrouting method has emerged in the recent past. OpenFlow routing is an instance of SDN found to improve routing in wired and small wireless networks. This study'smain objective was to compare the performance of OSPF and OpenFlow routing in a large mobile IP core network. Specific objectives weretocompare the performance of OSPF and OpenFlow routing on jitter, packet delivery ratio (PDR), throughput, and end-to-end delay. This research adopted a mixed design methodology consisting of exploratory and experimental research design where Objective Modular Network Testbed in C++ (OMNeT++) network simulator was used with simuLTE add-on to model two side-by-side mobile IP networks core architecturerunning OSPF and OpenFlow protocols. Applications that present high demand for processing and routing used; interactive gaming, VoiP, audio streaming, and Internet Protocol Television (IPTV) were used to test routing efficiency. The setup test environment consisted of 1000 UEs, 80 OSPF routers, and 80 OpenFlow switches which are considered a large network. Notably, Jitter was improved by 10 milliseconds when OpenFlow routing was used compared to OSPF. This improvement was consistent across the network with the addition of more UEs and Routers. In instances where OSPF improved PDR, the value is less significant, with a standard deviation of 0.7 mbs. This study demonstrated that OpenFlow could improve routing efficiency in large simulated mobile IP core networks compared to the OSPF routing network by over 60 percent across the four QoS metrics considered in the experiment. The study also confirmed that the network size impacts the QoS parameters. This study further recommends testing OpenFlow in large mobile IP production networks.
Masters Thesis
2022-01-01T00:00:00ZAn evaluation of hybirdmachine learning classifier models for identification of terrorist groups in the aftermath of an attack
https://repository.maseno.ac.ke/handle/123456789/4072
An evaluation of hybirdmachine learning classifier models for identification of terrorist groups in the aftermath of an attack
OKETCH, Peter Opiyo
Terrorist attacks have globally led to loss of life and property, fear, and general insecurity. Terrorist acts are planned and perpetrated by collections of loosely organized people operating in shadowy networks that are difficult to identify. Machine learning classifier algorithms have been used in accurate identification of terrorist groups and weapon types in India, Egypt, Pakistan, and United Kingdom. However, the urgency of responding to a terrorist attack and the subsequent nature of analysis required to identify the terrorist group involved in an attack demands that the performance of the classifiers yield highly accurate outcomes. The concept of combining classifier algorithms into hybrid is proposed as a new way of improving the accuracy. To date there has not been sufficient research that attempts to find combinations of Naïve Bayes, K-Nearest Neighbor, Decision Trees, Support Vector Machines and Multi-Layer Perceptron as base classifier algorithm modelsand resample sample size percent for optimum accuracy in the identification of terrorist groups in the aftermath of an attack. The aim of the study is to build and evaluate hybrid classifier algorithm models for identification of terrorist groups. Specifically, it builds and evaluates base classifier algorithm models, builds, and evaluates hybrid classifier algorithm models by combining and evaluating the base classifier algorithm models,and compares the performance of the classifier algorithm models. The study adoptsa randomized block experimental research design using Waikato Environment for Knowledge Analysis (WEKA) tool for building and evaluating the classifier algorithm models, and 1999-2017 sub-Sahara terrorist dataset from the Global Terrorist Database (GTD). The features country, region, attack type, target type, group name and weapon type are ranked highest of 23 attributes of the dataset for identification of the terrorist group name the using WEK Afilter-based search and ranker routine. Data imbalance in the dataset is addressed by varying resample sample size percent for optimum performance. The classifier algorithm models were evaluated and compared on accuracy and build time as performance metrics using 10-fold cross validation, test split and ANOVA test. The results suggest that hybrid classifier algorithm models yield higher accuracy rates, accuracy rates for 10-fold cross validation are higher than the rates for test split and that resample sample size percent as a technique to solve class imbalance affects accuracy and yields optimum accuracy rates at resample sample size percent of 1000 for the available dataset. The results show a significant improvement in accuracy between the control group and the experimental group.The study concludes that hybrid KD (a combination of K-Nearest Neighbor and Decision trees) outperformed all other classifier algorithm models at resample sample size percent of 1000 with an accuracy rate of 88.18% and build time of 0.03 seconds for 10- fold cross validation and accuracy rate of 87.66% and build time of 1.03 seconds for test split in the identification of terrorist groups in the aftermath of an attack for the sub-Sahara Africa dataset.The study makes contribution by developing a systematic process of building a hybrid classifier algorithm model and establishing a resample sample size percent of 1000 for optimum accuracy rates for the dataset.
Masters Thesis
2020-01-01T00:00:00ZIntegration of a Memory Analyzer to the Browser Reference Architecture
https://repository.maseno.ac.ke/handle/123456789/1437
Integration of a Memory Analyzer to the Browser Reference Architecture
KARIUKI, Kamau Harun
A Web Browser is a computer application used to access information on the World Wide Web.
The browser‟s parsing capability has advanced over years since its inception. The advancements
have consequently increased demand for memory as manifested by computer crawl.
Contemporary browsers are anchored on reference architecture that lacks memory control
mechanism that can limit maximum memory a browser can use thus posing a challenge in
multiprogramming environments with less memory thereby making the computer to freeze.
Enhanced browser reference architecture was developed for investigation. The main objective of
the study was to develop and integrate a memory analyzer to the browser with a view to
evaluating its performance in Web browsers. Specific objectives were to specify the functional
requirements for the browser prototype, to design and develop a browser prototype, to design,
implement, and integrate memory analyzer and to evaluate the performance of the memory
analyzer in the developed architecture. Prototyping technique and software reuse were adopted in
formulating the model. The memory analyzer component acted as a memory meter and a
memory optimizer. It controlled memory hogging by limiting memory usage to a particular value
set by the user and optimizing available memory by calling the garbage collector. Experiments
were carried out to validate the Mozilla–based developed prototype by using Mozilla Firefox
browser as a control. All tests were carried on windows environment in parallel. Memory
consumption between the two browsers was recorded and statistically analyzed to test the
researcher‟s hypothesis. To evaluate the performance of the analyzer, memory demands posed by
access to popular sites such as electronic mail service providers, social networks entertainment
and search engines were examined. Statistical T-test on memory consumption between the two
browsers revealed that memory analyzer-integrated browser consumed 38.65 MB and 52.08 MB
less with homogeneous and heterogeneous tabs respectively compared to contemporary Mozilla
Firefox browser. This value is computationally significant as it provides suitable environment
that facilitates concurrency in computer systems that have low memory. The study provides
insights on the performance of enhanced browser reference architecture with regard to memory
optimization. The study recommends further research on memory optimization approaches, as
browser memory consumption is dynamic and browser technologies change often.
2019-01-01T00:00:00Z