Browsing by Author "Sibanda, Khulumani"

Now showing 1 - 3 of 3
  • No Thumbnail Available
    Item
    The classification performance of Bayesian Networks Classifiers: a case study of detecting Denial of Service (DoS) attacks in cloud computing environments
    (University of Fort Hare, 2015) Moyo, Lindani; Sibanda, Khulumani
    In this research we propose a Bayesian networks approach as a promissory classification technique for detecting malicious traffic due to Denial of Service (DoS) attacks. Bayesian networks have been applied in numerous fields fraught with uncertainty and they have been proved to be successful. They have excelled tremendously in classification tasks i.e. text analysis, medical diagnoses and environmental modeling and management. The detection of DoS attacks has received tremendous attention in the field of network security. DoS attacks have proved to be detrimental and are the bane of cloud computing environments. Large business enterprises have been/or are still unwilling to outsource their businesses to the cloud due to the intrusive tendencies that the cloud platforms are prone too. To make use of Bayesian networks it is imperative to understand the ―ecosystem‖ of factors that are external to modeling the Bayesian algorithm itself. Understanding these factors have proven to result in comparable improvement in classification performance beyond the augmentation of the existing algorithms. Literature provides discussions pertaining to the factors that impact the classification capability, however it was noticed that the effects of the factors are not universal, they tend to be unique for each domain problem. This study investigates the effects of modeling parameters on the classification performance of Bayesian network classifiers in detecting DoS attacks in cloud platforms. We analyzed how structural complexity, training sample size, the choice of discretization method and lastly the score function both individually and collectively impact the performance of classifying between normal and DoS attacks on the cloud. To study the aforementioned factors, we conducted a series of experiments in detecting live DoS attacks launched against a deployed cloud and thereafter examined the classification performance in terms of accuracy of different classes of Bayesian networks. NSL-KDD dataset was used as our training set. We used ownCloud software to deploy our cloud platform. To launch DoS attacks, we used hping3 hacker friendly utility. A live packet capture was used as our test set. WEKA version 3.7.12 was used for our experiments. Our results show that the progression in model complexity improves the classification performance. This is attributed to the increase in the number of attribute correlations. Also the size of the training sample size proved to improve classification ability. Our findings noted that the choice of discretization algorithm does matter in the quest for optimal classification performance. Furthermore, our results indicate that the choice of scoring function does not affect the classification performance of Bayesian networks. Conclusions drawn from this research are prescriptive particularly for a novice machine learning researcher with valuable recommendations that ensure optimal classification performance of Bayesian networks classifiers.
  • No Thumbnail Available
    Item
    Dynamic adaptive cost model for wireless internet connectivity in African Rural communities
    (University of Fort Hare, 2010-01-14) Sibanda, Khulumani
    In today's dynamic technological landscape, wireless communication networks have become an important part of economic development. The emergence of wireless technologies raises hopes to extend communication to remote areas that have not seen any tangible deployment to. date. As developing nations pin their hopes to wireless technologies, cost models for wireless communication networks are becoming vital to support the emerging technologies. However, varying cost changes raise critical challenges to the estimation of both capital expenditure and operational expenditure. The network deployment process has numerous events that may cause adjustments to initially estimated project costs. These adjustments are necessary for a cost management plan and this plan includes monitoring cost performance and ensuring that only appropriate changes are made to the network project. The incidents that may cause cost changes cannot be entirely predicted as their distribution tend to change dynamically from time to time. Estimating network deployment costs in such a dynamic environment necessitates cost models. that can adapt to random occurrence of cost changes.
  • No Thumbnail Available
    Item
    Dynamic adaptive cost model for wireless Internet connectivity in African rural communities
    (University of Fort Hare, 2010) Sibanda, Khulumani; Muyingi, P
    In today’s dynamic technological landscape, wireless communication networks have become an important part of economic development. The emergence of wireless technologies raises hopes to extend communication to remote areas that have not seen any tangible deployment to date. As developing nations pin their hopes to wireless technologies, cost models for wireless communication networks are becoming vital to support the emerging technologies. However, varying cost changes raise critical challenges to the estimation of both capital expenditure and operational expenditure. The network deployment process has numerous events that may cause adjustments to initially estimated project costs. These adjustments are necessary for a cost management plan and this plan includes monitoring cost performance and ensuring that only appropriate changes are made to the network project. The incidents that may cause cost changes can not be entirely predicted as their distribution tend to change dynamically from time to time. Estimating network deployment costs in such a dynamic environment necessitates cost models that can adapt to random occurrence of cost changes. Widely used cost models are usually performed by experienced personnel whose engineering experience is derived from deploying similar networks. In this approach experienced personnel add a certain percentage to the cost estimate to cater for contingency costs. Certainly such an approach depends on individual opinion, making it subjective and void of mathematical estimating relationships which are of paramount importance in ensuring that estimated deployment costs are sufficient to deal with cost uncertainties. We observe that existing approaches can only explore a limited solution space and hence can lead to cost overruns if implemented in dynamically cost changing environments. This thesis presents a wireless communication network deployment cost model that incorporates uncertainties into the final cost estimate. The model is adaptive to unpredictable cost changes since it allows adjustments of confidence levels when calculating contingency costs. This allows dynamically updating the cost changes without the cost model being reconstructed from scratch. We make use of the Poisson process in modeling the occurrence of incidents that are responsible for causing cost changes during network deployment. We also show that the occurrence of the incidents causing cost change are random and tend to follow the Poisson distribution. Using different levels of confidence we model various cost contingencies and make sensitivity analyses to identify the probability of cost overrun when given different contingencies. The dynamic adaptive cost model can be used either at the strategic level to understand the cost of a particular technique or at the operational level, as a way to show how Poisson process in network deployment can compare with engineering experience and other estimating techniques. We believe that the model is useful for remote areas where deployment costs are volatile and the distribution of incidents causing cost change to original cost estimates are diverse and dynamically changing. Further we expect that our research improves the knowledge base of information about the costs for rural communities to connect to the Internet, consequently providing useful input to future policy debates. This work is further poised to be a utility function to help those planning internet infrastructure deployments in least developed regions.