| Popis: |
Internet age thrives on connectivity features and accessibility options, while denial of services attacks are detrimental to accessibility. In this context our research is making use of software exposures datasets from National Institute of Standards and Technology for analysis of denial of services (DoS) exposure of various products of three major vendors named as Microsoft (MS), Apple (AP) and Google (GO). A non-linear logit regression model has been used to force binary predictions for variety of DoS related vulnerabilities, given the values of independent variables established through the odd ratios of dependent variables including accessibility type, complexity and scores of attacks, authenticity level as well as confidentiality, accessibility and/or integrity violations. Interesting results compares and contrasts three vendors. Availability has bigger effects for two vendors (MS and AP) but not for GO in case of dependent variable DoS as well for all three vendors for dependent variable of (DoS +Info) vulnerability type which is also heavily influenced by Confidentiality in all three vendors, although impact on MS for both these independent variables on both of these dependent variables are too high as compared to other two vendors. Other heavy affects have been noted on (DoS +Priv) from Access and Integrity for MS and confidentiality for GO, on DoS Mem. Corr. from availability and on DoS Exec Code Overflow Mem. Corr. from complexity both for AP. |
