Intelligent Cloud Security Issues Detection Using Mamdani Fuzzy Logic

Abstract

Cloud computing is computing that provides, storage, databases, networking, intelligence, software, and analytics over the internet. Cloud services are delivered remotely and almost always from an offsite data center. Cloud services control computing infrastructure efficiently. This research presents cloud security issues and challenges and tries to find out its possible solution. It demonstrates some solutions dealing with cloud computing-related privacy and security challenges. This study uses a Multi-Layered Mamdani Fuzzy Inference System (ML-MFIS) model. To detect cloud security issues and challenges, the author proposed an Intelligent Cloud Security Issues Detection model using a Multi-layer Mamdani Fuzzy Inference System (ICSID-ML-MFIS), that can classify different types of threats. The ICSID model has eight input variables at layer-I and three input variables at Layer-II respectively. In layer-I input variables are threat-to-software (TS). Traffic Monitoring (TM), Networking Threat (NT), Resource Availability (RA), Platform availability (PA), Trusted-Service-Availability (TSP), Device Availability (DA), Network Availability (NA) that detects output condition of threats to be affected or Not-Affected. At layer-II input variables are Detect SAAS Threats (DSAAST), Detect PAAS Threats (DPAAST), and Detect IAAS Threats (DIAAST), which determine the inference-result is either affected or nor-affected. In the end, the output layer detects cloud security issues by determining various threats that occur such as lack of visibility of data, theft of data, failure to control data, hijacking, system weakness, social engineering attacks, data breaches, and no-security issues. In the end, the output determines the condition as Yes or No. The proposed model based on Fuzzy reached 91.5% of true positive cases.

Keywords: Big Cloud Security issues, Detection of threat, SaaS, IaaS, PaaS, fuzzy logic