Securing Multi-Tenant Cloud Environments with Graph-Based Models
Sai Kiran Reddy Malikireddy
Independent Researcher, USA
Abstract
Multi-tenant approaches like cloud computing are vulnerable because separate tenants share a single hardware and networking platform. Privacy, acquisition, and segregation/removal of data resources are big challenges that they face. As a result, the following security challenges can be mitigated by harnessing the fairly recent graph-based models that allow for a more logical depiction of tenants, resources, and services herein. This paper presents the use of graph theory in a multi-tenant cloud to enhance the cloud environments' security in terms of access control, anomalies, and risk measures. These changed cloud resources and the tenants' touch points can be modeled as graph structures to construct security models that are useful in continuously assessing risks, identifying pre-specified anomalies, and containing them where necessary. Moreover, the paper provides an overview of existing graph-based techniques and algorithms such as graph search, community detection, and machine learning for anomaly detection for security improvement of multi-tenanted cloud platforms. These models help prevent cross-tenancy data compromise and framework invasion and illustrate VM deployment for controlling the battles over scarce resources through an actual example's plausibility. The work also includes negative aspects such as scalability, privacy invasion, and integration with conventional security models, with corresponding research areas considering the interaction with AI and Blockchain. However, the models based on graphs offer a rather sound approach to providing specific multiple-tenant security in the cloud; further developments remain imperative for enhancing cloud security.
Keywords: Multi-Tenant Cloud Computing, Cloud Security, Graph-Based Models. Access Control, Anomaly Detection, Data Privacy, Resource Isolation, Risk Management, Machine Learning, Community Detection, Graph Theory
