Cloud Computing for Telecom Systems

Abstract: Context: Cloud computing is reshaping the service-delivery and business-models in Information and Communications Technology (ICT). The Information Technology (IT) sector has benefited from it in the previous 3-5 years. Despite the attraction of cloud computing, it is required to have an effective application migration strategy. Cloud computing with its diverse provisioning models makes it possible for telecom vendors and service providers to decide effective service and business models. Currently, cloud computing contains security, performance and dimensioning considerations for telecom companies. Objectives: This thesis assesses the trends and issues associated with the cloud, with telecommunications perspective, while leveraging the cloud to come to a decision on a suitable cloud environment for telecom grade applications. Analysis of maturity of public cloud (in terms of compatibility, consolidation, compliance and standardization) in general and Amazon cloud in particular, is part of the thesis objective. While doing so, deployment of a telecom-grade product in the Amazon cloud will be evaluated against the current on-premise deployment. We want to identify architectural difference between the two domains, and what issues are faced when a migration is planned. This evaluation between two systems, i.e. on-premise and the cloud will significantly contribute to the research and can be used when making business decisions. Methods: We conducted literature review, survey, and a case study, assessing the above mentioned objectives. Research papers from academia and industry were chosen for literature review; personnel, with experience in cloud computing, were chosen for the survey; and a telecom-grade platform was used to assess the migration issues on Amazon cloud in the case study. The Ericsson Composition Engine (ECE) was used to check what deployment issues it can have on Amazon cloud. Its on-premise Reference Deployment Architecture was compared with the cloud-based Reference Deployment Architecture. This case study served as a confirmation to results obtained in the literature review and survey. Results: In the literature review and survey, we found motivations, trends, current applications, and challenges of cloud computing for telecom. It was found from the case study that Amazon Web Services (AWS) lacks application and network centric attributes that are required in ECE deployment. We propose recommendations that can be integrated with ECE while deploying it in a public cloud. Conclusions: Companies are choosing cloud vendors that uniquely give ease of migration and control, based on application needs and compatibility. ECE cannot be directly migrated to AWS, unless we provide Amazon specific modifications in the architecture. The survey and literature review support a private and/or hybrid strategy for ECE, along with the inclusion of cloud networking into the ECE package.