Abstract
The paradigm of edge computing emerges as the next revolution in computer networking. Cloudlets – as an implementation of edge computing – are mobility-enhanced small-scale data centers on the edge of the network. They serve users located in their geographical proximity. This proximity brings various advantages, such as low latency and better service quality. Prior studies investigating the optimal placement of cloudlets assume a global view with complete information – an unrealistic and not really implementable assumption. This study presents a more realistic view on the matter and accounts for the information asymmetries among the stakeholders involved in a cloudlet scenario. Further, we introduce a system design that enables efficient exchange of relevant placement information, with the ability to dynamically adapt to contextual factors. Since the effectiveness of placement decisions depends on service usage, which is subjected to constant change, our system design can bring a significant improvement related to the cloudlet placement decisions.
Recommended Citation
Felka, Patrick, "Smart Communication Infrastructures: A Probabilistic and Context-Aware System Design for Efficient and Dynamic Cloudlet-Management" (2019). AMCIS 2019 Proceedings. 7.
https://aisel.aisnet.org/amcis2019/data_science_analytics_for_decision_support/data_science_analytics_for_decision_support/7
Smart Communication Infrastructures: A Probabilistic and Context-Aware System Design for Efficient and Dynamic Cloudlet-Management
The paradigm of edge computing emerges as the next revolution in computer networking. Cloudlets – as an implementation of edge computing – are mobility-enhanced small-scale data centers on the edge of the network. They serve users located in their geographical proximity. This proximity brings various advantages, such as low latency and better service quality. Prior studies investigating the optimal placement of cloudlets assume a global view with complete information – an unrealistic and not really implementable assumption. This study presents a more realistic view on the matter and accounts for the information asymmetries among the stakeholders involved in a cloudlet scenario. Further, we introduce a system design that enables efficient exchange of relevant placement information, with the ability to dynamically adapt to contextual factors. Since the effectiveness of placement decisions depends on service usage, which is subjected to constant change, our system design can bring a significant improvement related to the cloudlet placement decisions.