A team of researchers have proposed a method to use cloud and fog, or edge, computing structures to complement one another – rather than viewing edge computing as a replacement for the cloud. Using Software-Defined Networking (SDN) to manage the interaction between cloud and edge resources, a network can remain dynamic, agile and efficient while providing a better experience for the end user.
Increased use of mobile devices has created stresses on cloud networks, which will only increase as mobile device use increases worldwide. Creating a system where cloud and edge computing resources are unified is a potential response to the challenges of overtaxed resources and unexpected latency, which can cause a degraded quality of experience for the end user.
However, combining resources in the cloud and at the edge creates specific challenges for the system. There must be a local coordinator to divert tasks to the appropriate resources in real time, in a dynamic and unpredictable environment. The system must be constantly updated in real time to provide the best information regarding available resources, and have open, programmable interfaces to divert tasks in the most efficient ways.
The research team created a software-defined network (SDN) enabled architecture to meet these challenges and create a usable network that uses cloud and edge computing capabilities simultaneously.
First, the SDN provides real-time knowledge of available resources that is both flexible and reliable. A centralized controller allows for optimal decision making for each unit within the system as a whole, and a dedicated control channel allows translation of high-level policies to low-level configuration instructions to give the system fine-grained control.
Using SDN architecture to combine cloud and edge computing within a single network is intended to provide an improved quality of experience for the end user, as well as better resource pooling economics and agile network management.
The team conducted two case studies to test whether a combination of cloud and edge computing performed better than a cloud system alone. In the first, they analyzed the responsiveness of the system to computational crowdsourcing tasks. In the second, they considered whether SDN-enabled architecture could be used to identify which videos are most popular in a certain time frame, and cache them differently to allow easier viewing and an improved quality of experience for the end user.
In both case studies, the researchers concluded that using SDN-enabled architecture allowed a network to use resources of cloud computing and edge computing interchangeably, capitalizing on agile and dynamic system requirements, to deliver an improved service to the end users. However, further study is recommended that focuses on security of data in cloud-edge combined networks.