The Edge-Cloud Continuum: From IoT to Serverless Computing

Dear Colleagues,

Today, Cloud and Edge Computing technologies have reached a high level of maturity and are ready to become truly ubiquitous in countless areas for the benefit of business, economic growth, and society at large. These new technologies have the potential to realize a computing continuum that comprises end-to-end: user devices, intermediate hubs and private or public clouds.

The term ‘edge’ generically refers to the collection of devices with sensing, actuation, processing, storage, and communication capabilities available on the customer premises. Edge computing is the consolidated trend of shifting the balance of application execution from remote clouds to the edge. This edgification trend has been driven greatly by the proliferation of IoT systems, in which edge devices carry sensors that collect and possible store data temporarily; performing some part of processing on the edge or on hubs close to the edge is desirable.

For a cloud computing environment, serverless computing is the emerging paradigm for application design. The latter dictates that applications are created in a modular manner, so that application parts can be executed separately, on different resources, in an effort to achieve the optimisation goals. More specifically, the term ‘serverless’ refers to services that are server-agnostic, meaning that details like configuration, management, billing, etc. are abstracted from the end-user. Therefore, application code can be deployed on flexible platforms with auto-scaling features and pay-as-you-go plans with much finer granularity than other cloud abstractions.

The Internet of Things (IoT) is a paradigm of networking where the networked points are made of uniquely identifiable physical or virtual entities that are commonly referred to by “things”. Things are an abstraction used to denote devices that are equipped with embedded network connectivity at a minimum, along with sensors and (possibly) actuators to interact with the environment. Sensors are embedded in devices, vehicles, buildings, fields, etc, in order to serve as the end points on the edge of an IoT system. Among the most relevant enablers of the IoT paradigm are the widespread availability of low-cost computing devices, Internet connectivity, cloud computing, and big data tools for extracting intelligence from the gigantic amounts of data that are produced by the ubiquitous things. More and more large-scale IoT systems tend, or need to, operate on combinations of edge and cloud infrastructures. Numerous and multifarious sensing devices collect data rapidly, which need to be pre-processed close to their collection point, in order to avoid data loss, increase response time for time-critical actions, unload the system from network traffic and use the most central and powerful parts of the infrastructures for demanding data analysis and management.

Prof. Verena Kantere

Guest Editor