Abstract
The continuous demand for low latency, high reliability, and context-aware content has pushed the existing computational models to their limit. The cloud with its infinite resources can accommodate many of the existing scenarios, however, as new scenarios emerge in the IoT area, the cloud falls short. In this context, the Edge Computing model emerged as an extension to the cloud in support of low latency and high-performance applications, by placing part of cloud resources at the edge of the network, in close proximity to the data sources and applications. The goal of Edge Computing is to provide the same level of abstraction at the cloud but in a local context. However, since Edge Computing inherits many of the benefits provided by the cloud, it also inherits some of its drawbacks. One such limitation is the management overhead needed to set-up and continuously configure the Edge Computing applications. In the cloud space, this problem has been addressed using a new paradigm called serverless technology. Similarly, in the Edge Computing, few attempts are being developed to bring the concept of Serverless Computing at the edge. In this paper, we survey the main edge computing platforms that provide support for serverless computing comparing their characteristics and identifying issues and research directions.