Recognizing the limits of (distant) cloud computing, modern Internet of things (IoT) architectures recognize the so-called edge continuum between cloud applications and IoT devices and machines. “Continuum” implies that “edge” does not refer to a single spatial position but to a multitude of options for (physically) placing the different technologies required for a true end-to-end IoT solution.
Four of these technology options are briefly introduced, discussed, and superficially compared in the first part (eight pages) of the survey, with the bulk being devoted to selectively and shortly presenting current (13 pages) and potential future (four pages) research topics for the respective technologies.
Besides presenting the two well-known concepts of multi-access edge computing (still called “mobile edge computing,” which is now obsolete) and fog computing, I found the inclusion of transparent computing and cloudlets very interesting. The technologies are explained on a level accessible to anyone interested in information technology on a fairly general level, and largely good. As always, the authors fail to provide good criteria for differentiating between fog and edge.
Current research topics cited revolve around energy optimization, latency minimization, the ubiquitous security and privacy concerns, and, surprisingly, caching. Topics are typically described in a short paragraph and, apparently, not ordered or further structured in any way.
This style continues in Section 6, “Future Research Directions,” which contains possibilities such as block-streaming service loading, lightweight virtualization, and flexible network operating systems (and again security).
The survey is easily accessible and contains lots of concepts and stimulating facts. It is recommended for the non-IoT specialist with interests in one of the several topics indicated here. Experts, though, should be aware that the content reflects the state of the art as of January 2019.