Edge Computing: Vision and Challenges
CASE STUDY
Cloud Offloading
In the traditional content delivery network, only the data iscached at the edge servers. This is based on the fact that thecontent provider provides the data on the Internet, which istrue for the past decades. In the IoT, the data is produced andconsumed at the edge. Thus, in the edge computing paradigm,not only data but also operations applied on the data shouldbe cached at the edge
Another issue involves the collaboration of multiple edgeswhen a user moves from one edge node to another. One simplesolution is to cache the data to all edges the user may reach.Then the synchronization issue between edge nodes rises up.
Similar applicationsalso include the following.
- Navigation applications can move the navigating orsearching services to the edge for a local area, in whichcase only a few map blocks are involved.
- Content filtering/aggregating could be done at the edgenodes to reduce the data volume to be transferred.
- Real-time applications such as vision-aid entertainmentgames, augmented reality, and connected health, couldmake fast responses by using edge nodes.Thus, by leveraging edge computing, the latency and con-sequently the user experience for time-sensitive applicationcould be improved significantly.
Video Analytics
Cloud comput-ing is no longer suitable for applications that requires videoanalytics due to the long data transmission latency and privacyconcerns.
Here we give an example of finding a lost childin the city.
With the edge computing paradigm, the request of searchinga child can be generated from the cloud and pushed to all thethings in a target area. Each thing, for example, a smart phone,can perform the request and search its local camera data andonly report the result back to the cloud.