Trillions of gigabytes of data is being generated/captured by devices and network systems, which need to be analyzed and processed. Forbes estimates that in the year of 2025, 150 zettabytes of data will need to be analyzed and processed. Over half of this data is expected to come from the edge of the network (79.4 Zetta bytes by 2025). The world’s most valuable resource is no longer oil, but data. While oil is a limited natural resource which needs to be preserved and conserved. On the contrary data is experiencing explosive growth and showing no signs of slowing down anytime soon. Data need to be managed, processed and analyzed to derive value from it. Conventionally, this data is sent to the centralized systems usually in the cloud for processing, analyzing and deriving insights. It is an enormous task at hand. This method of processing the data incurs significant latency and huge amount transport cost. Which often renders the derived stale insights not useful for most latency sensitive applications.
Edge computing (EC) addresses and mitigates some of these issues. In this paper we look at different aspects of EC, comprising of addressing the latency in data processing, analyzing and deriving insights; architecture, standards and deployment considerations.