Mobile Edge Computing (MEC) – Meeting The Latency Requirements for 5G networks

In a recent article, I wrote about the use of multiple frequency bands to achieve a careful combination of coverage and capacity, to achieve connectivity over large geographic areas, faster speeds, exponential and intelligent connectivity, ultra high bandwidth, etc, in 5G networks.

Besides the above promises, 5G networks also hope to achieve a low latency of less than or equal to 1ms; this is critical for applications like immersive entertainment, remote/robotic surgery, autonomous cars etc., where a delay could have damning consequences.

Latency here refers to delay, and is primarily governed by physical laws. Using the speed of light (3×108 m/s), light or electromagnetic (em) waves would travel over a distance of 300m in 1µs. this means for a content to be delivered in 1ms, the content should be located within a distance of 300km. Through careful research, GSMA estimates that the content should be 1km away from real communication networks, in order to achieve a latency of 1ms. In essence, a low latency gives the users the perception of infinite capacity.

Edge computing, the distribution of computing and communication resources closer to the edge of the network or user, has been proposed for use in 5G networks, to help address the low latency requirements. Interestingly, this is an area that has been dominated by the cloud computing providers like Google, Microsoft, amazon etc. The use of edge computing in 5G networks has also been seen as a market opportunity for telcos looking to diversify their offerings beyond connectivity but they know they would face an uphill battle from the cloud computing operators.

Interestingly, it seems this would be a partnership opportunity rather than a competition, contrary to industry speculations. This is because the cloud computing operators, no doubt own the cloud platforms, but are also actively looking for ways to deploy closer to the user by integrating into IoT devices, supporting private and hybrid cloud deployments etc., in order to diversify their offerings beyond public cloud offerings.

Telcos, on the other hand, have various physical locations, to their advantage, where the edge computing resources could be deployed. They also know that they would benefit greatly from partnering with cloud computing operators to provide a distributed computing environment within their markets. Furthermore, they could equally largely benefit from specific applications, mandated by local laws to store and host data locally. Besides, this would significantly reduce the required investment (and expertise) needed to benefit from this market, should telcos decide to venture on their own.

Though, it looks like both parties (telcos and cloud computing operators) would benefit from partnerships, rather than competition, telcos however need to have a clear long term strategy when formalising such partnership arrangements and constantly stay ahead of the trends within the industry. This is because the cloud computing operators are known to be very innovative in meeting new and future customer demands and have proven to be very agile and flexible in the delivery of their services. Telcos, on the other hand, are known for being generally slow to react to varying customer demands. Only time will tell how this would eventually play out.