The current 4G mobile networks are strained to their limits, and it is estimated that the number of connected devices will reach 100 billion by 2025. 5G’s performance goals of ultra-fast download and ultra-low latency will be impaired if all 5G application traffic has to navigate its way to the centralised cloud for computing, and then back to the connected device. Mobile cloud computing can be either centralised or distributed in nature. The computing power required for social network services like Facebook, Twitter and Instagram, and navigation tools like Google Maps are all hosted on the centralised cloud, where all the computing processes occur.
Mobile Edge Computing, or Multi-access Edge Computing (MEC) aims to reduce congestion on mobile networks, by enabling the computing that would normally be done in a centralised cloud to be done more locally, on the edge of a mobile network, closer to the connected devices. MEC is already being deployed over WIFI and LTE networks, so it needn’t necessarily wait for 5G to happen, but the complementary benefits for 5G are huge.
MEC will move the computing of traffic and services from a centralised cloud and closer to the consumer on the edge of the network. Mobile Edge Computing involves the evolution of the base station from being purely communications based, to being computing and communications based.
The edge computing hardware at the datacentres themselves will be much like at any other datacentres; clusters of racked servers in cooled units providing the required local computing power.
These server clusters may be placed at the base station, or close to a number of base stations, all of which can use the pooled resources at the same datacentre whilst still retaining the low latency. Applications which are computing intensive or latency sensitive can then be hosted and processed there.
This would allow the mobile operator to host or provide a range of applications and services that benefit from low latency, as well as reduce the amount of data traffic that needs to be sent back to the core network.
Edge Computing can also be used to reduce network transport costs for the mobile operator – costs are reduced by processing data at the edge rather than transporting it back to the core.
Machine-to-machine communication – the foundation of the Internet of Things (IoT) – has been around for decades, but now there are so many connected devices, transmitting huge amounts of data at great speed, that centralised cloud computing is no longer a realistic option for devices which require data processing in the low milliseconds.
MEC brings the resources of computing, storage and networking closer to applications, devices and users. Ultra-low latency and uniform data rates will allow the transfer of high-resolution data in real time. Whilst 4G offers around responsiveness of around 50ms, 5G promises 1ms but there would be more lag if the data and computing had to travel all the way to the cloud for processing.
MEC allows the round trip for the data to be much shorter, enabling actions to happen much quicker. A good example of the need for low latency is autonomous driven cars: The car must sense the environment and respond immediately to apply its brakes if an obstruction appears in the road, thereby preventing an accident. Whilst the data may be sent to the cloud for monitoring and vehicle management, the critical millisecond reaction time data are collected, processed and analysed at the edge via MEC.
Virtual and Augmented Reality also require ultra-low latency to deliver fast and highly localised feedback, to deliver real-time information and overlay it on the real environment. The technology, in conjunction with MEC, will enable immersive education, shared personal moments and social interactions in real time, with a uniform experience and zero lag time.
This approach would be a low risk for the operator, as the infrastructure is only built when required. The telecom company acts as an enabler of edge computing, hosting the computing and storage hardware already connected to the network. The customer or partner would run its own software on the MEC.
One such example would be a virtual Content Distribution Network (CDN) model: The CDN service providers use the dedicated edge hosting to facilitate very fast delivery of content owned by the likes of Netflix, HBO and YouTube to end consumers.
The predicted growth in connected and even autonomous vehicles in coming years would see the computing resources of dedicated edge hosting also being sold to car manufacturers, who offer services to car owners.
In this business model, the telecoms company acts in the same way as a cloud provider, providing distributed storage and computing capabilities; a platform for developing applications on the edge infrastructure, and network services in an ‘as a service’ manner through a cloud portal as the customer interface.
The big difference to dedicated edge hosting for the network operator, is that there are no guaranteed customers beforehand. There is a higher risk for the telecoms company as it needs to invest in MEC coverage and widely deploy the edge infrastructure up front, before a revenue stream is established.
This model is more scalable however, and there is the possibility of a larger ongoing revenue stream, assuming sufficient adoption among customers.
Customers might be IoT application providers who want to optimise applications so that they can analyse the data from devices to trigger actions quickly. Other potential customers include large enterprises, systems integrators, CDN providers, content owners and other cloud providers.
This business model would offer custom turnkey solutions for enterprise customers with specific requirements, some of which are met by MEC functionality.
An example would be a city or local council investing in a MEC solution for their smart city project which would require deployment of the MEC infrastructure and any necessary hardware (sensors, actuators and devices), integration of different networks and systems, and orchestrating the development of the smart city solutions and applications.
This model would not require significant investment from the network operator as the MEC infrastructure would be under written by the customers.
More of an ‘off -the-shelf’ product than the systems integration model above, the telecoms company may offer MEC-enabled business solutions for government, enterprise or SME customers. This may be to improve existing processes, or to contribute to an end customer offering (B2B2X). These offerings would require significantly less integration work than systems integrati on projects.
A good example would be a service for large events with significant risk of network congestion (sports stadiums, music concerts) to enhance the attendees’ experience using MEC to provide immersive low-latency video, or live video streams from dif erent viewpoints. This can be monetised and off ered as an additional paid service for attendees.
Another model would be CCTV video surveillance. Transmitting all CCTV feeds for central processing is uneconomical, but footage could be analysed on the edge, and only events deemed important enough would trigger notifications to CCTV operators and send the feeds to the cloud for further analysis. Recurring revenues in this instance might be a service fee for each connected camera.
The risks associated with this business model are mainly due to the upfront service development and the uncertain take up of the solution by enterprise customers
In this model, the telecoms operator acts as a digital service provider for consumer applications. Applications in this category will leverage the benefits of MEC, like low latency, high throughput and context awareness, to provide consumers with innovative applications. These could be Internet of Things (IoT), Augmented Reality (AR), or Virtual Reality (VR) applications that require video transmission in real time, like live sport or gaming.
There is potentially high revenue in this business model, but significant investment would be needed to extend an existing content offering.
An ever greater supply of bandwidth will inevitably result in a diminished value perception of core services like voice, text, and data – and therefore pricing – for the end user. Consumers will be looking for content and value-added applications, whereas enterprise, industry and governments will be looking for solutions.
It will be up to mobile operators to assess which MEC business models offer them the best return on their 5G investments, and getting that choice right will be one of their key strategic objectives in the months and years to come.