What is Multi-Access Edge Computing and why does it matter?

A brief overview about Multi-Access Edge Computing (MEC), its benefits and potential usecases

‍

It is not an exaggeration to say that we are in the middle of a data revolution. The growing share of world population having access to the internet, the evolution of multifaceted and data-hungry applications or virtual worlds like the metaverse as well as their rising usage in every aspect of human life cause an exponential growth of data volume and consequently data load on telecommunication networks. In addition, they tremendously increase the complexity of data handling for telco operators, especially due to the increasing demand for real-time processing and the need to meet regulatory requirements. To respond to these circumstances, new concepts are required, with Edge Computing being a promising one. In the following article, Niklas Schleßmann and Leonard Michalas take a closer look at Edge Computing, outline promising use cases, and discuss why it is an integral concept of Telco’s future 5G network architecture.

‍

Executive Summary

• Edge Computing is an optimized way of Cloud Computing as it processes and stores data in data centers located nearby the end-user.
  ‍
• Several types of Edge Computing exist, with Multi-Access-Edge Computing being one of them.
  ‍
• Multi-Access-Edge Computing (MEC) clouds are deployed within a telecommunication network, i.e. storage and compute power can be provided directly out of a 4G or 5G network.
  ‍
• MEC benefits end-users and operators; It enables use cases with complex requirements regarding latency, speed, real-time, reliability, stability, and security. Additionally, MEC brings potential for telecommunication operators to open new revenue streams and to leverage network efficiencies.
  ‍
• There are various challenges in scaling MEC, with use case interoperability across different telco operator networks being one of them. The Telco Edge Cloud, an orchestration platform that sits on top of the operators’ edge capabilities, could be a potential solution.

‍

Demystifying Edge Computing

Before the mid-2000s, data was mainly stored and processed on infrastructure within the physical boundaries of its user. The private or enterprise users primarily relied on internally managed solutions, but Cloud Computing quickly replaced on-premises infrastructure in many fields. Although traditional Cloud Computing started a new era of flexibility, efficiency, time to market and scalability, there are limitations since servers that store and process data are distributed over multiple geographical locations far away from end users and thus cannot fit all requirements demanded.

‍
Edge Computing is an optimized way of Cloud Computing as data storage and processing is performed nearby end-users ’devices. The technology is still in an early stage of its development cycle and there is no common understanding among Edge Computing providers and customers of what the term exactly means. For a better understanding, Edge Computing can be subdivided into categories with respect to the distance between the location of the cloud where data is stored or processed and the end-users who use the data (see simplified graphic 1):

‍

‍

• In traditional Cloud Computing the data is stored and processed in data centers far away from the location where data is generated or used and thus needs to pass through the whole telco network and the internet to arrive at the cloud service provider’s data center location. (Stage 6)
  ‍
• Public Multi-Access Edge Computing is a type of Edge Computing where compute power, storage, and functions (e.g., 5G user plane or Open RAN functions) are located within a operators’ cellular or fixed-line network – bringing these resources closer to the customers’ end-devices. Subsequently, data from the end application usually does not need to be transmitted via the internet. Consequently, Multi-Access Edge Computing clouds combine the best of two worlds; they allow cloud flexibility and scalability paired with on-premises performance and security.
  ‍
• There are several locations to install Edge Clouds within the telecommunication network. Either in the access network at cell towers or street cabinets (Stage 2), in the transport network at regional data centers (Stage 3), in the core network at superregional data centers on a city-by-city level (Stage 4), or at peering points to networks of other telecommunication operators/internet gateways (Stage 5).
  ‍
• Customer on-premises Multi-Access Edge Computing stores and processes data in data centers that are installed on an enterprise premises, for example in a  5G private network (Stage 1)
  ‍
• End-user device edges store and compute data directly on customer´s end-devices, like smartphones, smart glasses, drones or IoT devices, among others (Stage 0).

‍

Taking a closer look at Multi-Access Edge Computing

As MEC clouds are deployed closer to the users than traditional clouds, MEC has advantages for end-customers and operators (see simplified graphic 2):

‍

• Potential for customers to experience more complex use cases on less powerful devices
  The reduction of the distance data packages/requests need to travel from end-devices via network lines to data centers and back again, leads to many benefits for end users. This includes the enablement of use cases that rely on ultra-low latencies (5-12ms at the Access Network vs. 30-75ms at the cloud), ultra-reliability & stability, or real-time processing. In addition, the local data storage increases security. Moreover, the reduced need for on-device computation, as processing is conducted by the Edge Clouds, enables new device designs and customer experiences, e.g., real-time VR smart glasses.  
  ‍
• ‍Potential for telecommunication operators to provide new services and leverage network efficiencies
  Telecommunication providers have a competitive advantage as they own critical assets, i.e., their telecommunication networks. Thus, they can deploy Edge Clouds themselves or with partners and provide new and differentiated services to their customers. By going down the value chain new B2C, B2B and B2P revenue streams are opened.
  Furthermore, many local Edge Computing use-cases do not require that their data travels throughout the whole telecommunication network or via the internet. This brings benefits to telco operators on the cost side, too, for example due to lowered network usage, distributed data workload, network backbone load reduction, or energy per bit reductions, among others.

‍

What are potential Multi-Access Edge Computing use cases?

Edge Computing use cases may be roughly separated into two groups; Firstly, use cases that rely on data that is time-sensitive or too large to efficiently transport, store, or process centrally or, secondly, use cases that have strong requirements regarding speed, reliability, stability, or security. In order to decide whether a use case rather requires a MEC or a traditional Cloud Computing infrastructure, a long list of features needs to be considered. For simplicity and convenience, we clustered and assessed some exemplary use cases in two dimensions:

‍

Data transmission requirements:

• Low latency, i.e., real-time
• Ultra-reliability & stability
• High bandwidth

‍

Data processing requirements:

• Data sovereignty / privacy / security
• Data storage and compute power
• Reduction of processing needs of an end device

‍

Exemplary results are shown in graphic 3:

‍

For each metric in the two dimensions, a point evaluation from not relevant to highly relevant has been conducted, leading to an overall score of 12 per dimension. In detail, not relevant = 0 points, somewhat relevant =1 point, relevant = 2 points, very relevant = 3 points, highly relevant = 4 points.

‍
As seen in the graphic, a very promising edge use case could be “instant cloudification”. This means users no longer have their data and compute power directly on their device but instantly download and upload their required data to an Edge Cloud for storage and processing. The storage and compute offload from smartphones to the MEC cloud requires high data transmission and data processing capabilities (9 points in each dimension). Customers want to access or stream their data in real-time as well as in an ultra-reliable, stable, and very secure way. Furthermore, the high-speed bandwidth of 5G allows having data instantly downloaded and uploaded. In addition, smaller device storage/processors enable smaller designs and totally new customer experiences, e.g., access to virtual worlds via smart glasses.

‍

Current Challenges of scaling the Edge Cloud

As promising as the topic of Multi-Access Edge Computing sounds some challenges regarding user mobility and interoperability still need to be solved. Although telco operators increasingly setup Edge Computing infrastructure, these infrastructures sit at distributed places within one operator’s network. Handling client movement within the network of one operator as well as the interoperability across edge resources of other national or international operators are still challenges.

Providing uninterrupted services to “on-the-move” clients by complex multi-operator Edge Computing architectures does not only jeopardize key benefits such as ultra-low latency and reliability, but also makes it difficult to deploy use cases. Fast and cost-efficient deployments of use cases are not possible if the Telco Edge resources are not interoperable with each other.

One solution for these challenges is the utilization of an orchestration platform. An orchestration platform sits on top of the operators’ edge capabilities. It manages deployments across the different infrastructures and efficiently handles the traffic of the deployed services between various Edge Clouds. This concept, also known as Telco Edge Cloud, allows to aggregate and consolidate all the operator edge services into a common near-by customer cloud environment. Application providers can utilize this Telco Edge Cloud like their public cloud setups, not worrying about the architecture of the telco network edge resources, while still leveraging its benefits such as ultra-low latency or data sovereignty. One example of such an aggregator platform is MobileEdgeX, which allows application providers and startups to build their 5G applications once and deploy them across all Telco Edge infrastructure without the need to worry about topics such as interoperability, traffic orchestration or handling client movement. Thus, Edge Cloud system platforms like MobileEdgeX are important elements to provide access to the edge resources across various network providers (see case study).

‍

Wayra Tech Lab - The place to shape the future of 5G & Edge Computing

Founders and developers who want to learn more about 5G and MEC can test out the technology at Wayra´s TechLab. Sponsored by Telefónica, the lab enables to trial 5G and MEC use cases in an easily accessible test space. Located in the middle of Munich, the Tech Lab provides necessary technologies, like a 5G public network (NSA) with 70 MHz Telefónica owned spectrum, a 5G private network (SA) with 100 MHz industrial spectrum, a global standardized Telco Edge Cloud and NB-IOT & LTE-M on LTE 800. Moreover, developers and founders can access a large entrepreneurial ecosystem, business expertise as well as development support to bring startups’ 5G and MEC solutions to the next level.

‍

‍

‍

‍

‍

‍