Multi-access Edge Computing (MEC) and 5G

Edge computing is a concept that:

  • enables low latencies for time-critical services, known in 5G as Ultra-Reliable Low-Latency Communication (URLLC) services
  • handles bandwidth-heavy applications closer to the end user in order to offload the transport network

European Telecommunications Standards Institute (ETSI) is leading standardization activities around Multi-access Edge Computing (MEC).

MEC provides compute and storage resources for applications geographically close to the end users. This enables high-bandwidth, low-latency access to services. Real-time radio network information can be used to optimize and adapt the services to the current radio situation.

Several 5G use cases are expected to rely on MEC to deliver added value for services to the end users e.g. inside factory buildings, hospitals, ports as well as in supporting technologies such as virtual and augmented reality.

It is estimated that by 2023, 5G will make up around one-fifth of all mobile data traffic, where 25{ab0bbb36d35f56bd2c9e68629e6e8f1a44a0c9776dc977281f4aeafa2fb487ad} of the use cases will depend on edge computing capabilities. The majority of new 5G revenue potential is expected to come from enterprise and IoT services, of which many will rely on edge computing.

In addition to providing an execution environment for running applications at the edge, MEC also enables services that supply information on end user and base station context (Radio Network Information Services), such as the radio channel quality of users and their location in the network, which allows design of context-aware applications. A context-aware system is one that can determine and react to the current physical and computing context of mobile users and devices.

The ETSI MEC Industry Specification Group (ISG) has been working on the development of standardization activities around MEC since 2013. Its first released document covers the reference architecture.

The MEC host provides the virtualization environment to run MEC applications, while interacting with mobile network entities via the MEC platform (MEP) to provide MEC services and data offload to MEC applications. Two MEC hosts can communicate via the Mp3 interface aiming at managing user mobility via the migration of MEC applications among MEC hosts.

The MEC platform (MEP) acts as an interface between the mobile network and the MEC applications. It has an interface (Mp1) for MEC applications to expose and consume MEC services, and another (Mp2) to interact with the mobile network. The latter is used to obtain statistics from the Radio Access Network (RAN) on User Equipments (UEs) and Next Generation NodeBs (gNBs), e.g. to provide the Radio Network Information Service (RNIS) and the Location Service, and to appropriately steer user plane traffic to MEC applications.

MEC applications run on top of a virtualized platform. MEC services provided by third-party MEC applications should be registered with the MEP and made available over the Mp1 reference point. Once registered, a service may be discovered and consumed by other MEC applications.

For the management plane, ETSI MEC introduced the Mobile Edge Orchestrator (MEO), which is in charge of the life-cycle of MEC applications (instantiation, orchestration and management), and acts as the interface between the MEC host and the Operations/Business Support System (OSS/BSS).

For the management plane, ETSI MEC introduced the Mobile Edge Orchestrator (MEO), which is in charge of the life-cycle of MEC applications (instantiation, orchestration and management), and acts as the interface between the MEC host and the Operations/Business Support System (OSS/BSS).

In the 5G architecture, the MEP will be integrated as a 5G Application Function (AF), trusted or untrusted, depending on the use case. It may request traffic redirection for a MEC application to a location “closer to the edge”. If the MEP is a trusted 5G AF, it can directly use the Policy Control Function (PCF) to generate a policy to offload traffic towards the MEC application. If it is not considered as a trusted 5G AF, it uses the Network Exposure Function (NEF) to access the Session Management Function (SMF), via its traffic filter policy API, and requests the traffic redirection.

One interesting question that often pops up is: what – or where – is the actual edge of the network? Depending on who you ask, you’ll be given a variety of very different answers. Obviously, there is no standardized definition of the word edge, so it very much depends on the application as to how close to the user the application needs to be. In some cases, edge computing requires that the hardware (and software) for the application needs to be physically co-located with the gNB inside a production plant, for example. For caching popular media content (e.g. a new episode of some popular TV series), a neighborhood or even a city can be considered to be the edge.  

Leave a Reply

Your email address will not be published.