- Traditional networks had been designed for voice providers and are now not appropriate for MBB providers.
- Re-transmitting MBB providers from the Internet to cell networks for every person wastes an enormous quantity of bandwidth.
Key takeaways:
Mobile Edge Computing (MEC) structure positions community capabilities and third-party functions on the community edge, enabling utility, content material, and community orchestration. Processing providers closer to the person additionally supplies customers with better-quality providers.
User expectations on service expertise have risen in tandem with wi-fi community speeds. Meeting this increased demand requires a brand new sort of structure and repair supply methodology – one that may ship extra thrilling providers.
This new structure is MEC, service options for which had been demonstrated at MWC 2016. Huawei’s MEC@CloudEdge is a 5G-oriented MEC resolution.
From flat to edge
Mobile networks have developed with much less service processing nodes and delayered structure. In the 2G period, customers had to move by way of the bottom station, base station controller, SGSN, and GGSN, earlier than connecting to the Internet, making already sluggish Internet speeds even slower.
By the time cell networks entered the LTE period, community structure was largely delayered and Internet entry had been simplified right into a two-hop system. To entry Internet providers, a person may join straight from the bottom station to the core community gateway, enormously bettering expertise.
However, we‘re now within the 5G period of ultra-low latency, ultra-high bandwidth, and large-scale IoT service necessities from completely different industries, necessities that present community structure can’t meet.
Low latency on the community edge
Since cell networks first supported information providers, every new era of cell know-how has aimed to enhance community throughput to improve person expertise. As throughput has elevated, Internet connection pace is now not a bottleneck for expertise, with latency turning into the important thing issue affecting person expertise.
Research on the applying aspect exhibits that, for prime quality, video providers have extraordinarily excessive latency necessities. For instance, augmented actuality (AR) and digital actuality (VR) video providers require latency of under 20 ms, with latency any increased for the latter inflicting VR illness. A service goal tied in with 5G is E2E latency of 1 ms to help the calls for of providers similar to Internet of Vehicles (IoV) and industrial management.
However, present cell know-how can’t reduce latency. With LTE, for instance, air interface throughput is elevated tenfold, however E2E latency is improved solely threefold due to sub-optimal community structure.
Although LTE networks possess two-hop flat structure, the bottom station and the core community are sometimes separated by a whole lot of kilometers, with a number of aggregation and forwarding units deployed in between. Add unpredictable congestion and jitter to the combo, and it’s not possible to guarantee ultra-low latency on LTE networks.
To help providers with very strict latency necessities, it’s obligatory to transfer community capabilities and repair processing capabilities to the sting of the community, closer to the person. This reduces the variety of intermediate layers, enabling low-latency service processing.
Content localization for ultra-high site visitors
Current cell networks had been designed for voice providers. Network structure adopts a tree construction, the place providers are converged at central nodes to be processed. This permits for frequent service concurrency and transmission effectivity.
But, with MBB providers, this type of community structure can encounter issues. For instance,1,000 customers watching a ten MB viral video would produce 10 GB of community site visitors as a result of the content material is re-transmitted from the Internet to the cell community 1,000 occasions. Therefore, 99.9 % of the community bandwidth used is wasted. So, caching the video in an edge node shut to the entry level would save the provider an enormous quantity of transmission bandwidth.
As air interface throughput is enormously elevated and community site visitors continues to develop, network-side inefficiency will improve, making content material localization important.
What verticals want
Mobile broadband networks have gotten the fundamental platforms for workplace work and advertising, with corporations anticipating personalized networks to meet their particular wants. Some, for instance, need information entry to the non-public cloud to be accomplished inside their enterprise campus networks to guarantee cell workplace safety, which means that community capabilities should be deployed on the campus community and help native breakout. Another instance is the concept of good billboards proposed by a Korean operator for good crowd evaluation primarily based on community information – the native content material modifications to improve the promoting conversion price. This sort of service requires a localized, open community platform.
5G edge cloud with MEC@CloudEdge
In Huawei’s 5G-oriented MEC resolution, MEC@CloudEdge, functions, content material, and a few MBB core community service processing and useful resource scheduling capabilities are deployed on the community edge, closer to the entry level. This allows each providers to be processed closer to the person and in addition utility, content material, and community orchestration. As a end result, customers are supplied with dependable, final service experiences.
MEC@CloudEdge: Deployment eventualities
Experience and effectivity should be balanced for deployment. Deploying the answer closer to the bottom station aspect requires fewer middleman nodes, which improves expertise. But, it additionally means there can be fewer customers accessing the node, lowering utilization.
Deployment location wants to think about scenario-based service necessities. For instance, some enterprises need non-public cloud information entry to be accomplished inside their enterprise campus networks, requiring the answer to be deployed inside their campus community. For sports activities stadiums that present dwell playback, interplay providers, on-line buying, and site providers for match-goers, the answer would wish to be deployed within the stadium.
Based on service necessities and useful resource effectivity, these calls for imply that MEC@CloudEdge ought to be deployed between the sting of the metropolitan space community (MAN) and the bottom station. This would possibly embrace central workplaces (CO) and numerous particular venues and campuses.
MEC@CloudEdge: Main capabilities
To meet service necessities, MEC@CloudEdge presents the next fundamental capabilities:
End of the person aircraft: runs core community capabilities similar to billing and coverage, and meets necessities for billing, monitoring, mobility, and O&M.
Local Breakout (LBO) for providers: makes use of LBO performance to help localized content material and utility processing, so the person can instantly entry native content material through MEC with out routing to the central core community gateway.
Third-party utility registration and administration: helps third-party utility integration, together with registration, discovery, and unified administration on duties like useful resource scheduling and well being checks. This permits community capabilities to be expanded and third-party providers to be personalized, in addition to service site visitors and content material consciousness and orchestration, so providers could be optimized to enhance expertise.
Network operate sharing: supplies an open platform for providers to be personalized for verticals and allows versatile third-party service deployment. With an open platform, inside community capabilities are shared, making seamless integration with third-party providers doable.
MEC@CloudEdge: Solution construction
MEC@CloudEdge relies on Huawei’s CloudEdge platform. It could be deployed on the identical {hardware}, with third-party functions primarily based on common servers. The resolution can share community capabilities to help new service innovation for operators, and allow on-demand modular deployment on the management, person, enabling, and administration planes.
MEC@CloudEdge: Key applied sciences
MEC tech used to be constrained by platform sharing applied sciences and industrial components earlier than precise deployment. But now, the maturity and utility of NFV know-how has prompted the trade to develop requirements for MEC and perform R&D on MEC options.
MEC@CloudEdge’s fundamental enabling applied sciences embrace:
NFV and cloud know-how to help multi-tenancy: NFV know-how helps a number of tenants on the MEC@CloudEdge. It additionally allows co-platform deployment and unified useful resource administration on gateway capabilities and third-party functions. However, because the deployment location is decrease utilizing NFV, node capability is comparatively small. This means utilizing NFV alone can’t make sure the reliability and efficiency offered by large-scale DCs.
Therefore, it’s obligatory to introduce cloudified software program structure on prime of NFV, so software program capabilities could be decoupled and deployed as layers primarily based on the properties of the completely different capabilities. This allows excessive reliability, excessive flexibility, and excessive efficiency when sources are restricted.
CU separation for versatile deployment of community capabilities: When MEC@CloudEdge is deployed shut to the entry level, core community gateway capabilities are distributed on the fringe of the community. This creates a considerable amount of interface configuration, connection, and commissioning. Thus, the core community’s management aircraft and person aircraft should be separated to enable the gateways to be flexibly deployed and to simplify the community.
Separating the Gateway CU strips away the complicated management logic of the gateway. Gateway capabilities are retained on the normal central gateway or built-in into an built-in management aircraft, reducing deployment prices and resolving issues with circuitous signaling routes and the burden on the interface.
Gateway CU separation first requires making the capabilities light-weight and stripping away complicated management logic capabilities. Next, the fundamental core capabilities which are retained are modeled, and the frequent forwarding aircraft fashions and object-oriented interfaces outlined. The forwarding aircraft is then programmable with glorious scalability. Once the complicated service capabilities are stripped away, light-weight configuration is carried out to help one-click deployment.
Service consciousness and good orchestration: To improve person expertise, MEC@CloudEdge helps real-time service consciousness and clever expertise optimization. When, for instance, wi-fi community high quality degrades, MEC@CloudEdge can sense the person’s video bit price and ship the site visitors to a video optimization module, the place the video stream is re-encoded, guaranteeing easy video for the person.
As MEC is deployed on cell networks, numerous thrilling new capabilities for work and leisure will emerge. These embrace MCDN-based HD video experiences; AR/VR-based cell video games; LBS advertising; and safe, dependable LBO-based enterprise cell workplace; good stadiums primarily based on MCDN and open platforms; and site visitors help programs for IoV. These new applied sciences will rework life in methods we can’t but think about.
