In recent years, the traffic of telecom operators has increased very rapidly, but in contrast, the income of telecom operators is still growing at a low rate, and the scissors gap between the two is getting bigger and bigger. With the advent of 5G and the development of cloud computing, IoT, AI, VR/AR and smart driving, digital transformation of the entire industry has entered a new stage. The increasing real-time computing capability not only raises ultra-low latency and higher bandwidth requirements for telecom networks, but also poses new challenges to the current cloud computing model. There is also an urgent need for integration of information and communication technology (ICT) and operational technology (OT).
Edge computing is the support and enabling technology for integration of ICT and OT, and the industry will enter an important period of opportunities. Edge computing is an inevitable choice for operators to expand the application value of industrial internet.
Introduction of MEC
As networks continue to expand, more and more connections need to be analyzed, processed, and stored on the edge side. With the continuous improvement of terminal capability and the further reduction of traffic tariff, high-traffic services will have a direct pulling effect on the average data traffic per user per month (DOU). It is estimated that when 5G is put into large-scale commercial use, the average traffic bandwidth of a single user will reach 5 to 10 times that of the 4G network, putting great pressure on the backhaul network.
The current network architecture and mobile technologies fail to fully optimize the network. The distance between base stations and the core network is often hundreds of kilometers, through multiple convergence and forwarding devices, coupled with unpredictable congestion and jitter. It is difficult to guarantee the scenarios of industry users with high requirements for delay and reliability.
Edge computing is a distributed computing architecture that processes data physically close to the data source. It can further reduce transmission delay, increase network efficiency, improve service distribution and transmission capabilities, and enhance user experience, thus meeting the key needs of industry users in the process of digital transformation, such as real-timeliness, intelligence, data aggregation and interoperability, security and privacy protection. This is highly consistent with the concept of being service-oriented and expanding 5G into vertical industries. Edge computing, as a native 5G function, will help to implement application localization, content distribution, and edged-based computing.
Different organizations give different definitions of edge computing. As defined by ETSI, multi-access edge computing (MEC) is a product of ICT integration and also a key technology to help operators transform their operations. Tens of thousands of edge DCs are excellent resources of operators over OTT, which makes MEC have a wide application space.
As an innovation in network architecture and business model, MEC can be a sharp tool that uses the network to serve vertical industries. It can also be an opportunity for operators to enhance network value and will probably promote the redistribution of industry chain values.
MEC Empowers Network Architecture
MEC will change the state of infrastructure in almost all enterprises, from the highly centralized to a mixture of centralized and distributed, that is, the combination of cloud computing and edge computing. With the emergence of MEC, operators have changed their pipes into intelligent ones. MEC can not only optimize pipes and services but also assist enterprises in interworking with their applications, thus solving the problem of dealing with massive connections and large data traffic. Moreover, MEC can be integrated with the field operation technology to achieve the integration of the whole OICT, that is, with the ICT-based operation technology, operators can help enterprises complete digital operation and better serve vertical industries.
The MEC-based future network architecture allows computing, storage, and network functions to be moved down to the pipes of the network domain, and can be deployed in the desired locations according to the service requirements of industrial applications (Fig. 1). These MECs deployed in different locations form a MEC cloud that supports real-time computing while making existing network pipes intelligent. The MEC cloud can not only be deployed based on existing 4G LTE networks, but also evolve seamlessly to support 5G and enable three major 5G application scenarios: eMBB, mMTC and uRLLC.
In the era of cloud computing, internet terminals are connected to the cloud through network pipes for services, which is a typical C/S computing mode. When MEC is introduced, the edge computing layer is available in the network pipes. The original C/S computing mode is transformed into the C/E/S computing mode. The MEC-enabled network architecture based on the SOA architecture, micro-service model, and cloud native applications can meet the digital transformation of enterprises.
MEC Helps Operator Transform
Because of its proximity, large bandwidth, low latency, and localized computing, MEC brings the following typical advantages:
—Through efficient computing at the network edge, the backhaul traffic of the cellular network can be reduced greatly and the transmission cost can be reduced.
—For delay-sensitive applications, edge computing is implemented to reduce the delay and improve the quality of experience (QoE).
—Edge data will become more private (personal health, facial or voice recognition data, private place interaction) or confidential (key factory internal data). Edge computing meets privacy regulatory needs by processing, storing, and/or discarding appropriate data, reducing the risk of data leakage.
Through low-cost edge solutions, telecom operators can interconnect or integrate and optimize enterprise infrastructure to facilitate digital transformation in vertical industries, thus increasing revenues. They can develop emerging B2B business models to provide edge cloud managed services for many third-party cloud service providers, thereby creating business opportunities and increasing revenues. Based on the MEC-enabled architecture, telecom operators achieve ICT integration. Also, they can successfully transform their operation mode from serving public customers with 2C services to serving vertical industry customers with 2B services in the mobile internet era.
MEC Based Transformation Road
ITU has defined three application scenarios for 5G: eMBB, mMTC and uRLLC. The vision and needs of 5G are to meet the explosive growth of mobile data traffic, massive device connections, and emerging new services and application scenarios in the future, while deeply integrating with the industry to meet diverse needs of terminal interconnection in vertical industries, achieve real internet of everything, and build the cornerstone for digital transformation of the social economy.
As 4G networks still serve as basic coverage networks for a long time, operators can transform MEC on existing 4G networks and continue to operate it after building 5G networks in the future. MEC can be applicable to both 4G and 5G networks. It enables telecom operators to make full use of the fully-covered 4G LTE networks and the upcoming 5G networks to achieve integration of mobile communication network and industrial internet services, enhance the added value of networks, and facilitate operation transformation.
From a business perspective, operators can transform from 2C and 2B services respectively. The transformation of 2C services can be based on existing 4G networks, and the QoS management and control capability can be improved through MEC in addition to original package services. This scheme improves user experience and thus increases the revenue. Transformation of 2B vertical services that require very low delay can be provided based on 5G networks. In a scenario not requiring low latency requirements, existing 4G networks can be used with MEC to provide office and production operations in an enterprise campus. As there are few uplink services on the 4G networks, 4G video services can also solve the problem of difficult video surveillance deployment.
Flexible security services are required. Because MEC is deployed inside network pipes and needs to provide localized computing and coordination with cloud computing, the privacy and security of enterprise data need to be considered. Based on the isolation implemented through virtualization, data security and security collaboration with cloud-based and on-site edge computing need to be further considered.
In terms of technology evolution, MEC transformation needs to consider plug-and-play infrastructure, flexible and reliable MEC platform, cloud native applications, automatic service orchestration, and building open MEC application capabilities.
The deployment of MEC is based on service requirements and scenarios, combined with network requirements, edge infrastructure, operation model, and maintenance and management needs, which is a balance between performance and investment. A strong ecosystem is the guarantee for MEC development. Perfect infrastructure, flexible network and platform capabilities, and abundant edge applications are the key to prosperity of the edge ecology.