By Critz Chan
Multi-access Edge Computing (MEC) is a cloud-networking distributed architecture that combines real-time low latency access with edge servers to better handle latency-dependent applications. As almost everything moves to the cloud, from office and smartphone applications to entire platforms, performance and security challenges are growing along with the huge increase in the number of migrated applications. In the APAC region, demand is growing rapidly for cloud computing and edge computing solutions that can accommodate vast volumes of data and meet low latency requirements. MEC is designed to serve mobile, Wi-Fi, and fixed users simultaneously, connecting them via dedicated interfaces. MEC extends the power of the cloud to where people and services are connected, with requirements varying by country and region. One of its key advantages is to push applications, data transfer, and services to devices at the edge of networks in order to reduce latency, improve user experience, and preserve transmission resources at the core.
In a recent Techwire Asia survey, nearly 97 percent of respondents agreed that edge computing will be relevant in their business and IT strategy in the mid-to-long run. In APAC, there are several important pilot deployments, even though edge computing is, in some places, still a vague concept, with more education needed to demonstrate its relevance in the short term and demonstrate how it can revolutionize business operations.
Our Asia-Pacific customers are seeing the following as the key applications and business needs driving their interest in edge computing:
- Mission-critical real-time apps, such as biometrics, facial recognition analytics, and artificial intelligence
- Improving the QoS and QoE (user experience) in interactive gaming and in augmented and virtual reality
- Edge security to allow for applications to prevent cyber-attacks by moving the security perimeter closer to the source
- Cost reduction to save expensive backhaul investment
- IoT, particularly in the areas of smart cities, data capture, and real-time analytics
- The range of SDN/NVF virtual functions, including vBRAS, vBNG, and vRouter
As mentioned, in APAC, MEC has been moving beyond small scale and lab test environments to early deployments in the past few years. This is particularly true where providers are moving time-sensitive and secure applications from the cloud to the edge for faster response time.
For example, China Telecom has completed a smart parking project based on MEC and narrowband IoT (NB-IoT). China Unicom deployed a virtual edge-cloud testbed in Tianjin City for video optimization and security monitoring. Their objective is to provide network security and allow video data from surveillance cameras to be executed with an AI interface at the edge for faster local decisions and real-time response.
In Taiwan, the city government of Taoyuan has leveraged MEC for a pilot smart streetlight system, with IoT sensors collecting traffic data and road construction information and allowing the system to make decisions and take actions without time delay. Another example is in a Taiwanese manufacturing operation where critical data is shared and insight from the edge is utilized across multiple locations thanks to data correlation, deep learning, and process coordination.
In Japan, NTT DoCoMo commenced a proof-of-concept that will enable video analytics from surveillance cameras using MEC.
In other parts of Asia, such as South Korea and Southeast Asia, we are seeing pilot projects that involve routing solutions with field-programmable capabilities and data offloading. This includes virtual routers and virtual BRAS solutions. Recently, SKT signed an MoU with Deutsche Telekom to introduce MEC to 5G in the Korea market.
The ultimate objective of MEC is to provide an optimized and low-latency computing infrastructure with deployment agility that can scale on demand. With MEC, network services and data are moved closer to end users for much better QoE and QoS, while reducing network core congestion and optimizing gateway interconnectivity costs.
MEC requirements will surely continue to gain strong traction in APAC, amid the skyrocketing demands for high bandwidth, faster speeds, and low latency. Following the rollouts of 5G networks, the requirements for MEC will become more obvious for time-critical applications. Nations leading the way in MEC adoption, such as China, Japan, India, and South Korea, have strong government initiatives and supports to encourage an advanced network infrastructure. This is particularly true where those countries have ambitious plans for smart city and smart country initiatives.
As one analyst organization noted, APAC is expected to grow the fastest in the virtual routers market from now until 2023. Major growth drivers include the increasing complexities in network infrastructure, desire for network optimization, and an urgency to deliver an enhanced customer experience. Currently, the region accounts for about 30 percent of the vRouter market, compared with 40 percent in North America and 25 percent in Europe. Just as 5G will make our businesses and lives smarter and more connected, MEC will enhance that by empowering time-critical and information-sensitive applications on 5G. As service providers continue to pilot launch 5G services, MEC will be riding along these test deployments and demonstrating its potential.
Ethernity Networks is engaged with MEC initiatives worldwide, and in Asia we are working with Tier 1 operators in China and Korea that are testing – in various stages – their MEC solutions.
A critical factor in enabling MEC to live up to its vast potential is the challenge of reducing power and equipment footprint, as these are much greater issues in edge sites than in a typical data center. That is where the ability to maximize data processing capacity, as Ethernity does in our FPGA-based SmartNICs, comes into play. These SmartNICs can slot into existing commercial off-the-shelf (COTS) servers, with multiple NICs residing in a single server as needed, reducing the requirement for additional boxes and saving space and power. FPGA-based SmartNICs also address the security requirement of the edge, enabling network isolation and user segregation to prevent attacks on edge sites and user devices, as well as offering IPSec VPNs and tunnel termination.
