Main Conference
Session KNWE-S1
Keynote
Conference
9:10 AM
—
9:50 AM KST
Local
May 26 Tue, 5:10 PM
—
5:50 PM PDT
6G - A Step Beyond Stretching 5G
Prof. Gerhard Fettweis (Technische Universität Dresden)
18
The 5G services require a network with “high bandwidth and ultra-low latency.” High bandwidth can be enabled by wider frequency bands. To achieve ultra-low latency, however, network operators have come up with the concept of “mobile edge.” By leveraging mobile edge, we can deliver novel 5G applications that can benefit from sub 10msec latency, such as cloud XR, cloud gaming, connected cars, cloud robots.
While providing ultra-low latency itself is useful, this does not fully justify the cost of deployment of numerous edge sites. In fact, it is not difficult to see mobile edge provides a couple of additional benefits: (1) huge volumes of data (that may be generated by, for example, connected cars) can be processed at the edge instead of sending them to a remote data center, which is extremely costly; (2) mission-critical and sensitive data from a smart factory or hospital can be processed at the edge without leaving the site. By enabling edge data processing and local security, mobile edge provides a unique opportunity for mobile service providers to bring new values to its B2C and B2B customers.
In this keynote, I propose that mobile edge should be “programmable” and “cloud native.” This does not mean just running a few VMs at the edge site. At SKT we are developing its mobile edge as a fully functioning cloud. SKT’s MEC or “mobile edge cloud” will provide virtualized infrastructure with Kubernetes, serverless, and service mesh support. We are also pairing our MEC with public clouds so that our users have options to quickly build new applications using widely understood cloud APIs and services. In addition, we will provide our unique service assets, such as telco APIs, natural language processing, real-time data processing, etc. “as a service” to developers so they can quickly build something that was truly not possible before.
I will conclude this talk by presenting several early use cases that we are developing on 5G MEC with our partners.
While providing ultra-low latency itself is useful, this does not fully justify the cost of deployment of numerous edge sites. In fact, it is not difficult to see mobile edge provides a couple of additional benefits: (1) huge volumes of data (that may be generated by, for example, connected cars) can be processed at the edge instead of sending them to a remote data center, which is extremely costly; (2) mission-critical and sensitive data from a smart factory or hospital can be processed at the edge without leaving the site. By enabling edge data processing and local security, mobile edge provides a unique opportunity for mobile service providers to bring new values to its B2C and B2B customers.
In this keynote, I propose that mobile edge should be “programmable” and “cloud native.” This does not mean just running a few VMs at the edge site. At SKT we are developing its mobile edge as a fully functioning cloud. SKT’s MEC or “mobile edge cloud” will provide virtualized infrastructure with Kubernetes, serverless, and service mesh support. We are also pairing our MEC with public clouds so that our users have options to quickly build new applications using widely understood cloud APIs and services. In addition, we will provide our unique service assets, such as telco APIs, natural language processing, real-time data processing, etc. “as a service” to developers so they can quickly build something that was truly not possible before.
I will conclude this talk by presenting several early use cases that we are developing on 5G MEC with our partners.
Session Chair
Song Chong (KAIST, Korea (South))
Session KNWE-S2
Keynote
Conference
9:50 AM
—
10:30 AM KST
Local
May 26 Tue, 5:50 PM
—
6:30 PM PDT
Into the Future Wireless
Dr. Jianmin Lu (Huawei)
12
The 5G has already been commercially deployed since last year and people, especially Korean customers, enjoy the benefit of 2C business. Although the standard of 3GPP R16 will be released shortly in 2020 and this will be the full formal 5G, i.e. IMT2020 compliant, the research and standard to evolve 5G will not stop. While 5G is opening the door of digital transformation of many aspects of our life, industry, business and even the whole society, the future of wireless is yet to be discovered. Since the first generation of mobile technology, the mobile industry has experienced significant growth driven by ‘subscription dividend’ and ‘traffic dividend’. The next dividend is believed to be the “connection dividend” or even “intelligence dividend”. In addition, sensing (including accurate positioning, imaging etc.) will be a novel capability of future network, enabling “everything sensing”, “everything connected” and “everything intelligent”. On the other hand, the sheer number of connected devices and objects will not only create unprecedented growth of data traffic and massive connections, but also create a substantial increase in energy consumption across all parts of the network. Energy efficiency in wireless networks is now a growing concern for network operators to not only reduce the network operation costs, but also as a social obligation, to reduce greenhouse gas emissions. Moreover, higher frequency (mmWave and THz) band provides abundant spectrum for Tbps data rate, while it also brings about paradigm shift in the whole system design. The research challenges and technology breakthroughs required to deliver the vision for future wireless will be presented in this talk.
Session Chair
Song Chong (KAIST, Korea (South))