Insight: Light Reading 5G Ecosystem Event Recap

Posted by Mae Kowalke on Saturday, April 18, 2015 with No comments

Accedian's VP of Solutions Marketing, Scott Sumner, attended Light Reading's "Building America's 5G Ecosystem" event this week in New York City, to keep a finger on the pulse of next-generation mobile communications development. Some of his observations from attending sessions at the event are summarized here. 

4.5G is the Road to 5G

During the "5G Future and How to Get There" session, Heavy Reading senior mobile analyst Gabriel Brown predicted--echoing Accedian's stance regarding the real-world reality of 5G--that the 4.5G network technologies like C-RAN, HetNets, various flavors of LTE (U/unlicensed, B, A, M/connectionless machine interface, D/direct device-to-device), software-defined networking (SDN), and network functions virtualization (NFV) will drive performance improvements and blur the gap with 5G objectives.

Those objectives themselves are, Brown noted, are already changing (we would add: predictably, before the ink is even dry on proposed standards) in response to lab tests. For example, following development teams' early attempts to trim latency, the goal of 1ms end-to-end is moving toward something more like 5ms--which would still be a significant challenge compared to today's networks, often clocking 200-300ms latency figures in real-world use.

Brown also explored two 5G-related concepts: "no-cell / no edge" system architecture, and network slicing. 

No-cell system architecture refers to devices communicating between multiple cells concurrently, allowing user traffic to flow directly between endpoints over multiple paths. Using this approach--also known as Coordinated Multi-Point (CoMP)--users are less likely to experience 'edge' issues, where performance degrades as they move towards the outer edge of a cell.

Network slicing (championed in NGMN's 5G white paper), refers to identifying different network embodiments for different use cases. These 'slices' would be controlled via software-defined networking SDN, incorporating specific network functions and services to meet demands of diverse applications such as Internet of Things (IoT) vs. ultra-low-latency (ULL) tactile internet. Performance assurance will need to address each slice with tailored monitoring, trending, and visualization. NFV is a key component of this vision, as these 'slices'--basically virtual networks--are spun up across shared infrastructure.

Other observations and predictions from Brown:
  • Industrial internet will be a key revenue driver
  • Candidate radio access technologies (RATs) for 5G are still mostly in university labs (e.g. Lund University trials of 100 antenna MIMO)
  • The 2020 Tokyo Summer Olympics will be a showcase for vendors, if they are able to get meaningful 5G solutions developed by then
  • Ultra-scalable use cases will target 200,000 users/km2 with 100 Mbps speeds
  • Most operators remain more interested in sub-6GHz spectrum bands than cm/mm wave due to penetration issues with upper bands, and availability of allocated, licensed spectrum.

What is 5G All About, Anyway?

During the "Vision for 5G" session, Qualcomm senior director of technical marketing Rasmus Hellberg explored the applications that distinguish 5G from 4G. 

One main application 4G does not address, according to Hellberg: mission-critical control and monitoring applications like healthcare and emergency response. He predicted this will be realized using CoMP connectivity, ensuring higher reliability through UE access path diversity. This could be achieved with LTE-A, but the industry sees CoMP integration with SDN and C-RAN (adding a layer of software-based resilience) as a 5G realization. 

Hellberg noted that 5G has been positioned in a very application-centric way because it's generally accepted that the many use cases proposed have competing performance demands that can't all be realized concurrently. For example, low cost, pervasive IoT connectivity and ULL networking are not generally compatible. Also, high bandwidth applications often do not require low latency; consider 8k video / 120Mbps (at 120fps). 

Other observations from Hellberg:

  • If IoT devices have very low power transceivers, they may have to create their own mesh networks to reach the RAN.
  • 'Awareness' is a 5G requirement throughout the network to improve reliability, security, and enhance scalability.

Can We Get There from Here?

During the "5G: Getting from There to Here" session, National Institute of Standards and Technology (NIST) chief scientist Kent Rochford examined the key spectrum and technical standard issues standing between current reality and 5G as it's envisioned. That journey, he said, will include addressing spectrum sharing, millimeter wave, massive MIMO, and public safety, among other topics. 

Rochford noted that:

  • Shared spectrum allows priority services to gain full access in critical situations. 
  • Upcoming LTE extensions include methods of aggregating more than three carriers, including unlicensed bands for LTE and Wi-Fi.
  • Scattered millimeter wave signals are showing the ability to carry bandwidth, and to some extent their limited reach could benefit small coverage areas with reduced wider-area interference.

5G Use Cases and Requirements

During the "NGMN 5G Whitepaper: Vision, Use Cases, Requirements & Architecture" session, NGMN Alliance General Manager Philipp Deibert explored some of the organization's recommendations and conclusions around 5G. Some of NGMN's targets for 5G include:
  • Smart office services that deliver 15 Tbps/kmtraffic density for office towers
  • Updated over-the-air (OTA) capabilities for adapting to network upgrades
  • Three-day smartphone battery life (our note: maybe graphite and aluminum batteries will essentially fulfill this need through much faster charging?)
  • Network-based positioning accuracy approaching <1m indoor or out (a new requirement vs. 4G)
  • Improvements to handset processors and radios, which currently contribute to latency

3GPP Roadmap for 5G

During the "3GPP Activities and 5G" session, NTT-DoCoMo research manager Satoshi Nagata looked at the 5G standardization timetable endorsed this March by the 3rd Generational Partnership Project (3GPP). 

Nagata said the most important technologies in the 5G development process are millimeter wave, massive MIMO, beamforming, and LTE-LAA (Licenced Assisted Access - really this means unlicensed). He explained that having LAA coexist with Wi-Fi on the same carrier for transmit/receive interference may be more spectrally efficient; a DoCoMo study showed a 60% improvement over 802.11ac. 

NTT plans to introduce 5G in 2020 using sub-6GHz bands, including LTE-LAA unlicensed. (Likely candidates: licensed 4.9GHz, unlicensed 5GHz; 100-200MHz bandwidth is only available for this frequency range in Japan.) 5G using 6GHz and higher bands will come later, due to constraints around upper frequency definition and licensing. 

A New Air Interface for 5G?
During the "Case for a New 5G Air Interface" sessions, Alcatel-Lucent wireless division CTO explored the technicalities of mobile air interfaces, defining and noting aspects of this technology that suggest a new air interface will be needed to achieve 5G goals. He started by defining millimeter "high band" wave as being 20GHz and higher; 6GHz-20GHz is cm wave.

5G will still require use of sub-6GHz band for control and continuity, he said, while high band on small cells ("ultra-narrowband") are a good fit together for massive capacity, extreme low latency and shorter coverage range.

Peters also predicted that IoT, currently considered a cellular backhaul function, will likely connect via home gateways using Bluetooth and Wi-Fi; this is a sensible and simple way to allow connectionless direct connections for IoT endpoints.

What's Your Take?
Did you attend the Light Reading 5G Ecosystem event this week, or follow coverage in the news feeds? What's your sense of how 5G visions and technology/market realities will converge as time goes on?

Categories: , , , , ,