3GPP agrees plan for first release of 5G specifications, including the New Radio access technology


The latest plenary meeting of the 3GPP Technical Specifications Groups (TSG#72) in Busan, Korea, has agreed on a detailed work plan for Release-15, the first release of 5G specifications. The plan includes a set of intermediate tasks and check-points to guide the ongoing studies in the various Working Groups. Key dates to watch out for:

  • September 2016: completion of Technical Report on New Radio (NR) requirements
  • December 2016: start of SA2 normative work on Next Generation (NexGen) architecture
  • March 2017: the beginning of the RAN Working Group’s specification of the 5G NR.

The 3GPP also made progress during the Busan meeting on defining the new radio (NR) access standard, agreeing the target scope for its eventual inclusion in Release 15. It has been decided that the NR must work in both standalone and non-standalone configurations – and by non-standalone they mean using LTE as the control plane anchor. Standalone mode implies full control plane capability for the NR. This is made possible because of the agreement to separate the control plane from the data plane, which is a familiar concept if you have been following TelecomTV’s coverage of NFV and next generation architectures.

There’s a good overview of all the possible permutations here (zipped PPT), courtesy of Deutsche Telekom, which shows how you can arrange two radio technologies (LTE and NR) and two core network concepts (the current EPC and a Next Generation Core Network) along with the radio level aggregation between them. In short, 11 new scenarios are possible. The question, though, is whether or not 3GPP needs to support all these architecture options, and how to ensure forward compatibility in order to avoid global fragmentation. At the moment, it would appear that the 11 might get reduced to 6 or 7, which is still a high number and therefore added complexity.

The 3GPP also agreed two target use cases: Enhanced Mobile Broadband (eMBB), as well as Low Latency and High Reliability to enable some Ultra-Reliable and Low Latency Communications (URLCC) use cases. We are not sure what happened to mMTC (massive Machine Type Communications), but it’s there in the current Technical Report draft. There was also agreement to use 6GHz as a spectrum key point, enabling study work to be conducted at frequencies below 6GHz and additional work above 6GHz.

During the discussion at the TSG meeting, the importance of forward compatibility – in both radio and protocol design – was stressed, as this will be key for phasing-in the necessary features, enabling all identified use cases, in subsequent releases of the 5G specification.

“We now have a more concrete plan to guide the studies in the Working Groups and to put us in the position to address both short term and long term opportunities of 5G,” said Dino Flore, Chairman of 3GPP TSG RAN.

The latest agreed requirements and deployment scenarios for NR can be found in this document, but be aware that these draft publications get updated on a very frequent basis.

However, the latest document does succinctly lay out the requirements for the NR, as follows:

  • tight interworking between the new Radio Access Technology (RAT) and LTE
  • at least dual connectivity between LTE and new RAT, to support high performing inter-RAT mobility and aggregation of data flows
  • multiple transmission points, either collocated or non-collocated
  • separation of control plane signalling and user plane data from different sites (C-plane/U-plane separation)
  • inter-site scheduling coordination
  • different options and flexibility for splitting the RAN architecture
  • deployment flexibility (e.g. to host relevant RAN, Core Network (CN) and application functions close together at the edges of the network, when needed, for example to enable context aware service delivery and low latency services
  • allow deployments using Network Function Virtualisation
  • allow for the RAN and the CN to evolve independently
  • allow for the operation of Network Slicing
  • support sharing of the RAN between multiple operators
  • allow the deployment of new services rapidly and efficiently
  • allow the support of 3GPP defined service classes (e.g. interactive, background, streaming and conversational)
  • enable lower CAPEX/OPEX with respect to current networks to achieve the same level of services
  • RAN-CN interfaces and RAN internal interfaces shall be open for multi-vendor interoperability
  • support operator-controlled sidelink (device-to-device) operation, both in coverage and out of coverage

That’s a long list, but necessary if all the envisaged new use cases can be enabled and supported by operators. A number of deployment scenarios have also already been identified, including various iterations of rural, urban and high-speed links for trains (no-one wants another fragmented rail system, like GSM-R). Interestingly, 3GPP has already conducted some evaluation work on support for connected cars, looking at two scenarios – highway scenario, for vehicles travelling along roads at high speeds, and an “urban grid” for densely deployed vehicles.

Other interesting points being considered (note, all of these are “works in progress” and subject to change) include:

  • confirmation that peak (theoretical maximum) data rate for downlink is 20Gbit/s and 10Gbit/s for uplink,
  • maximum aggregated system bandwidth is likely to be an ITU decision, not 3GPP,
  • control plane latency of 10ms,
  • user plane latency of 0.5ms for uplink and the same for downlink for URLLC cases,
  • user plane latency of 4ms for uplink and the same for downlink for eMBB cases,
  • mobility interruption time should be 0ms,
  • there is uncertainty about the need for voice interoperability with LTE,
  • maximum cell range without KPI degradation should be 100km,
  • target battery life for mMTC (IoT) devices of 15 years,
  • maximum connection density of 1 million devices per square kilometre,
  • support for connecting to a user travelling at a maximum speed of 500km/h.

The standardisation process might appear frighteningly complex to outsiders, but the 3GPP has been through this several times in the past and has tried and tested procedures in place.

“3GPP continues to actively coordinate radio access NR and Next Generation system level work to standardize target services on schedule,” confirmed Erik Guttman, Chairman of 3GPP TSG SA.