Five 5G Takeaways from the Brooklyn 5G Summit

The Brooklyn 5G Summit (B5GS) is a Nokia Bell Labs/NYU event that many from academia, industry, and the analyst community are invited to attend to discuss topics relevant to 5G. Now in its fifth year the B5GS anticipates imminent 5G takeoff and the Nokia 5G FutureX architecture being brought to life. In previous years, the B5GS focused on important 5G topics including Millimeter Wave (mmWave) channel modelling, high 10 Gbps throughput, Massive MIMO (Multiple Input, Multiple Output), and 5G use cases.

The many sessions and panels left us with a set of key takeaways on 5G deployment differences we should expect between different operators, and the performance to expect. We also gained insights into the challenges of deploying mmWave, Radio Access Network (RAN) evolution and virtualization, and commercial phased arrays.

During the two-day event, we heard from AT&T, KT, NTT DoCoMo, T-Mobile, and Verizon from the Mobile Network Operator (MNO) community, and we heard from various equipment and semiconductor vendors, including Ericsson, Huawei, Nokia, Intel, and Qualcomm.

Each of the MNOs outlined their plans for 5G, with KT highlighting its 5G trial experience from the recent Pyeongchang Winter Olympics. Approaches to 5G deployment vary from 5G underpinned by “5G Evolution” (AT&T), low-band support for coverage, mid-band for metro, and high-band for dense urban scenarios (T-Mobile). Verizon outlined its plans to benefit from first mover advantage and that 5G Technical Forum (5GTF) (pre-5G standard for Fixed Wireless Access [FWA]) is just one network slice with true mobile 5G being another. The company plans commercial FWA in 2018 and intends to quickly move from 5GTF to 3rd Generation Partnership Project (3GPP) standards.

The equipment vendors outlined trial and simulation results, with most expressing the opinion that 5G NR will comply with IMT-2020 standards. On mobile 5G in the mmWave bands, the consensus was that this is technically possible but may require special considerations for RAN deployment depending on the deployment scenario. Interestingly, Nokia talked about its plans to open Application Programming Interfaces (APIs) in the RAN (xRAN), management and orchestration (ONAP), and Artificial Intelligence (AI) to enable a “white box” networking equipment business.

Intel and Qualcomm both talked about their planned use of phased arrays in handsets (Qualcomm) and in automotive applications (Intel).

5G networks will be operator-specific as each operator deploys according to its spectrum assets in low, mid, and high band and configures its network as non-standalone or standalone. Passive infrastructure assets such as deep fiber for backhaul and fronthaul will also dictate network topology. 5G will build on a foundation of 4G/LTE to address each operator’s unique business models and value propositions.

Early 5G deployments will not be much faster than LTE Advanced Pro, at least according to T-Mobile, which will deploy for nationwide coverage in 600 MHz. However, throughput will increase over time as mid-band and high-band spectrum overlays the network, but these assets will be deployed in urban and dense urban hotspots.

High-band mmWave will more challenging to deploy than low- and mid-band due to its physics of propagation and the signal attenuation from common obstructions such as foliage and buildings. Verizon’s mmWave implementation leverages multipath and a knowledge of deployment topography for best node placement. Also, outdoor to indoor signal penetration will be difficult in the high bands and will require in-building wireless solutions, such as repeaters and small cells, according to KT. Ericsson also claims that outdoor-to-indoor coverage is possible but that it is challenging without a case-by-case targeted and tuned RAN.

RAN evolution and virtualization underpins 5G; however, there are many foundational and overlapping technologies for front haul that may form a headwind for full implementation. Nokia’s embrace of xRAN as part of its new open approach is just one example. Other initiatives (discussed in our recent foresightvRAN a Hot Topic at MWC18[IN-5064]) we are aware of are the vendor specific eCPRI, the NGFI IEEE1914.1 and 1914.3, the service provider centric ORAN, and Cisco’s Open VRAN. As MNOs look to mix and match select components from different vendors in their RANs to achieve the best performance, a truly open vendor-agnostic RAN is an ideal aspiration. In the real world, the hypercompetitive equipment vendors will likely strive to retain some form of vendor-specific hardware in the RAN to maintain their competitive differentiation.

Phased Arrays move out of the defense and radar verticals and into commercial mobile networks. No longer an exotic technology thanks to modern semiconductor and integrated circuit technology, phased arrays for massive MIMO and beam steering are an economic technology for commercial mobile applications either in the User Equipment (UE) or at the basestation. Much work remains to economically industrialize phased arrays and massive MIMO, with challenges remaining on power consumption, cost reduction, and complexity. ABI Research discussed massive MIMO and phased arrays as part of its The Rise of Massive MIMOreport (AN-2758).