SoftBank case study
Rapid small cell deployment in the urban jungle
Urban small cells deployed at street level are the next logical step to meet growing data traffic demand in city centres. Practical solutions need to be quick and easy to install, adapt seamlessly with tactical evolution and be resilient during outages. SoftBank installed CCS Metnet self-organising microwave backhaul in the challenging metropolis of Tokyo, delivering valuable insight into the behaviour of 5G microwave.
Planning deployment of urban small cells involves trade-offs between three key capabilities – backhaul, power and site availability.
The site preferred by RF planners may not have electrical power; fibre backhaul may be unavailable and/or have long lead times to install; some sites may be much more costly or inaccessible than others. Flexibility is critical for success.
In the SoftBank implementation, CCS Metnet simplified the planner’s dilemma by providing high capacity, highly resilient wireless backhaul which was able to be quickly deployed in the rapidly changing environment of Tokyo.
Operating in the 26GHz band that is normally constrained to Line-of-Sight, this live trials in downtown Tokyo has proven NLoS (Non-Line-of-Sight) as part of the mix.
Seven nodes were installed on existing telegraph poles, with two directly connected by Gigabit fibre backhaul.
The simplest links were direct line-of-sight along the street, which coped with reflections from buildings on both sides. The most remote node had three wireless links (hops) to reach fibre backhaul.
The topology of the deployment shown below is a live screenshot from the Management System.
No skilled tuning or calibration was required for these links – Metnet automatically senses, filters and adapts to optimise performance and minimise interference. This is a continuous on-going process, adapting to environmental changes such as traffic movement, weather and foliage growth.
All nodes use identical hardware with a single radio per location. Single RF channel operation does not require any RF planning, speeding and simplifying installation.
The node shown on the right is a typical street alley deployment with its neighbour node further along the canyon.
More unusual were NLoS (Non-Line-of-Sight) links such as one shown below, where Metnet resolved and adapted to multi-path reflections from nearby buildings. The 270 degree wide angle 16 antenna array built into each Metnet node captures a wide range of signals for extensive processing and filtering. This is much easier and more flexible than Point-to-Point radio links which require alignment, where instead Metnet nodes with their 270degree antenna array automatically detect all possible LoS and NLoS links and select them accordingly.
Dense environments such as this are likely to have small cells spaced closer than 50 meters apart. One link was 32 metres measured as a direct path, with NLoS reflected signals travelling 49 metres. The adjacent nodes had 37 different possible RF links with varying attenuation levels, for which the best options are automatically calculated and continuously reviewed. This demonstrated the high degree of multipath reflection that occurs in a typical street level urban canyon. All potential Metnet links are shown in the diagram below.
This link achieved an average signal-to-noise ratio/modulation of 12dB, supporting backhaul data throughput of 135Mbps. The link achieved 100% availability throughout the trial period. Round trip latency ranged from 0.5ms (1 hop) to 0.88ms (3 hops). The latest dual channel Metnet radios are capable of delivering 1.2Gbps using 256QAM in a typical 112MHz single channel pair.
Operating in a single channel pair without the need to radio plan, the SON algorithm select the best possible links and topology, and runs a dynamic spatial TDMA schedule which dictates exactly which node and antenna pair can transmit and receive so as not to cause interference in the cluster. This is made possible by the thousands of measurements which are taken every second to determine the real time state of the mesh. This cognitive approach to microwave networking provides a robust and flexible platform in the most challenging or urban environments.
Metnet addresses many of the issues of urban small cell deployment, simplifying, adapting and ensuring resilient high capacity backhaul can become available to all.
Our customer’s experience with Metnet
“A live deployment in downtown Tokyo is the ultimate challenge for a small cell backhaul system. Our engineering team was impressed with Metnet’s unique MPtMP mesh architecture and approach, and we wanted to test how this self-organising, self-optimising, and self-healing microwave solution performed in a dense urban area.
We designed the trial principally to assess Metnet’s automatic support for NLOS, and its ability to cope with multiple RF paths in a multipath propagation environment – all while delivering optimal performance and quality of service. Metnet’s capabilities fulfilled our expectations. The trial has provided a valuable insight into the behaviour of microwave and mmWave systems in future 5G bands, and the results of our trial demonstrate that Metnet can be well positioned to support these fifth generation networks.”
SoftBank Corp.Technology Strategy Office2