Jun 26, 2018
Qualcomm products mentioned within this post are offered by Qualcomm Technologies, Inc. and/or its subsidiaries.
June has been a very exciting month in the 5G world. Not only did we see the approval of a package of new projects that will expand 5G NR, in Release 16 and beyond, to new industries, but 3GPP also marked the completion of the 5G NR specifications for standalone (SA) mode. This milestone complements the non-standalone (NSA) specifications completed in December last year, and is significant, as it represents the final run towards 5G commercialization in 2019.
The SA specifications — in addition to supporting independent deployment of 5G NR with the new 5G core network — aim to enable new end-to-end network features, from network slicing to more granular Quality of Service (QoS) support. These network features are essential for enabling new business models.
With the SA deployment option for 5G NR, operators, mobile device manufacturers, and app developers are eager to get answers to a new set of questions:
Today, we are showcasing the next release of our 5G NR Network Capacity and User Experience Simulation in the Qualcomm booth at Mobile World Congress Shanghai. The simulation study conducted in Tokyo builds on top of the original platform that was released just four months ago, and it helps to answer these questions and deliver quantitative insights to the expected real-world performance and user experiences of 5G and Gigabit LTE devices, operating in the new SA deployment mode.
Let’s take a closer look at the results:
Unlike our simulations for Frankfurt and San Francisco, Tokyo modeled a SA 5G NR network (Figure 1, below) using 20 existing macro cell base stations with the new 5G NR cell sites co-located with existing LTE cell sites. The Tokyo 5G NR network operates on 100 MHz of 3.5 GHz spectrum, with an underlying Gigabit LTE TDD network operating across three LTE spectrum bands (3x20MHz). The propagation between the base stations and the devices was modeled based on high-definition 3D maps of Tokyo to account for path loss, shadowing, diffraction, building penetration loss, interference, and more.
In addition, RF capabilities were modeled to accurately depict real-world performance, such as massive MIMO capability for 5G NR with 256 antenna elements and 4x4 MIMO on LTE TDD.
We applied diverse traffic models that simulates popular users experiences — browsing, downloading, and streaming — to a mixture of devices with different capabilities.
Over 12,000 active user devices, of various capabilities, were randomly distributed across the network with approximately 50 percent of the users indoor and 50 percent outdoor. The simulation study showcased a downlink capacity increase of nearly 5x when migrating from an LTE TDD only network, with a mix of LTE devices of various capabilities, to a SA 5G NR network with multi-mode 5G NR devices and an increased mix of advanced Gigabit LTE devices. Another compelling benefit is the 3x increase in median spectral efficiency.
The simulation platform also provides insight on the user experience through user-level key performance indicators (KPIs) for different traffic patterns representing browsing (e.g., bursty traffic, web, social network feeds), cloud storage downloading of a 3GB high-definition movie, and adaptive bitrate 360-degree video streaming (8K resolution, 120 fps framerate).
An example of 5G NR device in the Tokyo network is shown below in Figure 2. The device achieved a DL data rate of 357 Mbps, which enabled 100 percent video playback at the max 8K, 120 fps video bitrate as shown in the bitrate distribution chart on the right. Figure 2 also showcases the device-level KPIs the simulation platform displays, including data rate, signal quality, spectral efficiency, MIMO rank, and spectrum bands/bandwidths.
Additionally, the simulation platform provides the ability to compare devices across the network for a traffic pattern under different signal quality: 10th percentile (cell-edge/challenging signal conditions), 50th percentile (median user), or 90th percentile (ideal signal conditions).
Here is a summary of key findings:
These findings are a testimony to the intended goals of 5G not only to deliver higher data rates, but also consistent performance even under cell-edge scenarios that open the door to new applications and use cases.
Lastly, the simulation platform provides system-wide comparison across all the device categories for a given traffic type. The overall performance metrics in Figure 4 below shows:
There are a lot of discussions about the promise of 5G speeds. Most, if not all, refer to the downlink and how it can enhance and enable new experiences with minimal reference to the uplink speeds and capabilities. The fact is, uplink is equally important, as app developers also need to anticipate the kind of uplink performance they can expect from 5G networks before they can start thinking about the new generation of apps or how to evolve existing ones.
That is why we also added the wireless industry’s first announced detailed uplink simulations to our 5G NR Network Capacity and User Experience platform, which is designed to deliver quantitative insights on the expected real-world uplink performance and user experience of 5G NR and Gigabit LTE devices in TDD spectrum, operating in SA multimode 4G/5G NR networks.
As shown in Figure 5, the simulation study showcased an uplink capacity increase of nearly 3x when migrating from an LTE-only network, with a mix of LTE devices of various capabilities, to a 5G NR network.
Figure 6 shows an example of an 5G NR device with traffic pattern that represents a PowerPoint file upload to the cloud. The device achieved a UL data rate of 78 Mbps, completing the transfer in less than 30 seconds. The figure also shows device-level KPIs such as signal quality, spectral efficiency, MIMO rank, and spectrum bands/bandwidths.
Another important user experience example is live video broadcast. The quality of that live video depends on uplink quality. With many social media platforms adopting live video broadcast, we see an increase in users interested in live videos broadcast. Figure 7 shows a comparison between a median LTE UL CAT 13 and a 5G NR user. In the simulation, the 5G user can broadcast the live video with 4K quality with no video impairments due to packet loss, while the CAT 13 LTE user doesn’t have enough bandwidth to broadcast anything over 240p, with some packet drops causing freezes in the live video.
The 5G Network Capacity and User Experience Simulation demonstrated the significant potential of 5G by offering insights into the anticipated real-world performance and user experience of 5G and Gigabit LTE in TDD spectrum, operating in the new standalone specifications. The findings also provided quantitative support for the significant gains in downlink and uplink capacities that can be realized by 5G NR, which can support new services and experiences.
In addition to the simulation, we are working on large-scale trials in the second half of 2018 utilizing the Qualcomm Snapdragon X50 5G modem, collaborating with OEMs, infrastructure vendors, and operators in preparation of the first wave of 5G consumer devices expected in first half of 2019.