The vision of a seamlessly connected society and of an ‘Internet of things’ is on the horizon. The first 5G NR specifications completed last year set the foundation for the mobile industry to work towards a full- scale development of next generation 5G wireless technology. We are familiar with the three major 5G use cases: enhanced Mobile BroadBand (eMBB), massive Machine Type Communications (mMTC) and Ultra-Reliable and Low Latency Communications (URLLC). (Note: For the Dec 2017 3GPP release, mMTC and parts of URLLC were considered ‘out of scope’ and will be part of a subsequent release.) From 1G to 4G, the mobile network has evolved tremendously from simple analog voice to the digital broadband communications we have today. Advances in technology enabled every generation to deliver revolutionary user experience – changing the way we work, live, and communicate. With 5G, the expectations are higher. The spectrum is already very congested below 6 GHz and we need to move to frequencies where higher bandwidths are available, e.g. into the millimeter wave bands, to meet the demand for higher speed and capacity. The air interface needs to be flexible to meet the diverse needs of future services and applications across industries. Mission critical applications also require the network to be able to provide ultra-reliable and low latency communications. To meet the 5G targets, 5G NR introduces both evolutionary and revolutionary changes in the physical layer with respect to LTE. We shall briefly explore the motivation and implications of some of these key changes and if you need more details, check out the webcast “Understanding the 5G Physical Layer”.
Waveforms / Frame Structure
The NR physical layer is designed to be flexible and scalable to support the diverse 5G use cases, coexistence with LTE, and forward-compatibility. We shall discuss three new concepts that help make this possible. They are scalable orthogonal frequency-division multiplexing (OFDM) numerology, numerology multiplexing, and dynamic time-division duplexing (TDD). One aspect of scalable OFDM numerology refers to the scaling of subcarrier spacing (SCS). The permitted SCS for different channel bandwidths are shown in Table 1. There are several reasons why SCS scaling is considered an advantage over the fixed SCS in LTE. At mmWave frequencies with very wide channel bandwidths, phase noise makes the use of LTE’s 15 kHz SCS impractical. Also, the length of each OFDM symbol in time is the inverse of the SCS in frequency and to enable very high throughput and low latency services for eMBB at mmWave frequencies, shorter symbols are beneficial both to increase data rates and reduce latency with shorter frame structures. Thus, at mmWave we see use of SCS up to 120 kHz for data channels and 240 kHz for broadcast and synchronization. At the other end of the use case scale, mMTC requires low power, high coverage and low data rate services that are best served using narrowband signals. In this case, SCS narrower than 15 kHz can be used, although this is not part of Release 15.
The NR frame structure has the same frame length (10 ms) and subframe length (1 ms) as in LTE but this is where the similarity ends. An NR slot comprises 14 OFDM symbols compared to the seven of LTE. In addition, the NR slot duration is variable. For any SCS spacing of 2 n × 15 kHz where n is an integer and 15 kHz is the subcarrier spacing used in LTE, the slot duration scales inversely as 1 ms/2 n . Hence, in a 1 ms subframe the number of slots will change depending on the subcarrier spacing used.
Subcarrier spacing can be selected based on the desired service therefore numerology multiplexing is supported, whereby a channel can support multiple numerologies (hence use cases) of 5G services simultaneously.
A slot can be all downlink, all uplink or mixed. The mixed configuration can be static, semi-static or dynamic. The first two configurations are just as implemented in LTE. The dynamic configuration however enables the network to adjust DL and UL resources flexibly according to the instantaneous traffic demands. To indicate link direction in dynamic TDD, Slot Format Indication (SFI) informs the user equipment (UE) whether an OFDM symbol is DL, UL or flexible. Unlike the fixed scheduling in slots, mini-slots have a flexible start position and can be allocated anywhere in a slot with a flexible duration. This is important for low latency cases where transmission needs to take place almost immediately without having to wait for the start of the next slot. A DL mini-slot can have 2, 4 or 7 symbols.
Massive MIMO, Beamforming and Beamsteering
Massive MIMO is a key innovation which is expected to be commercialized by 5G to help increase the spectral efficiency especially in the sub 6 GHz bands. Massive MIMO increases capacity by having far more antennas in the base stations than the number of user terminals. By using TDD channel reciprocity, the power-consuming calculations are done by the base station transmitter thereby making massive MIMO possible for battery-powered devices. The Keysight white paper, “Examining the challenges in implementing and testing massive MIMO for 5G” provides a deeper insight into massive MIMO starting from basics to addressing its challenges in test and implementation.
Massive MIMO beamforming increases capacity by multiplexing users in the same time and frequency, but how do we prevent all these signals from interfering with each other? Beamforming starts by estimating the instantaneous channel propagation conditions for each user. The transmitted signals are then pre-coded with the inverse of the channel to create beams of unknown shape which are then transformed by the channel into orthogonal signals at each receiver. In the corner case with no multi- path but just line of sight conditions, Massive MIMO creates narrow beams pointing at each user as in the simpler geographic beamsteering process which does not require channel estimation.
To meet the demands for more capacity and speed, 5G opens the possibility of accessing new frequency bands in the lower ends of the mmWave spectrum from 24 GHz to 52.6 GHz. Moving to higher frequencies gives the advantage of larger bandwidth and higher throughput but brings about higher propagation and penetration losses. It also unlocks the ‘Pandora’s box’ full of propagation challenges associated with shorter wavelengths such as scattering, diffraction and absorption. Smaller cells and advanced antenna techniques like beamsteering are used to get around mmWave propagation challenges. Beamsteering achieves a stronger signal-to-noise ratio by providing high gain in specific spatial directions. This helps to mitigate the high propagation losses at mmWave. So how does the UE find and connect to a base station if the downlink signal is directional?
This is where the differences in Initial Access comes in. As seen in Figure 1, the gNB (3GPP 5G Next Generation node B or base station) periodically transmits beams at different angles (beam-sweeping) by transmitting a SS/PBCH block (Synchronization Signal and Physical Broadcast Channel Block). When the UE identifies the strongest beam, it starts a Random Access Procedure, using timing and angular information, to transmit in the reciprocal direction of the best downlink beam. When the UE has completed the procedure, data transfer can take place on the UE-specific beam. Beam refinement procedures are then used to maintain beam reciprocity as the UE moves relative to the network. These procedures are not fully developed in Release 15.
Going Over-the-Air (OTA) for 5G testing
5G brings fundamental changes compared to previous mobile generations. It is the aim of the application of this new technology to realize a worthy vision of unifying the plethora of wireless access technologies and providing seamless connectivity of people and ‘things’. The same transformation also applies to the approach in testing 5G. It can’t get there in an incremental way. Keysight’s lead technologist, Moray Rumney highlights this in his blog, “We are going to have to move all of our testing to radiated, not just some of it like we do today, and that’s a big deal”. In previous generations (< 3 GHz), testing is done almost exclusively via connectors and cables. That changes when you move to massive MIMO at mmWave frequencies, or even the theoretical applications at sub-6GHz, requiring a paradigm shift in how devices and systems are tested. Antenna arrays used to realize techniques like beamforming and beamsteering are typically highly integrated devices, with antenna elements bonded directly to ICs, making it difficult, if not impossible, to connect with cables and test. The dynamic or active nature of antenna arrays also means it is not possible to extrapolate end-to-end performance from measurements of individual antenna elements. OTA is very much part of 5G testing (although some case like the sub 6GHz frequency range (FR1) will remain cabled), but this testing introduces a more challenging radiated channel between the component or device and the base station where the imperfections in the channel need to be accounted for during test. The Keysight white paper “OTA Test for Millimeter-Wave 5G NR Devices and Systems” addresses these OTA test challenges. It describes a compact antenna test range (CATR) which has now been approved by 3GPP as a permitted test method for far-field characterization. The CATR uses a parabolic reflector to collimate radiated energy from a test probe towards the DUT, thus providing the advantages of smaller footprint and lower path loss than the direct far-field method which can only be used for devices with antenna apertures below 5 cm. In addition to the CATR approach for RF tests and simple demodulation tests, there are different types of anechoic chambers available to allow testing of spatial aspects such as beam acquisition and tracking or for demodulation using spatially coloured signals.
5G NR is evolving rapidly. The push into mmWave frequency bands to access new spectrum for 5G NR is creating an abrupt and significant shift in the way commercial communications devices and systems will be designed and verified. Test and measurement tools and solutions must evolve with the technology and standards to keep pace with market developments.
Keysight is working extensively across the global 5G ecosystem, through engagements with the standards bodies, commercial partners, various consortia and collaborations to offer solutions that address the latest 5G test requirements and innovation across new and existing technologies in all stages of the design cycle. The race is well and truly underway and Keysight will help you take the lead—from evolution and revolution to reality.