Cellular IoT: How 5G Differs From LTE

  • 299
  • March 29, 2021

5G extends its scope beyond consumer to many new vertical and enterprise markets. Thanks to its flexibility and improved performance, 5G opens the door to many industrial applications.

When researchers and engineers began developing 5G in 2012, they began to look at use cases. The primary motivation for launching a new generation of wireless technology was insufficient spectrum. Most industry analysts predicted an explosion of data traffic that would result in saturation of existing spectrum resources.

Another motivation arose from the expected tremendous growth in the number of connected devices including many new device types for machine type communications (MTC) and Internet of Things (IoT) applications. This expansion gave rise to a variety of requirements that 4G missed. Here’s how 4G and 5G compare in relation to IoT.

5G’s definition took the shape of a now famous triangular icon with three sides depicting the three main 5G components. The triangle was subsequently modified, reused, and adapted by many companies throughout the wireless industry. Figure 1 summarizes the three 5G use cases.

  • eMBB: enhanced Mobile Broadband. Somewhat the same as 4G, but with faster speed and larger capacity. eMBB supports the accelerating growth in the number of consumer devices and to mitigate the expected saturation of 4G networks.
  • URLLC: Ultra Reliable, Low-Latency Communications. URLLC fulfills requirements of vertical market segments such as industrial, health, transportation, and aviation that have high demands for low latency and high reliability. These new use cases came from stakeholders outside of the traditional telecommunications world, such as automotive and energy.
  • mMTC: Massive Machine-Type Communications. mMTC supports a massive number of connected objects. While not necessarily requiring high data rates or low latency, these connected objects have other demanding requirements such as ultra-long battery life, small footprint, and simplicity needed to enable connections for almost any kind of object.

Figure 1. The original 5G triangle, from Recommendation ITU-R M.2083 [ITU-R, IMT Vision—Framework and overall objectives of the future development of IMT for 2020 and beyond]. Recommendation ITU-R M.2083, September 2015, https://www.itu.int/rec/R-REC-M.2083/en

With so many use cases and requirements, 5G needed versatility and support for these generally non-compatible requirements. These requirements were key in driving innovation in 5G design.

The new requirements imposed on 5G include not only a new radio (NR) interface. 5G adds an evolution in core network principles: convergence of wireless and wireline systems, new radio access networks (RANs) and new telecom network core architectures. These aspects are beyond the scope of this article.

5G Design Principles

To meet the three main use case requirements, 5G NR needed more flexibility and higher efficiency than 4G while providing more capacity, higher speed, and lower latency. To improve the capacity, concluded Shannon, one must either increase the bandwidth or improve the signal-to-interference-and-noise ratio (SINR). [Ref. 1]

4G and 5G differ in their use of spectrum. To increase the bandwidth for cellular, regulatory agencies look to repurpose spectrum from other uses. In the U.S., for example, frequencies formerly used for broadcast TV are now allocated for cellular. While 4G resides mainly below 3.8 GHz, 5G uses bands below 6 GHz (frequency range 1, FR1) and 24.25 GHz to 52.6 GHz (FR2).

In 4G, the use of unlicensed spectrum was introduced later in development with LTE assisted access (LAA) and LTE in unlicensed bands (LTE-U). In 5G, the use of unlicensed spectrum was considered early, under the NR-U (new radio unlicensed) name.

More efficient modulation schemes can increase spectral efficiency, resulting in the delivery of more bits per hertz. 5G uses 256 QAM and 1024 QAM, which provides greater spectral efficiency than lower-order modulations. New waveforms, generalization of multiple-input, multiple-output (MIMO) antenna schemes, and the introduction of improved forward error correction (FEC) techniques help to improve SINR.

Many of 4G’s principles continue in 5G. For example, OFDM, OFDMA, and MIMO all came from 4G, and the protocols are almost identical. [Ref. 2,3,4]

LTE vs. 5G

Because flexibility was deemed necessary to meet the needs of a wide variety of new use cases, the time-frequency grid must accommodate different numerologies µ (from 0 to 4), corresponding to the subcarrier spacing (SCS) of OFDM symbols. Numerology 0 refers to a subcarrier spacing of 15 kHz (same as LTE). Numerology 1,2,3 and 4 correspond respectively to 30 kHz, 60 kHz, 120 kHz, and 240 kHz), resulting in different slot durations (the number of OFDM symbols in a slot is kept constant at 14). Table 1 summarizes the numerology. With this flexibility, the NR frame design can accommodate low latency traffic (using very short slot durations), as well as variety of frequency bands (the higher the frequency, the higher the SCS).

Table 1. Various numerologies in 5G NR.

4G LTE protocol has two main frame structures: FDD and TDD. In contrast, 5G NR has 56 slot formats currently defined that operate in either duplex mode, FDD, TDD, or even in self-contained slots that contain downlink and uplink symbols. Such self-contained slots enable fast communication on the air interface, minimizing the transmission time interval (TTI).

Beamforming shows another difference. In 5G NR, all signals are beamformed, which provides better reach and limits the overhead (pilots are transmitted only when needed). The pilot structure is flexible, allowing adaptation to the channel characteristics. Front-loaded pilots let channel estimation occur first, and then demodulate received data symbols on the fly, for faster demodulation.

Additional innovations accommodate operation in mmWave bands. For instance, dedicated pilots such as phase reference symbols (PRS) counteract harmful phase noise.

5G also introduces low density parity codes (LDPC) as forward-error-correction codes for the data channels and polar codes for the control channels. Though polar codes are quite novel, LDPC were already being used in Wi-Fi.

While NR design didn’t introduce anything revolutionary, it is a better version of 4G. It can deal with larger bandwidths and higher frequency bands that LTE.

The first definition of 5G NR in 3GPP was made in the context of Release 15, completed in December 2017. For this release, the focus of standardization was on the eMBB use case, with some enablers for URLLC. 3GPP identified solutions for mMTC as LTE-M and NB-IoT, which were defined in Release 13. This raises more general questions about support of IoT in 5G.

5G IoT Support

4G LTE introduced MTC, which refers to two non-smartphone objects communicating with each other. MTC was originally considered only for low-data-rate devices and applications, generally known as IoT. 5G NR opens the door for communication of more sophisticated and higher data rate objects that must meet stricter latency and reliability requirements. This corresponds to the URLLC side of the ITU triangle (Fig. 1). These more demanding objects are sometimes referred to as industrial IoT or critical IoT objects to distinguish them from low profile IoT objects, mMTC.

3GPP defined 4G LTE in 2012 with Release 8. It was then improved in subsequent releases, adding higher throughput and more features. Release 13 (2016) added two new flavors specifically defined to address IoT: Category M (LTE-M) and narrowband IoT (NB-IoT, category NB). The former operates in regular LTE deployments, using the smallest possible channel size (1.4 MHz) and the latter operates in a 180 kHz channel. That lets it be deployed in standalone mode (typically reusing GSM channels), in regular LTE bands, or within LTE guard bands.

Think of LTE-M and NB-IoT as stripped-down versions of regular LTE, with the design target being low cost, improved (indoor) coverage, and very long battery life. That’s needed for battery-powered IoT applications – utility meters, wearables, alarm panels, and asset trackers. Primary design objectives of LTE-M and NB-IoT include:

  • Reduced cost, smaller footprint: LTE uses two antennas on the device side. LTE-M and NB-IoT use one, simplifying signal processing. Smaller channel sizes further simplify processing. Eliminating the duplexer (the specific filter that protects the receive path from the transmit signal) also simplifies the design in half-duplex FDD (HD-FDD), the mode used in LTE-M and NB-IoT. This simplified design lets a single hardware design operate globally.
  • Improved coverage. Removing one antenna negatively impacts receiver sensitivity. To compensate for this loss and improve the coverage (as necessary for deep indoor deployments such as smart meters), coverage enhancement (CE) modes were introduced. CE modes are simply signal repetitions, a low-cost technique for improving SINR.
  • Long battery life. New power-saving schemes and protocol optimizations let IoT devices enter deep sleep as fast as and for as long as possible, resulting in reduced power consumption.
    LTE-M is richer in capability than NB-IoT. It supports mobility, VoLTE, and a data rates up to 1 Mb/s while NB-IoT is limited to 30 kbps. NB-IoT achieves theoretically better coverage and lower power consumption.

In Releases 14 and 15, 3GPP improved LTE-M and NB-IoT from their initial release. When 3GPP submitted its proposal to ITU for 5G, it submitted NR for eMBB and URLLC, while LTE-M and NB-IoT were accepted as already meeting the requirements for the mMTC aspects of IMT-2020.

3GPP introduced efficient solutions to connect classical objects. Depending on the application, the user can select the most appropriate cellular technology to connect objects, per application, as illustrated in Figure 2. LTE-M and NB-IoT are officially part of 5G.

Figure 2. Cellular IoT offers a solution for every use case.

How can we ensure that all deployed objects using LTE-M, NB-IoT, or even LTE Cat 1 have support going forward? In many applications, IoT connected objects are expected to live in the field for many years (utility meters) and some were originally designed to operate with a 4G core network. The 5G core brings improvement, especially with respect to high-end quality of service, but does not bring any specific benefit to low-end IoT (and more problematic, some power optimization features were not supported in the Release 15 5G core network). There are three options for this problem:

  • IoT devices can support both 4G and 5G core networks, which leads to additional cost and complexity, thus wiping out the low-cost advantages.
  • Upgrade the devices over the air with new firmware when switching from a 4G to a 5G core, assuming an immediate transition on the network and the possibility to upload a complete firmware over the air despite a thin pipe.
  • Network operators can continue to support 4G core functionalities within the 5G core, allowing easy support of legacy LTE devices.

Option three is the most realistic.

5G: Critical and Industrial IoT

5G NR brings significant improvements in latency and data rate compared to 4G, and these improvements are key in meeting the strict requirements in vertical markets such as factory automation (industry 4.0), transport, energy, or entertainment, including augmented and virtual reality. Most of these improvements are defined within the context of the URLLC side of the 5G triangle.

URLLC services are enabled by the flexible frame structure (allowing a very short TTI), preemptive scheduling, and anticipated retransmission for fast turn-around, grant free transmission, etc.

Cellular connectivity, especially in harsh industrial environments, has an inherent advantage over Wi-Fi and even wired technologies. Wi-Fi is less secure and more susceptible to interference than cellular by design and wired technologies are less flexible and more difficult to update or change in a factory layout. Thus, 5G will be a key technology for industrial applications, especially when deployed as a private network where the network owner has full network control.

In Release 16, 3GPP introduced a dedicated working group to address Industrial IoT. The work item [Ref. 5] introduced improved reliability thanks to enhanced packet data convergence protocol (PDCP, an upper layer of the protocol stack) duplication, mechanisms to prioritize traffic between UEs and within a UE, and a means to support time sensitive networking (TSN) [Ref. 6].

TSN is a technique introduced by IEEE 802.1 group for an Ethernet wired network that provides deterministic transmissions by synchronizing various equipment components to a single master clock. [Ref. 7] Mechanisms to ensure deterministic delays and synchronization were defined by IEEE and the objective of 3GPP was to adapt these mechanisms to the wireless and 5G world. Industrial IoT and its subsequent ongoing work item in release 17 called in 3GPP Enhanced IIoT complements URLLC and is expected to fully support the most stringent requirements of critical and industrial connected objects. [Ref. 8]

5G will not replace 4G. Both will coexist for a long time, especially for the LPWA side of IoT for which LTE-M and NB-IoT will remain the solution of choice. With 4G LTE and 5G NR, 3GPP defined a unified toolbox to support professional IoT and a wide range of applications from very simple, low data rate types of connected objects to high performance industrial and critical IoT.

1. Michelle Effros and H. Vincent Poor, “Claude Shannon: His Work and Its Legacy,” EMS Newsletter, March 2017, p.20. https://www.itsoc.org/resources/Shannon-Centenary/shannon-work-legacy-paper.

2. Bob Witte, “The basics of 5G’s modulation, OFDM,” 5G Technology World, April 16, 2020. https://www.5gtechnologyworld.com/the-basics-of-5gs-modulation-ofdm/

3. Bob Witte, “OFDMA improves spectrum use in Wi-Fi 6,” 5G Technology World, June 4, 2020. https://www.5gtechnologyworld.com/ofdma-improves-spectrum-use-in-wi-fi-6/

4. Bob Witte, “More antennas, faster data transfer,” 5G Technology World, July 7, 2020. https://www.5gtechnologyworld.com/more-antennas-faster-data-transfer/

5. See RP-200797.zip available at: https://www.3gpp.org/ftp/TSG_RAN/TSG_RAN/TSGR_88e/Docs/RP-200797.zip

6. Packet Data Convergence Protocol (PDCP) specification (3GPP TS 38.323 version 15.2.0 Release 15), https://www.etsi.org/deliver/etsi_ts/138300_138399/138323/15.02.00_60/ts_138323v150200p.pdf.

7. Dave Cavalcanti, “Five reasons why TSN over 5G promises timely deliveries,” 5G Technology World, February 2, 2021. https://www.5gtechnologyworld.com/five-reasons-why-tsn-over-5g-promises-timely-deliveries/

8. See RP-201310 available at https://www.3gpp.org/ftp/tsg_ran/TSG_RAN/TSGR_88e/Docs/RP-201310.zip

© 2018 SEQUANS.