Editor's Note: Growing requirements for increased availability of IoT devices coincide with the emergence of cellular technologies well suited for the IoT . For developers, the need has never been more acute for more detailed information about cellular technologies and their application to the IoT.
Excerpted from the book, Cellular Internet of Things, this series introduces key concepts and technologies in this arena. In part one, the authors described the evolving landscape for cellular and its role in the IoT. This article reviews massive machine-type communications (mMTC) and ultra reliable low latency communications (URLLC).
Elsevier is offering this and other engineering books at a 30% discount. To use this discount, click here and use code ENGIN318 during checkout.
Adapted from Cellular Internet of Things, by Olof Liberg, Marten Sundberg, Eric Wang, Johan Bergman, Joachim Sachs.
Chapter 1. Cellular Internet of Things
By Olof Liberg, Marten Sundberg, Eric Wang, Johan Bergman, Joachim Sachs
From a service, applications, and requirements point of view, the IoT market is often said to be divided into at least two categories: massive machine-type communications (mMTC) and ultra reliable low latency communications (URLLC). The Next Generation Mobile Networks Alliance is in its 5G white paper  describing these two categories in terms of typical use cases and their associated requirements. Smart Wearables and Sensor Networks are two industrial vertical features mentioned as belonging to the mMTC market category. Smart wearables comprise not only, e.g., smart watches but also sensors integrated in clothing. A main use case is sensing health-related metrics such as body temperature and heartbeat. It is clear that if this trend gains traction, the number of devices per person will go far beyond what we see today, which will put new requirements on the capacity that must be supported by cellular networks providing IoT services. It can furthermore be expected that in order for clothing manufacturers to find wearables an appealing concept, the devices must be extremely compact to support seamless integration in the clothing. The devices must also be of ultra-low cost to attract clothing manufacturers as well as consumers.
Sensor networks is a family name for various utility meters such as gas, water, and electricity meters. Potentially, every home is equipped with a multitude of sensors that will put high requirement on the capacity of the communication system providing them with connectivity. As utility meters are associated with stringent requirements on coverage, which is radio resource consuming, the task to provide sufficient capacity for these becomes even more challenging. Meters may in addition entirely rely on battery power, which will put high requirements on device energy efficiency to facilitate operation for years on small and low-cost batteries.
URLLC can, on the other hand, be exemplified by high-end applications such as automated driving, industrial automation, and eHealth. The news is filled with articles about traditional car manufacturers and giants from the ICT industry competing in the development of autonomous vehicles. If such applications are to be supported by cellular communication networks, the network needs to offer close to perfect reliability combined with support for extreme latency requirements. It is also not far-fetched to imagine remote steering as an attractive alternative to, or perhaps first step toward, fully autonomous vehicles. In this case also requirements on support for high data rates to support high-resolution video may come into the picture. Similar requirements can obviously also be mapped to the industrial automation and eHealth verticals.
Figure 1.1 summarizes the just-made observations with a high level illustration of expected requirements for the mMTC category and the URLLC category in terms of coverage, number of supported connections, latency, throughput, mobility, device complexity, and device battery life. For comparison also typical mobile broadband requirements discussed in Section 2.1 are depicted. The center of the radar chart corresponds to relaxed requirements while the outskirts of the chart map to stringent requirements.
click for larger image
FIGURE 1.1 mMTC, URLLC, and mobile broadband requirements.
1.2.3 INTRODUCING EC-GSM-IOT, NB-IOT, AND LTE-M
The three technologies EC-GSM-IoT, NB-IoT, and LTE-M, described in Chapters 3-8 in this book, were to a large extent designed to serve use cases belonging to the category of mMTC. The work on LTE-M started in September 2011 with the 3GPP feasibility study named Study on Provision of Low- Cost MTC UEs Based on LTE , referred to as the LTE-M study item in the following. The main justification for this study item was to extend the LTE device capabilities in the low-end MTC domain, provide an alternative to General Packet Radio Service (GPRS) and Enhanced General Packet Radio Service (EGPRS) devices, and facilitate a migration of GSM networks toward LTE. As the aim was to replace GPRS/EGPRS as a bearer for MTC services, the study naturally used GPRS/EGPRS performance as the benchmark when setting its objectives. It was, e.g., required that data rates, spectrum efficiency, and power consumption should be at least as good as EGPRS. The main focus of the study was, however, to provide a solution with device complexity and cost on par with GPRS. In September 2012, it was decided to add a new objective on study of the feasibility of a coverage improvement of 20 dB beyond normal LTE coverage.
LTE devices have typically been considered far more expensive than GPRS/EGPRS devices, mainly because of the improved capabilities provided by LTE. Therefore for LTE to become competitive in the low-end mMTC market, a reduction in device cost and complexity was considered crucial. The coverage enhancement was considered to be needed to facilitate deep indoor coverage in locations where, for example, utility meters are expected to be located.
The work on EC-GSM-IoT and NB-IoT started one release later in the 3GPP feasibility study named Cellular System Support for Ultra-low Complexity and Low Throughput Internet of Things , referred to as the Cellular IoT study item in the following. Many aspects of this study can be recognized from the LTE-M study item. The overall objective did, however, change from targeting a solution comparable to GPRS/EGPRS to finding a solution competitive in the Low Power Wide Area segment, which at the time to a large extent was defined by technologies for unlicensed operation designed for ultra-low complexity, extreme coverage range, and long device battery life as discussed in more detail in Section 1.3.
As for the work on LTE-M, an important part of the Cellular IoT study was to meet an objective of extending coverage with 20 dB. While the LTE-M study item used LTE coverage as reference, the Cellular IoT study item defined the improvement in relation to GPRS coverage. The GPRS technology is often seen as one of the most capable technologies in terms of coverage, and today it is still the main technology providing cellular IoT connectivity. But just as for LTE, its capability was believed to be insufficient to cater for deep indoor coverage. Part of the reasoning behind this objective was that modern building materials besides their ability to provide excellent thermal insulation often are characterized by high attenuation of radio waves. A further aspect that was taken into consideration is that many IoT devices have a compact form factor where efficient antenna design is not prioritized. This limits the achievable antenna gain and further increases the need for coverage and increased Maximum Coupling Loss (MCL).
MCL is defined as the maximum loss in the conducted power level that a system can tolerate and still be operable (defined by a minimum acceptable received power level). Hence, it can be calculated as the difference between the conducted power levels measured at the transmitting and receiving antenna ports, and it is an attractive metric to define coverage because of its simplicity. Because it is defined using the antenna connector as the reference point, the directional gain of the antenna is not considered when calculating MCL. Coverage can also be expressed by the Maximum Path Loss (MPL) a technology can support. The path loss is defined by the loss in the signal path (e.g., distant-dependent propagation loss, building penetration loss, etc.) of the radiated power. Hence, it can be calculated by the difference in radiated power levels at the transmitting and receiving antennas. To determine the MPL also the antenna gain at the transmitter and receiver need to be considered. The difference between MCL and MPL is illustrated in Figure 1.2. MCL has been chosen by 3GPP as the metric to evaluate coverage enhancements. In the remainder of the book, the focus will therefore be on MCL with a few exceptions when describing the capabilities of technologies operating in unlicensed spectrum. The concept of MCL is described in Section 188.8.131.52 in detail.
click for larger image
FIGURE 1.2 Illustration of coupling loss and path loss.
In addition to coverage enhancement, it was required in the Cellular IoT study to show that device battery life of at least 10 years were feasible to support, for example, utility meters with no access to a mains power source. To minimize the maintenance cost of applications and equipment relying on limited battery capacity, low energy consumption leading to long battery life is of high importance. A network supporting a large-scale deployment of millions of devices requiring a fresh battery every year would be difficult to maintain on commercial terms.
Furthermore, the candidate technologies were required to present a network capacity sufficient to support IoT type of services for central London under an assumption of 40 connected devices per household, resulting in over 60,000 devices/km2. Last but not least, it was requested to achieve all this with stringent requirements on keeping the device complexity ultra-low to secure as low device cost as possible. Table 1.1 summarizes these performance objectives required to be fulfilled by the candidate technologies developed within the Cellular IoT study item, including EC-GSM-IoT and NB-IoT.
The next article in this series discusses low power wide area networks.
Reprinted with permission from Elsevier/Academic Press, Copyright © 2017