1.1.1 Major Drivers Influencing the Growth of Future Mobile Traffic

The evolution of wireless communications is driven primarily by the anticipated data traffic explosion and its economical perspective. Many white papers and reports present some of the major global data traffic projections and growth trends. According to the CISCO White Paper [7], the forecast for global mobile data traffic per month ranges from about 7 EB to 49 EB between 2016 and 2021, which means a compounded annual growth rate (CAGR) in the overall mobile data traffic of 47% from 2016 to 2021. The key figures in [8] are quite similar: the total mobile data traffic is estimated as 8.5 and 69 EB/month in 2016 and 2022, respectively, with a CAGR of 45% compared to 70 and 170 EB/month for the total fixed data traffic in 2016 and 2022, respectively, with a CAGR of 20%. Better user experience and lower prices of “smart” devices will accelerate its proliferation. Thus, the projected IPv6 mobile traffic forecast will increase over the same time period from 1.03 EB in 2016 to 27.4 EB in 2021, which means an increase from 14.7% to 56% of the total mobile traffic noted earlier. The CAGR in IPv6 network traffic is then 92% from 2016 to 2021 [7].

Some drivers and trends that impact traffic growth are described in detail in Report ITU-R M.2370 [1], containing global International Mobile Telecommunications (IMT) traffic estimates beyond 2020 from several sources, where the global mobile traffic estimates per month beyond 2020 with and without M2M traffic are illustrated in Figure 1.2. These traffic estimates are significantly higher than those of the aforementioned White Paper [7] for 2020 and 2021. They reach 62 EB/57 EB, 607 EB/543 EB, and 5016 EB/4394 EB with and without M2M traffic in 2020, 2025, and 2030, respectively. This means a CAGR of around 55% and 54% in 2020–2030 with and without M2M traffic, respectively. These estimates anticipate that global IMT traffic will grow in the range of 10–100 times from 2020 to 2030 [1].

Figure 1.2 Mobile traffic estimates per month in EB between 2020 and 2030 without and with M2M traffic. All data values are taken from [1]. The network capacity unit ExaByte is 10¹⁸ bytes.

Among many characteristics and trends that are expected to impact the anticipated traffic growth and the overall traffic demand in 2020 and beyond, there are three major drivers influencing the growth of future mobile traffic [1]:

• ▪ Enhanced mobile broadband, in particular by video usage

• ▪ Application uptake, that is, the rate at which applications are being adopted

• ▪ Device proliferation, accompanied with an evolution toward ever smarter mobile devices in different form factors and with continuously enhanced capabilities and intelligence, which require increasing bit-rates and bandwidth

All are expected to evolve over time.

Within the different service types behind this anticipated traffic growth is M2M with a traffic volume of about 7% and 12% of the total in 2020 and 2030, respectively [1]. By far the largest traffic volume will be consumed by the usage of video-on-demand services (video usage), which will continue to grow, and resolution of these videos will continue to increase. Beyond 2020, people will watch more ultra-high-resolution audio-visual content, regardless of the way the content is delivered. Video streaming with an enhanced mobile broadband is expected to account for almost 75% of all global mobile traffic in 2021, being 4.2 times and 6 times higher than non-video in 2025 and 2030, respectively [1]. The popularity of ultra-high resolution and large screens as well as the increasing number of high-performance smart devices such as smartphones and tablets, widely used around the world, together with the growth of mobile subscriptions will dramatically increase the mobile traffic volume consumed by each user according to Figure 1.3. The figure indicates that each subscriber will consume 5.3 GB, 39.4 GB, and 257.1 GB of data traffic per month on average in 2020, 2025, and 2030, respectively.

Figure 1.3 Estimations of global mobile data traffic per subscriptions per month from 2020 to 2030 without M2M traffic. All data values are taken from figure 10 in [1]

Moreover, the rate at which applications are being adopted by users is accelerating. The annual global downloading of applications was 102 billion apps in 2013 and grew to about 270 billion in 2017. This mobile application uptake and usage of applications will contribute to increased mobile broadband traffic [1] and the broadband access growth rates given by various laws.

Nielsen’s Law of Users’ Internet Bandwidth, that is, users’ internet connectivity in bits per second, is based on data from 1983 to 2016. It states that high-end users’ connection speed grows by 50% per year, reaching 120 Mbit/s and 240 Mbit/s in 2014 and 2016, respectively [Reference Nielsen9]. This means it doubles roughly every 21 months. This 50% CAGR is close to the more established Moore’s Law, which states that computers double in capabilities every 18 months, which corresponds to about 60% annual growth. An observation similar to Nielsen’s has been made by Ulm et al. in [Reference Ulm, Cloonan, Emmendorfer, Finkelstein and Fioroni10]. It reflects a 50% CAGR for the downstream and about 30% CAGR for the upstream capacity, respectively. Similar exponential curves, that is, fitted straight lines of data rates plotted logarithmically against time, can be found for the wireline, wireless (cellular), and nomadic (WLAN) technologies according to Edholm’s Law of Bandwidth. Some trends can be observed from the figures of data rates plotted logarithmically against time [Reference Fettweis and Alamouti11–Reference Cherry13]:

• ▪ Backhaul network bandwidth (wireline) is larger than the ones for cellular (wireless) and WLAN (nomadic) technologies. For example, in 2010, data rates of cellular connections were around 10 Mbit/s, for WLAN connections at about 1 Gbit/s, and for Ethernet connections of office desktops about 10 Gbit/s.

• ▪ Backhaul network bandwidth grows annually at 20–30%, while the bandwidth of wireless and nomadic technologies exhibits a 50% CAGR according to Nielsen’s Law.

• ▪ Since wireless and nomadic data rates are growing much faster than wireline ones, the curves will converge sometime in the future.

• ▪ Wireless and nomadic bandwidth lines are almost parallel with a gap in between of about 100 times, however, gradually converging and cutting at around 2030. This is because use cases for mobile data on devices will converge. Consumers only care about applications and services, not about the underlying network; thus a truly heterogeneous network is needed, leading to a new inflection point of technology [Reference Fettweis and Alamouti11].

Because the highest data speed offered is a determining factor for sizing the network, some open questions arise [Reference Ulm, Cloonan, Emmendorfer, Finkelstein and Fioroni10]: will this broadband demand for wireless and nomadic technologies end, that is, Nielsen’s Law break as Moore’s Law or will service providers offer a residential 10+ and 100+ Gbit/s Internet service by 2025 and 2030, respectively, according to Nielsen’s Law?

As mentioned earlier, another driving factor is the device proliferation accompanied with an evolution toward ever smarter mobile devices in different form factors and with continuously enhanced capabilities and intelligence, which require increasing bit-rates and bandwidth. Globally, the growth of wireless devices, accessing mobile networks worldwide, grew to 8.0 billion in 2016, and it will grow to 11.6 billion by 2021 at a CAGR of 8%, where 8.3 billion will be handheld or personal mobile-ready devices and 3.3 billion M2M connections according to the 2017 CISCO White Paper [7]. Again, the values for the personal mobile-ready devices are close to the one in [8]. The growth of global mobile subscriptions between 2020 and 2030 was estimated in 2015 by the International Telecommunication Union [1], where the values in 2020 and 2021 are higher than the ones of the aforementioned CISCO Report. It estimated the following number of global mobile subscriptions: 10.7 billion in 2020, 13.8 billion in 2025, and 17.1 billion in 2030, respectively. This is illustrated in Figure 1.4, together with the estimated number of mobile connected M2M devices, which would be around 7 billion in 2020, 34 billion in 2025, and 97 billion in 2030, respectively. Hence, it is expected that the total number of devices connected by global mobile communications networks will surpass 100 billion by 2030, which will be more than 10 times greater than the world population at that time, which was estimated at 7.52 billion in July 2017, 7.76 billion in July 2020, 8.14 billion in July 2025, and 8.5 billion in July 2030 [14].

Figure 1.4 Estimations of global mobile and global M2M subscriptions in billions between 2020 and 2030 (top). In addition, the global mobile subscriptions are separated into three categories (bottom). All data values are taken from [1]

Better user experience and lower prices of smart devices will accelerate the growth rate of the penetration of smart devices. Thus, concerning the share of total devices, there will be a rapid decline of non-smartphones or feature phones from about 40% (3.3 billion) in 2016 down to 13% (1.5 billion) in 2021. At the same time, the share of smartphones, including phablets, will increase from about 45% (3.7 billion) in 2016 up to 53% (6.1 billion) in 2021. The shares of tablets, PCs, and other smart devices will smoothly go up from 4% in 2016 to 5% in 2021, while a noticeable growth is expected in M2M connections with shares of 10% in 2016 to 29% in 2021 [7]. Again, the estimation of global mobile subscriptions, separated for feature phones, smartphones, tablets, and other smart devices between 2020 and 2030 from [1], is given in Figure 1.4.

1.1.2 5G Activities, Vision, and Objectives

The framework, the overall objectives, and timelines of the development of IMT, including user and applications trends, usage scenarios and capabilities, technology trends and technical feasibility, traffic growth estimation, and spectrum implications as well as the expected evolution is regularly defined and released in Recommendations of the International Telecommunication Union (ITU), offering at the time of its publication some visions of future systems and networks. The most important ones for mobile communications are taken as basic references for this book chapter.

The Recommendation for IMT-2000 in 2000 [15] provides a single set of standards for 3G, which marks a major change in cellular mobile communications evolution. Based on IMT-2000 (3G), the ITU has released Recommendation ITU-R M.2012 in the terrestrial radio interface of IMT-Advanced in 2012, after nine years for its development, where 4G specifications are based on the IMT-Advanced requirements. Along with the enhancement of IMT-2000 (3G) and IMT-Advanced (4G) systems, a new IMT vision was released in 2015 [4], defining the framework and overall objectives of the development of future radio interfaces with far more enhanced and new capabilities of IMT in 2020 and beyond, called IMT-2020 (5G).

In parallel, different activities, involving academic research groups and telecommunication companies, have worked on or are still working on 5G scenarios, test cases, new system concepts, network architectures, and technologies, influencing the standardization and regulatory processes. A brief overview over some activities is given in [Reference Mitra and Agrawal6]. An example is the largest 5G project METIS (Mobile and wireless communications Enablers for the Twenty–twenty Information Society) within the Framework Program 7 (FP7) of the European Union, where top telecommunication companies and top academic institutions have been evolved. Their comprehensive outcomes are summarized in the METIS final project report in 2015 [16] (Deliverable D8.4), based on in total 29 deliverables, including scenarios, requirements, and key performance indicators for 5G mobile and wireless system (D1.1), channel models (D1.2, D1.4), test-bed/demonstration results (D1.3), requirements and general design principles for new air interface (D2.1), novel radio link concepts and state of the art analysis (D2.2), components of a new air interface-building blocks and performance (D2.3), proposed solutions for new radio access (D2.4), multi-node/multi-antenna transmission technologies (D3.1, D3.2, D3.3), network-level solutions (D4.1, D4.2, D4.3), spectrum needs and usage principles (D5.1, D5.3, D5.4), METIS system concept and technology roadmap (D6.6), and architecture and system evaluation results (D6.4, D6.3). In the subsequent METIS-II project (https://5g-ppp.eu/metis-ii/), the key objectives have been

Stimulated by the METIS outcome and on the aforementioned 5G/IMT – 2020 Report [4], Wei Xiang et al. published a comprehensive book in 2017 [Reference Xiang, Shen and Zheng3], in which a large number of authors have been involved, revealing the enabling techniques for 5G networks and addressing the challenges and opportunities of 5G mobile communications.

Based on these three major literatures on 5G, its vision is briefly summarized in the text that follows, starting with the major market drivers beyond 2020, which is expected to be MI and IoT, including M2M or D2D connectivity [Reference Xiang, Shen and Zheng3, 4]. Both MI and IoT will trigger a massive number of use cases, for example

• ▪ Mobile Internet (MI): Ultra-high density (UHD) and 3D video, augmented reality, virtual reality, online gaming, mobile cloud, remote computing, tactile Internet, 3D connectivity to aircrafts and drones, collaborative robots, smart office, as well as

• ▪ Internet of Things (IoT): Smart grid and critical infrastructure monitoring, mobile surveillance, environmental monitoring, industrial automation, eHealth services, smart wearables and smart body area networks, sensor networks, smart homes/buildings, smart cities, smart transportation, self-driving and connected cars (Internet of Vehicles).

All these use cases can be grouped into three usage scenarios [Reference Xiang, Shen and Zheng3, 4] or generic 5G services [16], addressing different use case characteristics according to Figure 1.5:

• ▪ Enhanced/extreme mobile broadband (eMBB) for improved performance and an increasingly seamless user experience, offering human-centric use cases for access to multimedia services at least with IMT capabilities (see Table 1.1)

• ▪ Massive machine-type communications (mMTC) characterizes a large number of connected devices, typically with relatively low-volume, non-delay sensitive data at low cost and very long battery lifetime

• ▪ Ultra-reliable and low-latency communications (URLLC) with stringent requirements for capabilities such as throughput, latency, and availability for applications such as industrial process monitoring and manufacturing, remote medical diagnosis and surgery, as well as safety and automation in smart grids.

Figure 1.5 Usage scenarios (generic 5G services) for mobile communications beyond 2020 with key use case characteristics and some exemplary use cases according to [4, 16].

While MI is focused on people-oriented communications, IoT provides communications between things and between things and people. Some of its use cases are indicated in Figure 1.5. UHD and 3D videos, over-the-top services such as social networking, as well as desktop cloud, augmented reality, and online gaming will significantly drive up data rates, even at high mobility in high-speed trains and connected cars (Internet of Vehicles), where massive data activities of a vast number of users will require a high system capacity. For services such as smart homes, smart cities, as well as environmental monitoring, future networks have to support massive connectivity of devices with low energy and cost. While some use cases such as desktop cloud, augmented reality, and online gaming will additionally need “imperceptible” latency, “safety-critical” use cases such as e-banking, eHealth, self-driving cars, smart grid, and critical infrastructure monitoring face extreme security requirements, that is, in fast response with very low latency in the lower millisecond range and with nearly absolute reliability and perception of 99.999% availability [Reference Tripathi and Khaparde2–4, 16].

It is expected that 5G/IMT-2020 will provide far more enhanced capabilities than 4G/IMT-Advanced, defining nine key performance and efficiency indicators summarized in Table 1.1.

Among these key capabilities, the user experienced data rate, connection density, and latency might be the most fundamental ones. While all key capabilities may to some extent be important for most of the aforementioned use cases, the relevance of certain key capabilities may be significantly different for the individual use cases or usage scenarios [Reference Xiang, Shen and Zheng3, 4, 16]. In these references, the importance of key capabilities for the three usage scenarios are given in form of a “spider-web” and “blooming flower” pictures together with some explanations, and in addition, the enhancement of key capabilities from 4G/IMT-Advanced to 5G/IMT-2020.

Compared to 4G, 5G should have three to five times higher spectrum efficiency and more than 100 times improvement on energy and cost efficiency to outperform previous generations of mobile communications systems. Moreover, to keep up with the rapid traffic growth, some core design objectives of 5G wireless networks are [Reference Tripathi and Khaparde2–4, 16, 17]:

• ▪ Implementation of massive (up to 1,000-fold) area traffic capacity (see Table 1.1) and massive device connectivity for at least up to 100 billion devices according to the predictions in Figure 1.4.

• ▪ Providing a fiber-like user experienced data rate by 10–100 times (see Table 1.1), that is, of up to 10 Gbit/s, for example, to support mobile cloud service, capable of extremely low latency and response times.

• ▪ Support for an increasingly diverse set of services, applications, and users – all with extremely diverging requirements as discussed earlier for the usage scenarios.

• ▪ Flexible and efficient use of all available noncontiguous spectrum for wildly different network deployment scenarios, which will be described in the next section.

1.1.3 Spectrum Allocation

To cope with the anticipated 1,000 times higher traffic capacity and the up to 100 times higher typical user data rate in 5G wireless networks, considerably more spectrum is required than currently available for mobile and wireless communication systems. Hence, in the 5G vision, another significant driver is the vast amount of spectrum available in the mm-wave range [Reference Xiang, Shen and Zheng3, 16, 18–20], for example, for cellular hot-spot coverage to satisfy consumer demand for high-speed wireless access with ultra-low latency. Technical feasibility of radio interface technology and systems operating in frequency bands between 6 GHz and 100 GHz, considering propagation characteristics, antenna technology, active and passive components, physical layer, and medium access control design as well as deployment architectures, is carried out by simulations and performance tests and trials and are published in [18, Reference Rappaport, Heath, Daniels and Murdock19, Reference Rappaport21].

Spectrum has been in the past and will also be in future one of the most valuable resource for mobile communications. Therefore, agencies and standardization organizations worldwide aim for international harmonized spectrum and full-spectrum access, especially above 6 GHz. Hence, beyond the sub-6 GHz bands for 5G in Europe 3.4–3.8 GHz, USA 3.1–3.55 GHz and 3.7–4.2 GHz, Japan 3.6–4.2 GHz and 4.4–4.9 GHz and China 3.3–3.6 GHz, 4.4–4.5 GHz and 4.8–4.99 GHz for 5G phase I, frequency bands in the mm-wave range are already foreseen for 5G phase II: in Europe 24.25–27.5 GHz and 31.8–33.4 GHz, United States 27.5–28.35 GHz and 37–40 GHz, Japan 27.5–29.5 GHz, China 24.75–27.5 GHz, South Korea 26.5–29.5 GHz. Moreover, some studies and tests are carried out by Samsung at 13.4–14 GHz, 18.1–18.6 GHz, 27.0–29.5 GHz, and 38.0–39.5 GHz. China allocated 40.5–42.5 GHz and 48.4–50.2 GHz for fixed point-to-point wireless access systems as well as 42.3–47.0 GHz and 47.2–48.4 GHz for mobile point-to-point wireless access systems [Reference Xiang, Shen and Zheng3].

In the V-band, a continuous spectrum is allocated for wireless communications in unlicensed mode: 57–64 GHz in the United States, 57–66 GHz in Europe, 59–64 GHz in China, and 59–66 GHz in Japan [Reference Xiang, Shen and Zheng3, Reference Rappaport, Heath, Daniels and Murdock19, 20]. In Europe, the European Telecommunications Standards Institute (ETSI) is working to facilitate the use of the E-band from 71.0 to 76.0 GHz and 81.0 to 86.0 GHz, and in the future, on the channelization of the W-band from 92.0 to 114.5 GHz and the D-band from 130.0 to 174.8 GHz for large-volume (high capacity) backhaul and front-haul systems as well as for innovative solutions for fixed broadband access [20].

Most of the GEO- and MEO-HTS make efficient use of both, K_u-band and K_a-band (e.g., O3b downlink 17.7–20.2 GHz, uplink 27.5–30 GHz). LEO high-throughput satellite constellations also aiming to operate in the K_u-, K_a- and V-bands.

International harmonized spectrum and spectrum regulation will support future mobile network implementation and other services, enabling large-scale usage of the spectrum. Parts of the aforementioned bands do not have global mobile allocation, and since several other services such as different satellite services, radiolocation, and radionavigation will use portions of these bands, some compatibility studies to determine the feasibility of using these bands must be discussed and managed on a global level, for example at World Radiocommunication Conferences [Reference Martin, O’Keefe and Finucan22].

1.1.4 Key Technology Drivers

The deployment of 5G networks is expected to emerge between 2020 and 2030. In [Reference Tripathi and Khaparde2-4, 16, 17] different deployment scenarios and environments are considered in detail such as various indoor hotspots (offices, shopping malls, stadiums) and various urban and rural, high-speed, and highway environments, as well as detailed requirements on the key performance indicators and many more capabilities, including also the requirements for the next generation of radio access network (RAN), which should support a wide range of intercell coordination schemes and accommodate new mobile access technologies for massive capacity, huge numbers of connections, and ultra-fast network speeds. Figure 1.6 illustrates some deployment scenarios for 5G and beyond cellular systems, indicating some usage scenarios such as cellular 5G, LTE-A, HSPA or GSM, D2D, mm-wave- and intervehicular communications in macro-, micro-, pico- and femtocells, where external communication flow is provided by a core network, which might consist of hybrid radio backhaul and/or fiber access networks, including satellites. Hence, 5G will realize networks capable of providing connectivity between people and machines/devices (M2M and H2M) over very small as well as long distances. Moreover, 5G radio access will be built upon both, evolved existing wireless technologies (GSM, HSPA, LTE, LTE-A, M-WiMAX, and WiFi) and new radio access technologies, that is, 5G core networks will also be equipped to seamlessly integrate current mobile core networks.

Figure 1.6 Deployment scenarios envisioned for 5G and beyond cellular systems.

While previous generations of wireless networks were characterized by fixed radio parameters and spectrum blocks, 5G network nodes will be flexible/configurable based on cloud, software-defined networking, and network function virtualization technologies [4].

Integration of mass-scale cloud architectures or cloud-based RAN will infuse mobile networks with capabilities for delivering services in a very flexible way, available to the customer on demand, which might leverage new services and applications. It will provide on-demand resource processing, storage, and network capacity (in a centralized server) wherever needed in a rapid speed, in order to cope with the tremendous growth in mobile data traffic, the diversification of mobile app innovation, IoT connectivity, and security, but also to facilitate on-demand customization. Thus, it will ensure QoS, and at the same time, reduce network cost and energy consumption. Software-defined air interface technologies will be seamlessly integrated into 5G wireless access network architectures toward a “hyper transceiver” approach to mobile access, which will help to realize the joint-layer optimization of how radio resources are efficiently utilized [17].

In addition, new integrated access node and backhaul design will enable ultra-dense networking of radio nodes. Plug-and-play will become essential to deployment, where such nodes will need to access and self-organize available spectrum blocks for both access and backhauling. This capability will be a key for enabling high-frequency spectrum radio access. All these adaptive network solutions will allow a much higher flexibility to utilize “any” spectrum and “any” access technology for the best delivery of services and to speed up the creation of massive-scale services and applications [17].

Moreover, if cognitive radio technology can be used for 5G cellular systems, then, it would enable the user equipment/handset to look at the radio landscape in which it is located and choose the optimum RAN, modulation scheme, and other parameters to configure itself to gain the best connection and optimum performance [Reference Tripathi and Khaparde2]. For this, reconfigurable baseband and RF hardware architectures are required to enable computationally intensive and adaptive air interfaces.

Hence, 5G will use advanced heterogeneous networks and new kinds of network deployments, in order to create an integrated system that can provide and support a variety of services and applications with flexibility, including ubiquitous ultra-broadband network infrastructure, mass-scale cloud architectures, ultra-dense radio networking with self-backhauling, D2D communications, and dynamic radio access infrastructure sharing. Thus, in 5G and beyond cellular systems, users will concurrently be connected to several wireless access technologies and seamlessly move between them (pervasive networks). With different access schemes, it will be possible to link, for example, devices to nearby devices to provide ad-hoc wireless networks for much higher data flow and low-latency connectivity. Furthermore, in current cellular networks, data rates fall toward the cell edge, where interference levels are higher and signal levels lower. To overcome this, that is, to make high data rates available over a wider area of the cell, group cooperative relay will be applied in 5G networks [Reference Tripathi and Khaparde2].

As enabler for designing programmable/flexible air interfaces, various advanced waveform technologies combined with advances in modulation and coding as well as advances in multiple access schemes such as Filtered OFDM, Filter Bank Multi Carrier, Pattern Division Multiple Access, Sparse Code Multiple Access, Interleave Division Multiple Access, and Beam Division Multiple Access are essential to achieve continuing improvements in spectral efficiency. In the latter concept, an orthogonal beam is allocated to each mobile station, that is, this technique will divide the antenna beam according to locations of the mobile stations for giving multiple accesses to the mobile stations, which correspondingly increases the capacity of the system [4, Reference Rafay5, 16, 23].

Other techniques such as flexible backhaul and dynamic radio access configurations can also enable enhancements to the radio interface. Moreover, flexible spectrum usage, joint management of multiple radio access technologies, and flexible uplink/downlink resource allocation such as Time Division Duplex (TDD) and Frequency Division Duplex (FDD) joint operation and dynamic TDD provide technical solutions, which address the growing traffic demand and allow more efficient and flexible use of radio resources. The latter one could also be attained by advanced RF-domain processing, for example using single-frequency full-duplex radio technologies, where simultaneous transmission and reception on the same frequency with self-interference cancellation could increase spectrum efficiency significantly. Improvements in these areas will drive overall network costs down while achieving improved energy efficiency [4, 17, 23].

Another major element of any 5G cellular system in the future are smart antennas, to alter the narrow beam direction to enable more direct communications and suppressing interference effectively, thus increasing the overall cell capacity and spectrum efficiency. This can be done by using advanced antenna technologies such as analog or digital 3D-beam-forming, active antenna systems, massive MIMO, and network MIMO. However, a significantly more advanced baseband computation and RF and antenna hardware are required to meet the complex requirements of digital 3D-beam-forming and mass-scale MIMO [23].

Finally, to cite [17]:

1.2.1 High-Throughput Satellites in Geostationary and Medium Earth Orbits

These HTS in the geostationary orbit make efficient use of the allocated frequency spectrum, both at K_u-band and K_a-band, use much larger capacity payloads with several hundred transponders and extensive multibeam and steerable beam antenna arrays. Moreover, adjustable spot beams enable greater flexibility for the operator to direct capacity according to the customers’ demands, considering an average lifespan of a GEO satellite of about 15–20 years during which market requirements can change significantly.

The use of multibeam antennas to provide coverage of small service areas, typically with 100–250 km diameter footprints (see Appendix 1), rather than the individual wide beams or spot beams of earlier satellites, has provided a significant increase in data handling capacity and throughput for all primary satellite communications services fixed satellite service broadcasting satellite service and mobile satellite service, because of frequency reuse, and higher EIRP on the transmit side and higher G/T on the receiver side. The design and principles of multibeam antennas are given in chapter 15.2 of [Reference Ippolito24, Reference Minoli26, Reference Swinford and Grau27], where the total available bandwidth in Hz and the capacity or overall throughput of the multibeam satellite in bit/s can be determined by

$B_{\mathrm{tot}}=\frac{P\cdot N_b}{N_c}{B}_w\mathrm{and}{C}_{\mathrm{sat}}={\eta }_s{B}_{\mathrm{tot}}$(1.1)

where $P=1$ if polarization diversity is not employed and $P=2$ for dual-polarization frequency reuse and $N_b$ is the number of beams of the satellite, $N_c$ is the cluster size, $F_r=P\cdot N_b/N_c$ is the frequency reuse factor, $B_w$ is the bandwidth allocate to the satellite, and ${\eta }_s$ is the spectral efficiency in bit/s/Hz of the modulation and coding scheme.

By software-defined payloads, these GEO-HTS will provide flexible coverage, power, and connectivity [Reference Suber28, Reference Dios29]. Moreover, advanced modulation and coding schemes in the next-generation modems allow users to increase channel throughput, and adaptive coding enables more efficient use of higher frequency bands that are intrinsically susceptible to rain attenuation. Typical services that GEO-HTS provides are high data rate very small aperture terminal (VSAT), transmission control protocol/Internet protocol (TCP/IP) over satellite, IP-TV, HD-TV, and UHD-TV [Reference Ippolito24, Reference Minoli26].

Few providers plan to build up global networks with GEO-HTS, such as Intelsat, for applications such as maritime and aeronautical, cellular backhaul, consumer broadband, and enterprise networks. Moreover, some providers started with K_u-band transponders to develop the market, then moved to the K_a-band as demand increased [Reference Suber28, Reference Dios29]. One of the earliest broadband satellites in geostationary orbit was Thaicom 4 (IPSTAR), launched in August 2005, operating in the K_u-band and providing multibeam coverage with 84 two-way spot beams, 3 shaped broad area beams, and 7 one-way broadcast beams. Examples of other broadband, high-capacity satellites are the following, with brief descriptions of their capabilities according to [Reference Ippolito24, 30]:

• ▪ K_a-Sat 9A, built by Eutelsat, launched December 2010, location 9° east, lifetime 15 years, 82 K_a-band spot beams, 250 km footprint, 10 gateways, 70 Gbit/s total throughput capacity

• ▪ ViaSat-1, built by Space Systems Loral, launched October 2011, position 115° west, lifetime 12 years, 56 K_a-band transponders, 72 spot beams, coverage continental United States, Alaska, Hawaii, and Canada, 140 Gbit/s total throughput capacity

• ▪ EchoStar 17, built by Space Systems Loral, launched July 2012, position 107° west, lifetime 15 years, 60 K_a-band downlink beams

• ▪ lntelsat 29e Epic, built by Boeing Satellite Systems, launched January 2016, global coverage, transponders: 14 C-band, 56 K_u-band, 1 K_a-band, wide beams, area beams, 25–60 Gbit/s total throughput capacity per satellite of the Intelsat Epic^NG platform

• ▪ lntelsat 33e Epic, built by Boeing Satellite Systems, launched August 2016, position 60° east longitude, lifetime >15 years, coverage: Europe, Africa, Asia, transponders: 20 C-band, 249 K_u-band, 1 K_a-band, wide beams, area beams

• ▪ ViaSat-2, built by Space Systems Loral, launched June 2017, position 69.9° west, lifetime >15 years, K_a-band transponders, coverage North and South America, 300 Gbit/s total throughput capacity, hybrid (chemical and electric) propulsion

To reduce latency, HTS can also be placed in a MEO, which can result in three to four times lower delay compared with GEO. However, in these MEO-HTS, users require steerable antennas to track the satellite and retain signal connectivity by moving from one satellite to another one within the MEO constellation [Reference Ippolito24, Reference Minoli26]. An example of an MEO-HTS system is the O3b satellite constellation, operated by O3b Networks, founded in 2007 and that is now a wholly owned subsidiary of SES S.A. It is designed for telecommunications and data backhaul from remote locations, providing high-speed connectivity to Internet service providers and phone companies. The first four satellites were launched in June 2013, and eight more in 2014. However, only nine are used operationally. There are plans to extend this to 20 satellites. The satellites were deployed in an MEO along the equator at an altitude of 8062 km (4.8 times closer to Earth, 14 dB less path loss, that is, 20 times lower power requirement than GEO), each making 4 orbits a day. Its service area can be covered continuously by a relatively small constellation compared to LEO. Each satellite is equipped with 12 fully steerable K_a-band (downlink 17.7–20.2 GHz, uplink 27.5–30 GHz) antennas (2 beams for gateways, 10 beams for remotes) that use 4.3 GHz of spectrum (2 × 216 MHz per beam) with a proposed throughput of 1.6 Gbit/s per beam (800 Mbit/s per direction), resulting in a total capacity of 16 Gbit/s per satellite. Each beam’s footprint measures 700 km in diameter. O3b claims a one-way latency of 179 ms for voice communication, and an end-to-end round-trip latency of 140 ms for data services compared to about 500 ms for GEO. The satellites weigh approximately 700 kg each [31–Reference Barnett34]. More examples and details about GEO- and MEO-HTS are given in chapter 15 in [Reference Ippolito24] and chapter 3.12 in [Reference Minoli26].

The growth in HTS has been accelerated in recent years as costs for ground terminals and satellite transponders have been coming down. Globally, the number of potential subscribers for satellite services is expected to rise quickly from about 1.5 and 3.1 million subscribers in 2010 and 2015, respectively, up to about 6 million in 2020 according to the projected growth in satellite broadband subscribers by ITU assessments. Thus, the number doubles every 5 years. This rapid anticipated growth highlights the need for continued development and launching of HTS networks, which promise a more competitive market and a key enabler to meet universal broadband targets than previous iterations of satellite technologies [Reference Martin, O’Keefe and Finucan22, Reference Ippolito24].

1.2.2 Low Earth Orbit High-Throughput Satellite Constellations

The demand for connectivity anywhere, anytime and to any device, will continue to grow in traditional sectors such as wireless connectivity and enterprise networking, but it will also open up new opportunities in providing services to the fast-growing mobility sector, that is, people/vehicles-on-the-moveFootnote ², aiming in particular for MI (including ultra-high density and 3D video, augmented reality, virtual reality, online gaming, mobile cloud, remote computing, tactile Internet, and 3D connectivity to aircrafts and drones and smart office) and to extend IoT (including M2M and H2M) to broadly distributed entities [Reference Ippolito24, Reference Emery35]. This could include access to people’s smartphones, tablets, and laptops, as well as to various smart devices in recreational vehicles, buses, trucks, in future autonomous driving and connected automobiles, trains, boats, ships, and airplanes. Advances in smaller, less costly low-orbit satellites, capable of delivering high-speed, bidirectional services will make this possible. By positioning in low orbit rather than geostationary, low-power processors and wireless transceivers for the device back on the ground have similar technology gains [Reference Kelly36].

To ensure global coverage, mobility with low latency, high throughput, and reasonable signal power will require to deploy huge satellite constellations in LEO and/or hybrid communications solutions combining wireless terrestrial and satellite networks that work together seamlessly, since no single technology or company might reach all the possible markets and customers that will be available in the future [Reference Ippolito24, Reference Emery35]. Thus, satellite connectivity will play a critical role.

According to [Reference Fitzsimons37], nearly 4.4 billion people around the world have never been online, at least are “unconnected or underconnected” [Reference Kelby38–Reference Grossman40]. For example, in India, 80% lack access to the Internet. Even in the United States and in Europe, full coverage is not achieved, and a certain percentage of the population are lacking high-speed broadband. This is because of both economic and technological factors. For example, rural areas have much lower average revenue per user than urban ones, since it would entail significantly large backhaul infrastructure [Reference Velivela32].

This motivates new ventures to provide reliable and fast Internet to the masses via new satellite constellations. By December 2019, at least 20 systems are announced for Earth observation and at least 6 new LEO systems for telecommunications [Reference Grossman40–43]. Google, SpaceX (Space Exploration Technologies Corporation), and Boeing recently announced large investments in LEO constellations around the globe to provide Internet to rural and developing areas of the world. Google announced plans to deploy 180 small LEO satellites to provide Internet access to underserved regions of the globe. SpaceX’s proposed constellation Starlink would comprise a network of 4,425 K_a- and K_u-band small operational satellites in the category between 100 kg and 500 kg to make global Internet service a reality. Launching with SpaceX’s own rockets, the full constellation would take about 10–15 years, starting in 2019. Latency is assumed under 20–30 ms with speeds up to 1 Gbit/s through ground terminals [Reference Velivela32, Reference Kelby38, Reference Henry44]. Moreover, SpaceX proposes an even larger constellation of 7,500 V-band satellites, which would circle in an orbit below the first constellation [Reference Henry44]. Also, Boeing applied for a license of the US Federal Communications Commission (FCC) to operate between 1,396 and 2,956 satellites in LEO to provide Internet access in the V-band around the world [Reference Alleven39, Reference Werner41]. With all these satellite constellations, price per bit is aimed to approach terrestrial cellular and fiber offerings.

Because of first-mover advantage, another entity, OneWeb (founded by Greg Wyler of O3b), seems to be ahead of the schedule of all the aforementioned campaigns [45]. Among many investors are SoftBank Group, Qualcomm, Virgin Group, Hughes, Intelsat, and Airbus Group. However, in March 2020, OneWeb faced a liquidity crisis and considered bankruptcy. Then, in July/August 2020, the UK Government announced that it had acquired a 45% stake in OneWeb Global for US$500 million in a joint venture with Sunil Mittal's Bharti Global of India who would hold 55% (formerly a partner of OneWeb). It was reported that the UK would repurpose the satellites for its own Global Navigation Satellite System. It was also announced that Hughes Network Systems would invest US$50 million in the consortium [46].

The original schedule of OneWeb was to design and build 648 (+250 spare) small communication satellites in a constellation, which will offer high-speed Internet with global coverage, connecting both Internet gateways (cellular backhaul, seamlessly extend the networks of mobile operators) and end users for satellite broadband, emergency services, and mobiles, connecting people/vehicles-on-the-move. The 648 satellites in the 150 kg category will operate in circular LEOs (in 18 orbital planes of 36 satellites each) at approximately 1200 km altitude, transmitting and receiving in the K_a (20/30 GHz) and Ku (11/14 GHz) bands. Their mission life time is expected to be at least 5 years. OneWeb aims to be the first fully global, pole-to-pole HTS system with a total throughput capacity of each satellite of more than 7.5 Gbit/s and low latency less than 30 ms round trip delay and a downlink data rate to the user terminals of 50 Mbit/s, making interactive applications such as 5G mobile telephony, game playing, and web surfing possible [Reference Dios29, Reference Velivela32, Reference Henry44, 46, Reference Azzarelli47]. According to [Reference Kelly36], each satellite is designed to provide an aggregate downlink capacity of 17–23 Gbit/s. The first 74 satellites of OneWeb had already been launched in 2019. Generally, to launch a large number of satellites for LEO constellations is challenging the satellite launch services industry, since even each rocket launch might carry about 35 satellites, it requires several tens of launches [48, 49]. For example, the establishment of the aimed OneWeb constellation would have required the greatest rocket campaign in the history of spaceflight. Arianespace, OneWeb’s primary launch partner, booked already 20 Soyuz rockets to launch clusters of 32–36 satellites into orbit. But OneWeb also had contracts with Virgin Orbit for 39 launches, using its still-in development LauncherOne dedicated smallsat vehicle as well as BlueOrigin for 5 launches with a future New Glenn rocket [Reference Henry44, 45]. The high-cost structure of OneWeb's contracted launches and competition, particularly with SpaceX's self-launched Starlink constellation, might have accelerated OneWeb’s failure [46]. However, this means not the end of LEO constellations. At the moment, the race is led by Elon Musk’s Starlink, which has already launched more than 600 satellites till August 2020, but whose target is 42000 satellites. Moreover, Amazon just got approval from the Federal Communications Commission (FCC) of the United States to launch its Kuiper project with its 3326 satellites (784 at a height of 590 km, 1296 to 610 km and 1156 are located at 630 km) into space, but it is not the only one with this goal in mind [49].

1.2.3 Evolving High-Altitude Platform Stations

Beside terrestrial and satellite platforms, other high-altitude platform stations (HAPSs) are evolving due to the growing demand of broadband services, using solar-powered air balloons and space glider or drones. The ideas are not all new, but emerging technologies and innovative Internet entrepreneurs might leverage them. In the project Loon, Google aims to build up an inexpensive ground-based infrastructure with solar-powered, remote-controlled air balloons within Earth’s atmosphere to provide 3G/LTE connectivity directly to the users on the ground. These HAPSs shall be placed above 20 km height to provide a wide area below compared to the very small ones with cell towers, but with latency-like terrestrial technologies [Reference Martin, O’Keefe and Finucan22, Reference Velivela32, Reference Kelby38]. Facebook works on a small solar-powered space glider or drone called “Aquila,” with a wingspan roughly the same as that of a Boeing 737, but weighs only about 450 kg. The upper surface of the wing is covered by solar cells to power the aircraft’s four electric motors and to charge the batteries as power storage for night flights. Aquila will be able to stay at an altitude between 18 and 27 km for months at a time. It uses lasers to transmit data between planes and to terrestrial stations within 50–80 km, which can then provide Internet service locally [Reference Martin, O’Keefe and Finucan22, 50–52]. Also, Google has shown interest in high-altitude drones to deliver wireless Internet access around the world, using mm-waves, which could offer up to 40 times higher datarate than today’s 4G LTE systems [Reference Martin, O’Keefe and Finucan22, Reference Velivela32].

The advantages of using HAPSs as alternatives to satellites are their ease of deployment and ability to move easily to new locations in order to meet the demand and the changing requirements of the operator or service provider’s business plan. These concepts of HAPSs are realizable now because of the improvements in composite materials of the lightweight aircraft technology, low-power computing, battery technology, and solar panels. With respect to spectrum resources, the ITU Radio Regulations currently designated several frequency bands for HAPSs in 2 GHz, 6.5 GHz, 27–31 GHz, and 47–48 GHz ranges. As part of its agenda, WRC-19 will consider additional spectrum requirements for gateway and fixed terminal links for HAPS in the 21.4–22 GHz, 24.25–27.5 GHz, and 38–39.5 GHz bands allocated to the fixed service and could take a decision on designation of some additional bands for HAPSs [Reference Martin, O’Keefe and Finucan22].

1.2.4 Satellite Markets and Perspective

Revenues in the satellite industry are a relatively small percentage of global telecommunications revenues. In 2015, they comprised about 4% according to the Satellite Industry Association (SIA). But in terms of money, it is still a large market with an overall worldwide satellite industry revenue of $260.5 billion in 2016 with a 2% growth rate [43]. Figure 1.7 exhibits the global satellite industry revenues and growth rates from 2007 to 2016, indicating a 10-year global industry growth of two times.

Figure 1.7 Global satellite industry revenues and growth rates from 2007 to 2016 according to [43].

The satellite industry revenue in [43] is subdivided into four major segments:

• ▪ Satellite Services is the largest industry segment, with revenues reaching $127.7 billion, but remained flat with 0.2% growth. It includes broadcast, radio, and broadband among the consumer communication services with $104.7 billion, but also revenues from fixed satellite services with $17.4 billion, comprising transponder agreements and managed network services, mobile satellite services with $3.6 billion as well as Earth observation with $2.0 billion.

• ▪ Satellite Manufacturing revenues dropped down from $16.0 billion in 2015 to $13.9 billion in 2016, suffering the greatest revenue loss of 13% globally.

• ▪ Satellite Launch Services Industry revenues, which include revenues for all commercially competitive launches that occurred in 2016, increased by 2% to $5.5 billion after decreasing by 9% from 2014 to 2015. There were 64 worldwide commercially procured launches in 2016, subdivided in orbits: 28 GEO, 2 beyond GEO, 6 MEO, and 28 LEO.

• ▪ Satellite Ground Equipment includes network equipment (gateways, control stations, very small aperture terminals) and consumer equipment (satellite TV dishes, satellite radio equipment, satellite broadband dishes, satellite phones and mobile satellite terminals, and satellite navigation stand-alone hardware). The revenues rose by 7% over 2015 to reach $113.4 billion.

By the end of 2016, the total number of operational satellites was 1,459, with an increase of 47% over the preceding 5 years, where 35%, 14%, and 1% are for commercial, government, and nonprofit communications, respectively; 19% and 1% for Earth and space observation, respectively; 12% and 5% for R&D and scientific purposes, respectively; 7% for navigation; 2% for meteorology; and 6% for military surveillance [43]. Currently, there are 1,738 satellites orbiting Earth [Reference Grossman40].

The growing importance of the industry is underlined by both the continued increase in the sheer number of operational satellites in orbit and by the announced plans for new satellites with increased capabilities as well as multiple constellations with very small, lightweight satellites, having very low production costs to make them economically viable. The challenge is enormous, since satellites have never been mass produced before. A requirement to produce several small satellites a day has inspired engineers and researchers to develop innovative designs and processes that will dramatically lower the cost in large volumes for high-performance space applications [45]. These smallsats will take advantage of recent advances in consumer electronics, incorporating the latest technologies and many sophisticated functions from smartphones.

Hence, constellations of inexpensive smallsats will be a major market driver (of the manufacturing and launch markets) for the next decade. The shape of the next decade is strongly reliant on the deployment of constellations with inexpensive smallsats. For the future LEO satellite market, it is expected that more than 65% of the total 3,832 satellites will be launched between 2017 and 2025 (2,067 until 2020 and 1,765 from 2021 to 2025) will operate in constellation, experiencing a 385% growth as compared to the previous decade, where the peak of satellites per year to be launched will be around 2021 [Reference Puteaux42]. These new classes of commercial non-geostationary orbit (NGSO) systems have received a wave of investment from key players inside and outside the traditional satellite industry.

Future LEO satellite constellations aim to provide direct broadband capacity to the user all around the globe, extending terrestrial broadband connectivity and providing direct-to-customer Internet connectivity in remote areas. These emerging networks will have a global impact on humanity

• ▪ By delivering ubiquitous high-bandwidth communications to underserved but hard-to wire regions of the world, which contain nearly 60% of the world’s population, including remote and rural (difficult-to-reach) areas, for example to obviate the need for expensive fiber infrastructure

• ▪ To support maritime, aviation, and other vertical markets

• ▪ To enable signaling offload or creating an emergency network, for example an emergency 5G slice in the future, during times of disaster when terrestrial networks are damaged [Reference Martin, O’Keefe and Finucan22, Reference Kapovits25]

The use of electric propulsion for keeping a satellite in its operating position has already changed the global satellite industry. Although it takes months with this type of propulsion technology (because electric thrusters produce very small flow), rather than weeks with traditional chemical propulsion (which ejects large amount of propellant) to reach the final operating position or to maneuver the satellite into another slot during the satellite’s lifetime (which could be up to 25% of an operator’s fleet), economic advantages make it attractive to replace chemical propulsion by all-electric propulsion in a satellite. This approach aims to reduce the spacecraft weight, which could be only 50% of the full-chemical propellant one, and hence it could reduce the launch cost significantly, and possibly extend the spacecraft life [Reference Minoli26].

Moreover, new launch platforms such as the Falcon 9 rockets operated by SpaceX, LauncherOne by Virgin Orbit, and New Glenn rocket by BlueOrigin are approaching the market. Competition and modular design with reusable rocket technology will lower the cost of launching satellites [Reference Kelly36, Reference Henry44, 45].

1.2.5 Interoperable Satellite Networks

In the past, there had already been schemes of interoperable satellite systems in geostationary and low orbits, to combine their strengths such as low latency, low loss, small terminals, high look angle, and pole coverage for LEO as well as wide beams for global coverage (e.g., for broadcast) or spot beams for high-density areas (e.g., for broadband services) for GEO, using its fixed positions. With the emerging new LEO constellations, schemes are going ahead, using GEO and LEO interoperable networks [Reference Suber28, Reference Dios29], relying on the core wireless standard 3GPP to seamlessly integrate with future networks such as 5G phase II and 6G. According to [Reference Suber28], “Intelsat is working closely with players outside of the satellite industry such as QUALCOMM to integrate their chip technology into terminals. This will transform satellite communications from last mile solution to become an integral part of the Telecommunications network,” where the future architecture and different path to HTS are shown in some illustrations there.

1.2.6 Toward Hybrid Terrestrial 5G/6G–Satellite Networks

Besides direct broadcasting, satellite communication technology has been associated with expensive proprietary solutions, addressing primarily niche markets and applications. In the past years, satellite technology has improved both its technical performance and capabilities as well as becoming much more competitive with terrestrial solutions and infrastructure. However, in the long run, satellite networks, even LEO constellations, might not exist as stand-alone islands in the sea of connectivity. Up to now, satellites have been used in terrestrial cellular mobile networks primarily for backhaul services for long-distance and international connections only, but not to any great extent in terrestrial cellular mobile networks. However, this begins to change as fourth-generation mobile systems move into the fifth and sixth generation, where IP-based services will be widely deployed, allowing true explicit end-to-end device addressability. The integration of satellite communication promises to provide a powerful networking infrastructure. Hence, the emerging 5G and 6G visions opens up a new chapter in communications, since in the move toward 5G phase II and 6G there will be a convergence of service delivery via multiple networks and systems by aiming to merge several technologies, integrating terrestrial cellular networks alongside and in combination with satellite systems and possibly high-altitude stratospheric platforms, operating on global distribution levels. The satellite will require very high capacity links, extending the performance of current HTS systems and that might operate in expanded and new frequency allocation bands to support increasing traffic capacity [Reference Ippolito24–Reference Minoli26].

Obviously, there are some characteristic constraints of satellite communications technology compared to terrestrial solutions: (1) the low signal level; (Reference Tripathi and Khaparde2) the comparatively large propagation delay, that is, the large latency, even for LEO constellations; and (Reference Xiang, Shen and Zheng3) the relatively low capacity, even for GEO-HTS, since satellites are difficult to build and launch and power is limited even considering powerful satellites, relying on solar power or batteries.

All these factors might be improved with the new LEO-HTS systems in future. At least for long-distance communications, latencies can be better than fiber, since speed of light is around 40% higher in space than in silica glass and satellites need fewer hops with less repeaters [Reference Velivela32]. But can LEO-HTS systems bypass existing providers? Most probably not, since these new networks will be reliant on high-bandwidth interconnects. LEO constellations are undoubtedly an efficient solution for delivering service with an aggregate downlink capacity of several Gbit/s into large expanses as well as inaccessible or expensive to cable or terrestrial network areas. Here, satellite technology offers the potential for lower capital expenditure. But its downlink capacity is a fraction of modern domestic fiber interconnects higher than 100 Gbit/s.

In opposite, there are also some features and capabilities of satellite communication that distinguish it from terrestrial solutions: (1) the capability to address large geographic regions, even continents, using a minimum amount of infrastructure on the ground; (2) its flexibility; and (3) its intrinsic broadcast capability, which enables satellite communication to deliver the same content to a very large number of network nodes and user devices with unparalleled efficiency. Because of these capabilities of satellites, it should be considered as complementary, rather than a replacement for fixed or terrestrial mobile infrastructure, extending coverage into areas where it is not cost effective to do so terrestrially or supporting terrestrial solutions in delivering on the 5G and 6G promises, some of which are very challenging [Reference Kapovits25, Reference Velivela32, Reference Kelly36].