Small satellite constellations, with hundreds of small spacecrafts flying in low orbits and working all together as a communication network, are envisioned as an attractive solution to support and complement 5G New Radio (NR) and beyond 5G communications [1, 2, 3, 4]. Specifically, Low Earth Orbit (LEO) constellations, deployed between 500 and 2000 km over the Earth’s surface, are envisioned to provide service to the three generic use cases defined for the 5G NR: massive machine-type communications (mMTC), enhanced mobile broadband (eMBB), and ultra-reliable low-latency communications (URLLC). eMBB has the primary goal of increasing the user data rates and network spectral efficiency, whereas URLLC and mMTC are the cornerstones of machine-type traffic and, thus, enablers of a wide range of Internet of Things (IoT) applications. However, these three use cases were originally defined for terrestrial 5G networks and their features/requirements must be reconsidered in the physical context of satellite communications. Moreover, the use of small low-cost satellites is key to complement and compete with terrestrial networks, but it severely limits the capabilities of the space infrastructure. The main contributions of this paper are: 1) a detailed description of the small satellite impairments and the main characteristics of LEO constellations, and their role in 5G; 2) a taxonomy for the physical/logical link types; and 3) a systematic review of the relevant technologies to support the three 5G use cases through satellite constellations.
Ii Characteristics of LEO small-satellite constellations
The small satellite category encompasses miniaturized satellites of low cost, low size and weight under 500 kg, in contrast to traditional big satellites of more than 1000 kg. Naturally, there is a close relation between the mass and the capabilities of spacecrafts. High-throughput eMBB or general-purpose space missions will typically require heavier spacecrafts (above 100 kg). This is the case of, e.g., the commercial missions Space X’s Starlink and OneWeb .
A great advantage of satellite vs. ground communications is coverage, which is especially beneficial for mMTC-type services with vast number of devices. A constellation is typically organized in several orbital planes, defined by a given altitude and inclination. The ground coverage area of a satellite is defined as a region of the Earth where the satellite is seen at a minimum predefined elevation. A pass is the period in which a satellite is above the local horizon and available for radio communication with a particular ground position. Due to their low altitude, LEO satellites do not maintain a fixed position in the sky, but rather move fast over the Earth, with each pass lasting a few minutes. As a consequence, a densely populated network is required in the space segment to ensure that any ground terminal is always covered by, at least, one satellite. In the ground segment, one or few dedicated ground stations (GS) are responsible of major control and management tasks of the satellite constellation.
A major challenge in using LEO is the strong Doppler effect between ground terminals and satellites due to the rapid movement with respect to the Earth’s surface. The propagation delay between ground terminals and LEO satellites, being approximately 3 ms at 1000 km above the Earth surface, is non-negligible, but still much lower than for the traditional Geostationary Earth Orbits (GEO), which is beyond 100 ms at a height of 36,000 km. Therefore, LEO constellations can still support a wide range of latency-critical applications as long as the latency requirements are within tens of milliseconds. However, the use of LEO constellations is not viable for the most extreme cases defined in NR, requiring less than 1 ms user-plane and 10 ms control-plane round-trip time (RTT) latency.
A small satellite has stringent energy constraints, exacerbated by the long transmission distances. Small spacecrafts must have active lifetimes of up to five years to prevent frequent redeployment. Therefore, they usually incorporate photo-voltaic solar panels to generate electrical energy from the sunlight and keep their batteries charged during the mission lifespan; these batteries are then employed in eclipse conditions. The periods of direct sunlight on the satellite, when solar energy can be harvested, are determined by its orbital period (around 100 minutes at an altitude of 1000 km above the Earth’s surface), but also by the inclination of the orbit. Hence, the availability of solar energy only depends on the positions of the satellite and the Earth relative to the Sun. Such predictability can be used to operate at maximum capacity and achieve an energy balance.
In terms of communication technology, both free-space optical (FSO) and traditional radio frequency (RF) are considered for inter-satellite and ground-to-satellite links. FSO links have very narrow beamwidth with increased transmission range . While highly susceptible to atmospheric effects and pointing errors, FSO links offer high transmission rates and create less interference. FSO has been demonstrated in ground-to-satellite communication in numerous scientific missions  and several planned commercial LEO constellations, such a SpaceX, Telesat, and LeoSat, will deploy laser communication equipment for high-throughput FSO inter-satellite links . Nevertheless, RF links are crucial as fallback solution if FSO communication is infeasible (bad weather) in hybrid RF-FSO systems, and to enable integration into RF-based systems. For example, ground-satellite communication in 5G NR will either take place completely in the S band around 2 GHz or in the Ka band, where the downlink operates at 20 GHz and the uplink at 30 GHz .
Iii The role of satellite constellations in 5G
Truly ubiquitous coverage is one of the 5G drivers, and this will be possible only by a close integration of satellite networks into 5G and post-5G networks. For that, 3GPP is working in a few study items of Non-Terrestrial Networks (NTN) to define the role of satellite communications in future releases of NR [2, 3]. The goal is to ensure an end-to-end standard in the Release 17 timeframe, tentatively in 2021, thereby bringing order to the present situation where satellite operators use a mix of proprietary and standard-based technologies. A dedicated study for NTN IoT will start soon, paving the way to introduce both NB-IoT and eMTC support for satellites.
The 3GPP work includes traditional satellite networks in Medium Earth (MEO) and GEO orbits, but also LEO constellations . The use cases for satellite networks are divided into three categories:
Service continuity: Continuous coverage to mobile ground terminals that have been previously granted access to 5G services, such as terrestrial vehicles, ships, and airborne platforms.
Service ubiquity: 5G access in areas without terrestrial coverage, including areas where the terrestrial coverage is interrupted by a natural disaster, such as earthquake or flood.
Service scalability: Support to the terrestrial infrastructure in massive multicasting (downlink) or IoT (uplink) applications, as in ultra-high definition TV and ultra-densely IoT deployments.
LEO constellations support these three categories, and the use of low orbits is especially favorable to support latency-sensitive services. The user- and control-plane latency have been suitably redefined to 50 ms RTT for LEO systems .
An interesting application of LEO constellations is as a backhaul of fixed or moving cellular base stations (called gNBs in NR). The main advantage is that locations where terrestrial network coverage is impaired by geographical or economic reasons can easily be reached by the satellite constellation.
Currently, two 5G satellite implementations are envisioned: bent pipe and full satellite gNB. In the first one, the satellites merely serve as relays toward the ground gNBs and, in the second one, satellites are a fully functional gNB. In 5G nomenclature, Xn is the logical interface interconnecting two gNBs and can be used to connect a ground gNB to a space gNB in the constellation, exploiting then fast FSO links among satellites.
A sketch of the architecture to integrate satellite communications into 5G and post-5G networks is illustrated in Figure 1. It encompasses: (1) satellites, which can be of different sizes and in different orbits that play the role of flying 5G gNBs; (2) and ground terminals, which can be end-user nodes (UEs or IoT devices), gNBs, terrestrial gateways, or dedicated GSs. There are two options to connect the UEs to the constellation. The first one is that the UE communicates through a terrestrial gateway (i.e. a relay node), which uses the constellation for backhaul. The big advantage of this approach is that no alterations are needed in the ground terminals that have been already deployed. The second approach is that ground terminals communicate directly with the satellite (e.g., a satellite phone or a direct satellite IoT transmission), where the main challenges are the limited capabilities of the ground terminals.
The satellite constellation is in charge of forwarding the gathered data to the destination; the latter can be a GS, a gateway, a UE or even a big satellite, with high computational and communication capabilities. The satellite segment is supported by one or few GS, which have also much greater capabilities. They are in charge of the command/control of the satellites and can also download data from the satellites to Earth. It is possible for a spacecraft to communicate with more than one GS at a time.
Until now, the 3GPP work addresses only GEO, MEO, and LEO constellations separately. However, it is expected that hybrid architectures combining different orbits will play a major role in future networks , which is reminiscent of the evolution towards heterogeneous cellular networks. The diversity in orbits and satellite choice can complement each other favorably. Thus, the short transmission times between ground terminals and LEO satellites can be combined with the wide coverage and the great communication and computation capabilities of GEO satellites. Besides, MEO satellites are mainly used for navigation (such as the GPS, Glonass and Galileo constellations). Hence, a hybrid orbit deployment largely increases the flexibility of the network and its capacity to guarantee widely different application requirements.
There are three types of data traffic in a LEO constellation: (1) user data, (2) control data, and (3) telemetry and telecommand (TMTC) data. In wireless networks user data and control data are typically separated to facilitate a more efficient management of the wireless resources. For machine-type communications, the size of the control information is similar to the data size, a massive number of devices needs to be handled, and in many cases stringent latency constraints must be met. These requirements are in stark contrast to classical mobile broadband communications.
Moreover, the basic function of a satellite requires extensive contact with ground stations for control, command, communication, and telemetry data return. The acronym TMTC encompasses these spacecraft control-related data, which is inherently different from network control data. Telemetry parameters describing the status, configuration, and health of the spacecraft payload and subsystems are downlinked to ground. In the uplink, commands are received on board of the spacecraft for controlling mission operations and managing expendable resources. Often, TMTC uses separated antennas and frequency bands (UHF/VHF).
V-a Physical and logical links
The physical links in satellite communication are broadly classified in ground-to-satellite links (GSLs) and inter-satellite links (ISLs). Ground refers to any transceiver located on Earth, which can be either a GS, a UE or a gateway (see Figure
in satellite communication are broadly classified in ground-to-satellite links (GSLs) and inter-satellite links (ISLs). Ground refers to any transceiver located on Earth, which can be either a GS, a UE or a gateway (see Figure1). The availability of a GSL is determined by the satellite pass. Since these passes are short in low orbits, the ground end-terminal must perform frequent satellite handover to maintain the connection; this is depicted on Fig. 1 through a feeder link from the GS to a LEO satellite. Moreover, the Doppler shift can be computed as , where is the relative speed between the transmitter and the receiver, is the carrier frequency, and is the light speed. With satellite speeds of several km/s, the Doppler shift can be of several hundred kHz, representing a major challenge of the GSL.
The ISLs are classified in intra- and inter-plane links, where intra-plane links are used for communication within the same orbital plane, and inter-plane links are used for communication between satellites in two different orbital planes. Intra-plane ISLs are usually more stable because the distance between neighboring satellites is mostly constant over time. Conversely, achieving stable inter-plane links is more challenging because the distance between satellites in two different orbital planes changes rapidly, which greatly restricts the time a specific inter-plane link can be maintained (the inter-plane contact time). Furthermore, the implementation of inter-plane links requires frequent handovers, which involves neighbor discovery, neighbor selection (matching), and connection setup (signaling). Contact times or link opportunities are defined as the time a pair of satellites are within communication range. The contact time depends on numerous parameters such as the constellation geometry, the directivity and gain of the antennas, and the transmission power. Also depending on the topology, the Doppler shift can be significant in the inter-plane ISL.
A logical link is a path from the source transmitter to the end receiver. Hence, data travels over many different physical connections, which may not be known by the two end-points. When a LEO constellation is used, there are two different kinds of end-points, satellite [S] and ground [G], which enables the definition of four logical links:
Ground to ground [G2G]. This is the classical use of the network to, for example, relay information between two distant points on the Earth surface.
Ground to satellite [G2S]. For instance, for maintenance and control operations initiated by the ground station.
Satellite to ground [S2G]. One example is Earth observation, in missions where the satellite collects information from several nodes to be downloaded to Earth.
Satellite to satellite [S2S]. This is relevant for satellite-related control applications like, e.g., exploiting the swarm intelligence or other autonomous operations in the space segment.
Each of these four logical links utilizes one or several of the two physical links, GSL and ISL.
The use of the logical and physical links is tightly related to the final application. In LEO constellations, we identify applications that are inherited from terrestrial networks, but also some other space-native cases.
One exemplary application is the use of the constellation as a multi-hop relay network to increase the coverage of IoT deployments in rural or remote areas, where the cellular and other relaying networks are out of range . In such scenario, IoT devices wake up periodically to send a status update. This update is received by the available satellite and forwarded within the constellation until reaching the nearest satellite to the desired ground terminal. This end-to-end application utilizes the [G2G] logical link, and the use of multiple hops to reach the destination can challenge the latency and timeliness requirements.
Another example is to use the constellation for Earth and/or space observation, both native applications of satellite networks. The spacecrafts are equipped with cameras and sensors. The [S2S] might be exploited for cooperation among satellites, for example to point the cameras to a particular position when a first spacecraft detects an unusual event. The [G2S] is needed to retrieve the information in ground.
Figure 2 illustrates the different types of logical links and the carried data for different applications. For each logical link, examples of relevant functions supported by each link are indicated in this figure.
In this section, we describe key technologies for supporting communications with LEO constellations. The technologies are all well-known in terrestrial networks, but important adaptations are required to make them suitable for space and to support the three 5G use cases, as highlighted in Table I.
|Technology||Requirements to support 5G services|
|Physical layer and||Highly flexible OFDM- or GFDM-based candidates for NR in space.|
|radio access||Beamforming for interference mitigation and macro-diversity.|
|Flexible use of grant-based and grant-free RA for the different 5G use cases.|
|Caching||For high data rate (eMBB) and latency-sensitive (URLLC) traffic.|
|In-network caching techniques and tailored policies for mMTC in IoT networks.|
|Distributed processing||Distributed architectures provide robustness against failure and ultra-reliability.|
|Mobility management||Seamless handovers for eMBB and long data transmissions.|
|Flexible RA protocols for mMTC to deal with a great number of access requests.|
|Topology control and Self-||Self-organization and automation are relevant for all services.|
|Organizing Networks (SON)|
|Network Slicing||Natural slicing in hybrid GEO-LEO architectures for different 5G use cases.|
|Relaying and routing||Relaying for enhanced latency-reliability in point-to-point communication.|
|QoS-aware routing protocols that benefit from the heterogeneity of end-user equipments for critical applications.|
Vi-a Physical layer
The waveform defines the physical shape of the signal that carries the modulated information through a channel. In NR, the defined waveform is based on Orthogonal Frequency Division Multiplexing (OFDM) [8, 9], which is very sensitive to Doppler shifts. In LEO, the satellites are moving at a very high relative speed to the ground terminals, making an accurate Doppler compensation and large subcarrier spacing necessary . An alternative approach would be the use of generalized frequency division multiplexing (GFDM), which allows for higher robustness against Doppler shifts at the cost of higher equalization complexity.
NR supports Quadrature Amplitude Modulation (QAM) schemes. Within QAM, the more robust versions, Binary Phase-Shift Keying (BPSK) and Quadrature Phase-Shift Keying (QPSK), are often used in satellite communications, although Amplitude and Phase Shift Keying (APSK) is the preferred technique in LEO commercial missions . The main benefit of APSK in space is its low peak-to-average power ratio (PAPR), which makes it suitable when using power amplifiers with nonlinear characteristic. Terrestrial gNBs adapt the modulation and coding scheme to the current channel conditions, for which the UEs must transmit information about the channel quality to the gNB . In satellite systems, rain fade, inclined orbit satellite operation, antenna pointing errors, noise and interference can affect the satellite link conditions, which can be addressed by a suitable Adaptive Coding and Modulation (ACM). However, ACM can be challenged by the fact that information provided by the terminals may be outdated due to the the large satellite-ground delay and the fast relative movement of LEO satellites.
Beamforming gives the possibility to increase the spectral efficiency in a multi-user scenario by multiplexing the users based on their positions. This is known as spatial division multiple access (SDMA). To allow accurate SDMA, the ground terminals have to be placed with a certain minimum distance between each other. Distributed beamforming, where several satellites form a large antenna array with an antenna spacing of several km , enables SDMA at a much lower distance between the ground terminals, but requires a closer coordination.
Vi-B Radio access and slicing
Radio access in the GSL is, essentially, random access (RA), due to the large amount of nodes and the fact that both the number of UEs and the nature of traffic are not known in advance. Two main types of RA protocols exist: grant-based and grant-free. Grant-based RA is the go-to solution in 5G. However, its excessive signaling overhead, limited resources, and the large two-way latency in protocol handshakes severely limit the scalability of IoT applications. Instead, grant-free RA is preferred for the transmission of short and infrequent data packets that characterize IoT. Nevertheless, the long distance between end points and the star topology prevents the use of traditional channel sensing protocols . Instead, non-orthogonal medium access (NOMA) techniques may be better suited.
In the intra-plane ISL, the transmitter and the receiver do not change, because the relative positions and distances are preserved. Therefore, fixed access schemes like Frequency Division Multiple Access (FDMA) or Code Division Multiple Access (CDMA) are simple and attractive solutions . With FDMA, the frequency reuse to mitigate interference along the orbit must be properly designed, which comes at the cost of higher bandwidth requirements. On the other side, the challenges of CDMA, e.g., synchronization or near-far effects, can be overcome by using NOMA.
In dense LEO constellations, there are situations in which multiple satellites want to establish an inter-plane connection with a given satellite at the same time. The inter-plane ISL can be seen as a mesh network. Unlike terrestrial mobile ad-hoc networks, the position of the satellite neighbors can be predicted if the orbital information is available at each node. Protocols from mesh networks for connecting directly, dynamically and non-hierarchically to as many other satellites as possible and cooperate with one another can therefore be adapted to be used in space systems. These must be optimized to deal with specific conditions in satellite constellations, e.g., relative velocities of satellites in different orbital planes, which, in some cases, may be extremely high. Figure 3 summarizes the radio access conditions in the GSL, intra- and inter-plane ISL.
A general-purpose satellite constellation must support the heterogeneity of eMBB, URLLC and mMTC services. Besides, the user, control and TMTC traffic transmitted through the constellation have widely different characteristics and requirements. For instance, control and TMTC data present much more strict reliability and latency requirements than user data, and can be either unicast, multicast, or broadcast. On the other hand, IoT user data is typically unicast and delay-tolerant.
Network slicing is a key 5G feature for the support of heterogeneous services by ensuring that each service is allocated resources that provide performance guarantees and isolation from other services. In the RAN, the conventional approach to slicing is to allocate orthogonal radio resources at the expense of a lower network efficiency. The multiplexing in time, frequency and space of services and data traffic with very different characteristics and requirements brings major challenges, and it requires priority-aware mechanisms in the data link and medium access layers to guarantee the efficient delivery of the packets.
Vi-C Advanced processing
Caching is an effective way to smooth out network traffic during peak traffic hours and to reduce the latency by drawing the content closer to the end user. There are two main components: the caching decision and the caching replacement strategy. The former refers to selecting the content to be cached, considering the limited cache size. The latter is needed when the size of new content exceeds the amount of free space left in the cache. It is desirable to delete items that are at least-likely needed again in the future. As this is usually not known in advance, contents are replaced based on usage prediction. LEO networks have a major role to play in content caching near the edge. In GEO satellite-terrestrial networks, usually only GSs have caching capability. With the introduction of LEO constellations, the addition of the caching capability to the space segment opens a new dimension, where the LEO layer and its wide coverage can be exploited for caching and, e.g., efficient multicast of the most popular content.
Distributed processing architectures are a promising research topic due to the success of microsystems and microelectronics and the possibility to continuously evolve a satellite network without disrupting its functionality  . Following a federated satellite architecture, satellite networks can exploit resources such as downlink bandwidth, storage, processing power, and instrument time that would be wasted otherwise.
The combination of federated learning techniques and mobile edge computing (MEC) is a promising solution to greatly expand the computational capabilities of LEO networks. Federated learning is a variant of machine learning in which edge nodes contribute to the global model based on locally stored data without sending their data to a centralized entity. Hence, federated learning can take place in the device terminals when the global model is provided by a satellite, but also at the satellite itself when the data is provided by the device terminals or when metadata is collected by the satellite (but these data is not transmitted further).
. Following a federated satellite architecture, satellite networks can exploit resources such as downlink bandwidth, storage, processing power, and instrument time that would be wasted otherwise. The combination of federated learning techniques and mobile edge computing (MEC) is a promising solution to greatly expand the computational capabilities of LEO networks. Federated learning is a variant of machine learning in which edge nodes contribute to the global model based on locally stored data without sending their data to a centralized entity. Hence, federated learning can take place in the device terminals when the global model is provided by a satellite, but also at the satellite itself when the data is provided by the device terminals or when metadata is collected by the satellite (but these data is not transmitted further).
Vi-D Mobility management and network layer
Mobility management is in charge of guaranteeing the continuation of the service as the transmitter and/or receiver move. Unlike terrestrial networks, the need for handover in satellite constellations is mainly determined by the fast movement of the space segment, whereas the speed of ground terminals is negligible. The selection of the next servicing satellite can be planned beforehand, thanks to the predictability of the satellite movements. For this selection, several criteria may be used, such as maximizing the service time, maximizing the number of free channels, or minimizing the distance. The ISL has a key role in implementing handover in LEO constellations.
Topology control and self-organization are of paramount importance in dense constellations. In particular, on-board intelligence and decision making capabilities ensure the accomplishment of the mission goals and the optimization of the network performance. Control automation is essential to reduce the TMTC traffic, which represents one of the bottlenecks in the scalability of a constellation. Exclusively managing a great number of satellites from a few dedicated GSs may be inefficient or unfeasible. Conversely, with a less ground-dependent space segment, the capacity of the ground stations can be exploited to control the traffic. Like in cellular networks, the introduction of Self-Organizing Networks (SON) is important for automatic configuration and optimization, and to reduce the interaction with ground human control as much as possible. One step further in the constellation control automation is the introduction of Artificial Intelligence (AI). This enables, for instance, swarm intelligence, where a population of simple agents interact locally with one another and with their environment.
are of paramount importance in dense constellations. In particular, on-board intelligence and decision making capabilities ensure the accomplishment of the mission goals and the optimization of the network performance. Control automation is essential to reduce the TMTC traffic, which represents one of the bottlenecks in the scalability of a constellation. Exclusively managing a great number of satellites from a few dedicated GSs may be inefficient or unfeasible. Conversely, with a less ground-dependent space segment, the capacity of the ground stations can be exploited to control the traffic. Like in cellular networks, the introduction of Self-Organizing Networks (SON) is important for automatic configuration and optimization, and to reduce the interaction with ground human control as much as possible. One step further in the constellation control automation is the introduction of Artificial Intelligence (AI). This enables, for instance, swarm intelligence, where a population of simple agents interact locally with one another and with their environment.
Minimizing fuel consumption is of utmost importance for small satellites due to their limited suply. Therefore, optimizing the trajectory of the satellites beforehand, but also during flight, is essential.
Relaying in LEO constellations aims to exploit spatial and interface diversity by transmitting directly from source to the destination, but also to one or more neighbors. These are the relay candidates, which can be in LEO, higher orbits, or on the ground, and must be effectively coordinated. Depending on the relay candidates, different interfaces and physical links can be used.
Routing decisions, on the other hand, must be dynamically reconfigured in a LEO constellation as the ground terminals and/or the satellites move. In centralized routing, a central entity creates the routing tables to be distributed along the satellites in the constellation. Conversely, in distributed routing, each satellite may only make simple routing decisions (i.e., up to a couple of hops away) to transmit the packet closer to the destination according to a predefined distance metric. This is an area where opportunistic and geographical-based routing can offer important benefits, specially when combined with network coding schemes. ML techniques can also be used to identify and exploit repetitive patterns in the constellation geometry and minimize routing computations.
Another important aspect of routing is the addressing method tho define the destination, which can be a single node or a plurality of nodes. In LEO constellations, we can have unicast, broadcast, multicast, anycast or geocast. Unicast refers to a one-to-one transmission from one point in the network to another point, both identified by a network address. Broadcast aims to reach all the possible recipients within range, in a one-to-all association. Multicast uses a one-to-many or many-to-many association, differing from broadcast in the sense that only a subset of accessible nodes is addressed. Anycast is also a one-to-one association, but the packet is routed to any single member of a group of potential receivers that are all identified by the same destination address, and typically the nearest node will be selected according to some distance measure. Finally, geocast refers to the delivery of information to a group of destinations in a network identified by their geographical locations.
LEO satellite networks pose an opportunity to fulfill the 5G promise of truly ubiquity. 3GPP is working fast to have LEO satellite constellations in Release 17, which could be known as 5G advanced. In this paper, we have described the main opportunities and challenges of these networks, and discussed their adaptation to be used in space for the three 5G NR use cases: eMBB, URLLC and mMTC. These generic use cases need to be adapted to the physical context of the satellite networks. The architectural evolution of wireless networks based on LEO satellites and GEO/MEO satellites does have a certain similarity with heterogeneous cellular networks, but, again, the satellite context brings a new perspective to the problem. Finally, we have discussed several enabling technologies and outlined their role in supporting 5G connectivity through LEO satellites.
-  3GPP, “5G: Study on scenarios and requirements for next generation access technologies,” TR 38.913 V14.2.0, May 2017.
-  ——, “Study on new radio (NR) to support non-terrestrial networks,” TR 38.811 V15.1.0, Jun. 2019.
-  ——, “Study on using satellite access in 5G,” TR 22.822 V16.0.0, Jun. 2018.
-  B. Di, L. Song, Y. Li, and H. V. Poor, “Ultra-dense LEO: Integration of satellite access networks into 5G and beyond,” IEEE Wireless Communications, vol. 26, no. 2, pp. 62–69, 2019.
-  I. del Portillo, B. Cameron, and E. Crawley, “A technical comparison of three low earth orbit satellite constellation systems to provide global broadband,” Acta Astronautica, vol. 159, 03 2019.
-  H. Kaushal and G. Kaddoum, “Optical communication in space: Challenges and mitigation techniques,” IEEE Communications Surveys Tutorials, vol. 19, no. 1, pp. 57–96, First quarter 2017.
-  W.-C. Chien, C.-F. Lai, M. S. Hossain, and G. Muhammad, “Heterogeneous space and terrestrial integrated networks for IoT: Architecture and challenges,” IEEE Network, vol. 33, no. 1, pp. 15–21, 2019.
-  O. Kodheli, A. Guidotti, and A. Vanelli-Coralli, “Integration of satellites in 5G through LEO constellations,” in Proc. GLOBECOM, Dec. 2017.
-  A. Guidotti et al., “Architectures and key technical challenges for 5G systems incorporating satellites,” IEEE Transactions on Vehicular Technology, vol. 68, no. 3, pp. 2624–2639, Mar. 2019.
-  M. Röper and A. Dekorsy, “Robust distributed MMSE precoding in satellite constellations for downlink transmission,” in Proc. IEEE 2nd 5G World Forum, Oct 2019.
-  S. Cioni, R. De Gaudenzi, O. Del Rio Herrero, and N. Girault, “On the satellite role in the era of 5G massive machine type communications,” IEEE Network, vol. 32, no. 5, pp. 54–61, 2018.
-  R. Radhakrishnan et al., “Survey of inter-satellite communication for small satellite systems: Physical layer to network layer view,” IEEE Communications Surveys Tutorials, vol. 18, no. 4, pp. 2442–2473, Fourth quarter 2016.
-  C. Verhoeven et al., “On the origin of satellite swarms,” Acta Astronautica, vol. 68, no. 7, pp. 1392–1395, 2011.
Beatriz Soret [M’11] received the M.Sc. and Ph.D. degree in Telecommunications from the Universidad de Malaga (Spain), in 2002 and 2010, respectively. She is currently an associate professor at the Department of Electronic Systems, Aalborg University (Denmark). Her research interests are within satellite communications with LEO constellations, low-latency and high reliable communications, and 5G and post-5G systems.
Israel Leyva-Mayorga received the M.Sc. degree (Hons.) in mobile computing systems from the Instituto Politécnico Nacional (IPN), Mexico, in 2014 and the Ph.D. (Cum Laude) in telecommunications from the Universitat Politècnica de València (UPV), Spain, in 2018. He is currently a postdoc at the Department of Electronic Systems, Aalborg University (AAU), Denmark.
Maik Röper [S’18] received the B.Sc. and M.Sc. degree in electrical engineering and information technology from the University of Bremen, Germany, in 2014 and 2016, respectively. Since then, he has been a research assistant with the Department of Communications Engineering at the University of Bremen, where he is currently pursuing the Ph.D. degree. His research interests include satellite communications, precoding and distributed signal processing.
Dirk Wübben [S’01, M’06, SM’12] is a senior researcher group leader and lecturer at the Department of Communications Engineering, University of Bremen, Germany. He received the Dipl.-Ing. (Uni) degree and the Dr.-Ing. degree in electrical engineering from the University of Bremen, Germany in 2000 and 2005, respectively. His research interests include wireless communications, signal processing, multiple antenna systems, cooperative communication systems, and channel coding. He has published more than 130 papers. He is a board member of the Germany Chapter of the IEEE Information Theory Society and an Editor for IEEE Wireless Communications Letters.
Bho Matthiesen (S’12 – M’20) received the Diplom-Ingenieur (M. Sc.) degree and his Ph. D. (with distinction) from Technische Universität Dresden, Germany, in 2012 and 2019, respectively. He is a postdoc at the Visiting Excellence Chair in the Department of Communications Engineering, University of Bremen, Germany. He serves as a publication chair for ISWCS 2020. His research interests are in signal processing and communication theory, with a focus on global optimization methods for resource allocation.
Armin Dekorsy [SM’18] received the B.Sc. degree from Fachhochschule Konstanz, Germany, the M.Sc. degree from the University of Paderborn, Germany, and the Ph.D. degree from the University of Bremen, Germany, all in communications engineering. He is currently the Head of the Department of Communications Engineering, University of Bremen and the head of VDE/ITG Expert Committee “Information and System Theory.” He has authored or coauthored more than 180 journal and conference publications, and holds more than 19 patents in the area of wireless communications.
Petar Popovski [S’97, A’98, M’04, SM’10, F’16] is a professor at Aalborg University, where he heads the Connectivity Section, and holder of a Visiting Excellence Chair at the University of Bremen. He received his Dipl.-Ing./ Magister Ing. in communication engineering from Sts. Cyril and Methodius University in Skopje and his Ph.D. from Aalborg University. He received an ERC Consolidator Grant (2015) and the Danish Elite Researcher award (2016). He is an Area Editor for IEEE Transactions on Wireless Communications. His research interests are in wireless communications/networks and communication theory.