Spectrum Sharing in Satellite-Terrestrial Integrated Networks: Frameworks, Approaches, and Opportunities

Bodong Shang, Zheng Wang, Xiangyu Li, Chungang Yang, Chao Ren, and Haijun Zhang Bodong Shang, Zheng Wang, and Xiangyu Li are with Eastern Institute for Advanced Study, Eastern Institute of Technology, Ningbo, China.Chungang Yang is with the School of Telecommunications Engineering, Xidian University, China.Chao Ren and Haijun Zhang are with University of Science and Technology Beijing, China.

Abstract

To accommodate the increasing communication needs in non-terrestrial networks (NTNs), wireless users in remote areas may require access to more spectrum than is currently allocated. Terrestrial networks (TNs), such as cellular networks, are deployed in specific areas, but many underused licensed spectrum bands remain in remote areas. Therefore, bringing NTNs to a shared spectrum with TNs can improve network capacity under reasonable interference management. However, in satellite-terrestrial integrated networks (STINs), the comprehensive coverage of a satellite and the unbalanced communication resources of STINs make it challenging to effectively manage mutual interference between NTN and TN. This article presents the fundamentals and prospects of spectrum sharing (SS) in STINs by introducing four SS frameworks, their potential application scenarios, and technical challenges. Furthermore, advanced SS approaches related to interference management in STINs and performance metrics of SS in STINs are introduced. Moreover, a preliminary performance evaluation showcases the potential for sharing the spectrum between NTN and TN. Finally, future research opportunities for SS in STINs are discussed.

I Introduction

Due to the boosting need for broader and more reliable wireless network coverage and data transmission, non-terrestrial networks (NTNs) have emerged as a promising solution. In urban areas with heavy traffic or remote areas where terrestrial networks (TNs) face geographical deployment limitations, NTNs can provide seamless access and ubiquitous connectivity. Spectrum integration in satellite-terrestrial integrated networks (STINs) is a necessary step to improve the utilization efficiency of scarce spectrum resources. However, the coexistence of NTNs and TNs may be challenging, especially with regard to spectrum utilization and interference management. When NTN nodes such as low-earth orbit (LEO) satellites are incorporated into existing TNs via certain frequency bands, interference patterns can become more complicated due to different spatial propagation characteristics of signals. For example, the use of the Ka band or the Ku band for satellites may cause severe co-channel interference in existing communication systems. Moreover, within a fixed spectrum allocation, it is possible that while certain portions of the spectrum in NTNs are fully occupied, other segments within TNs may remain unutilized. The imbalance will lead to under-utilization of available spectrum resources and a significant hindrance to substantial improvements in system performance.

In recent years, the rapid growth of NTNs has paved the way toward realizing STINs. STINs, a promising candidate for future 6th-generation (6G) wireless communications, incorporate LEO nodes into existing TN to take advantage of NTN and meet diverse communication requirements [1, 2]. By using STINs, the coverage, flexibility, and spectrum efficiency of the wireless network can be significantly promoted, as terrestrial users can access various TN and NTN nodes from vast geographical areas. Inevitably, the spectrum resource will become scarcer in STIN as more NTN nodes are introduced. This encourages us to manage the spectrum when TN and NTN coexist in the STIN [3]. Meanwhile, the licensed spectrum to the static TN nodes cannot always be utilized efficiently, as spatial spectrum holes cannot be eliminated within the geographic area of interest. Hence, mobile NTN users can admit access to users within the spectrum holes and promote spectrum efficiency [4]. In existing work, one of the promising technologies for TN and NTN coexistence is spectrum sharing (SS), which allows the NTN to share the spectrum pre-allocated to the TN. SS allows the NTN to work in the same frequency band when the TN users are active, as long as its interference to the TN can be tolerated [4, 5]. In this way, the STIN can provide high-quality and seamless wireless services at a slight cost of performance loss under the practical restrictions of limited spectrum resources.

With the distinct advantages above, STINs can overwhelm conventional TNs in a variety of practical fields. As the key technique in STINs, SS becomes the main bottleneck in network performance promotion and faces several challenges [2]. First, a universal SS architecture cannot always meet the diverse needs of the STIN. Therefore, the SS frameworks for all possible STIN transmission scenarios should be carefully designed, including the uplink and downlink of both TN and NTN. Second, since interference in STIN leads to system performance degradation, mitigating interference becomes the most critical issue. Generally, such mitigation can be implemented by isolating users from the interference source and carefully managing the TN and NTN spectrum [6, 7]. Meanwhile, the TN and NTN spectrum should be allocated appropriately, and the spectrum access scheme should be carefully designed [8]. In addition, since STINs are hierarchical and heterogeneous, the proper selection of performance metrics for system evaluation is also essential to the system, depending on the design objectives.

To address the above concerns, we first consider the isolation of STIN users. One of the well-known approaches for this is the utilization of a protection zone for TN, where NTN users within the protection zone are not allowed to access the spectrum shared by TN and NTN [6]. However, the appropriate definition of protection zones still remains a challenging problem. Meanwhile, some studies have been carried out for spectrum allocation and management. Typical work involves spatial SS, dynamic SS, and game-theoretic SS approaches, which are potentially beneficial for STINs. In spatial SS, the objective is to find the geographic areas that are good for NTN access in STINs [9]. In dynamic SS, the objective usually becomes finding the temporal holes in the spectrum resource by treating the NTN as a secondary network [3, 10]. Both spatial SS and dynamic SS can be considered as frequency reuse schemes in STINs. For spectrum allocation and sharing, game models are competitive candidates in finding the optimal scheme by modeling the TN and NTN as game players [11, 12]. All the above topics can be further studied to meet the performance requirements of future STINs.

Refer to caption — FIGURE 1: An overview of SS in STINs.

This article focuses on the SS frameworks for various STIN scenarios and the state-of-the-art approaches to SS in STINs. Our main contributions are as follows.

•

We provide detailed discussions of possible SS frameworks in STIN and summarize their geographic and transmission characteristics, including all uplink and downlink scenario combinations for TN and NTN.
•

We investigate several state-of-the-art approaches for possible SS in STINs to promote spectrum efficiency and system utilization from the spatial, temporal, and frequency domains. Detailed illustrations of these approaches are also provided.
•

We summarize the potential performance metrics for evaluating the efficiency of SS in STINs and validate the metrics via a practical case study. Finally, future research opportunities of SS in STINs are highlighted.

The remainder of the paper is organized as follows. Section II discusses the SS frameworks for various STIN scenarios, including uplink and downlink for TN and NTN in the network architecture. In Section III, we discuss advanced approaches to SS in STINs. In Section IV, we perform practical SS case studies to validate the above frameworks for STINs. Section V summarizes the future research directions of SS in STINs. Finally, the conclusion remarks are given in Section VI, and an overview of SS in STINs is provided in Fig. 1.

II Frameworks of SS in STINs

In this section, we present four frameworks of SS in STINs based on different combinations of uplink and downlink in NTN and TN, as illustrated in Fig. 2. In scenarios without SS, NTN and TN operate on their respective dedicated frequencies. However, this leads to the inefficient use of TN’s dedicated spectrum in some remote areas. By implementing SS for NTN and TN in the development of STINs, the total bandwidth is divided into two segments: 1) shared spectrum for both TN and NTN, and 2) reserved spectrum exclusively for NTN, as shown at the top of Fig. 2. This approach allows NTN and TN to share a substantial amount of spectrum resources flexibly and efficiently. Notably, a portion of the spectrum is reserved for NTN because specific existing TN equipment may not be compatible with the higher frequency bands used by NTN. Therefore, this reserved spectrum can only be utilized by NTN. The following subsections provide a detailed introduction to the four frameworks depicted in Fig. 2.

II-A NTN Downlink and TN Downlink SS

When a satellite transmits signals to an NTN user and a BS transmits signals to a TN user, they can access the shared spectrum, as shown in Fig. 2(1). In this case, TN users can be regarded as primary users and NTN users as secondary users. Designing a three-dimensional (3D) protection zone for each TN user is essential. By adopting such a strategy, NTN users should ensure that there are no TN users within the radius distance of the protection zone before receiving signals. In addition, satellites cannot be present within the 3D area that includes TN users and BSs. This is because the sharing of spectrum between TN and NTN will not only enhance the communication performance of TN users by virtue of the expanded bandwidth but also result in interference. The protection zone is mainly set to protect the quality of service (QoS) of TN users. Theoretically, NTN users can also be affected by interference leakage from BSs, considering their beam directionality on the antenna surface in the horizontal plane. However, near-vertical transmissions from satellites can hardly be influenced. Moreover, satellites can mitigate mutual interference by constraining spectrum access within a specified distance range.

II-B NTN Downlink and TN Uplink SS

Consider a scenario involving sharing a reverse link spectrum in which an NTN user receives signals from a satellite. As depicted in Fig. 2(2), both the satellite and the NTN user can access the shared spectrum through TN uplink communications. Satellite spectrum sensing can be performed to detect the strength of TN uplink signals to reduce mutual interference. If the uplink signal exhibits a low level of strength, the NTN downlink transmission can be operated to prevent significant mutual interference. To be more specific, due to the directionality of satellite downlink signals, the corresponding interference to TN uplink transmissions on the near-horizontal plane is relatively small. In addition, owing to the extended propagation distance and limited antenna gain associated with the satellite’s side lobes, TN uplink users located within the reception range of these side lobes exert minimal interference on the satellites’ spectrum sensing capabilities. In particular, spectrum sensing is conducted because satellites and NTN users can neither determine TN users transmitting in the current time slot nor predict aggregated interference.

II-C NTN Uplink and TN Downlink SS

When an NTN user transmits signals to a satellite, they can opportunistically access the shared spectrum with TN downlink communications, as shown in Fig. 2(3). In this situation, TN users are considered primary users, while NTN users are secondary users. This framework is feasible due to a couple of factors. First, the BS is tilted downward for TN downlink communications and employs beamforming techniques to serve TN users. Because of the BS’s antenna angle and beamforming, interference leakage from the BS to the satellite is minimal. Second, during NTN uplink communications, the NTN user and the satellite can collaboratively perform spectrum sensing. This allows them to opportunistically access the spectrum while minimizing interference to and from TN users. It is important to note that, based on spectrum sensing conducted by NTN users, the aggregated interference power generated by NTN users is relatively lower than the desired signal power of the TN downlink communication. This is due to the high transmit power associated with TN BS, which also employs beamforming techniques.

II-D NTN Uplink and TN Uplink SS

As shown in Fig. 2(4), allowing NTN uplink communications to share the spectrum with TN uplink communications is generally not advisable for several reasons. First, numerous TN uplink users operate within a satellite’s broad coverage area, and the distances between these users and NTN uplink users can be significant. This results in limited detection capabilities, making it difficult for either type of user to perceive one another effectively. Second, TN users typically have a limited number of antennas, leading to considerable interference leakage from TN users in the vertical direction. This interference can overwhelm the desired signals from NTN uplink communications. Third, NTN users are likely to experience a high rate of missed detections regarding the presence of TN users. Consequently, this inability to accurately sense TN users can lead to significant interference from NTN uplink transmissions to TN communications. As a result, TN users’ communication performance often falls short of expectations. Sharing the spectrum between these two co-directional uplink links is unfavorable from the perspective of interference and sensing capabilities in many situations. However, this spectrum-sharing framework may be viable with appropriate spectrum-aware strategies when the number of TN uplink users is relatively small.

III Advanced Approaches to SS in STINs

This section presents various approaches for achieving SS in STINs under different scenarios.

III-A Protection Zone-Based Approach

SS technique enables STIN users to sense and access the spectrum occupied by other users to improve the spectrum efficiency, and interference to the primary system can be tolerated when performance degradation is insignificant. However, such interference cannot be ignored when users are close to interference sources. Therefore, an auxiliary approach for SS is to define some geographical areas where users cannot access the spectrum of the primary system. Known as protection zones, these geographical regions are usually located around the primary users, which can isolate interference for the users inside them, as shown in Fig. 3(a). With protection zones and user locations in STINs, the SS technique enables intelligent monitoring of the spatial spectrum holes and shares the spectrum outside the protection zone. Since the NTN is mobile and provides flexible spectrum access for users, finding the boundary for the protection zone becomes a crucial and challenging issue. In addition, the dynamics of STIN architectures also have significant impacts on the protection zones, and we should consider them in the successful design of SS strategies.

III-B Spatial Spectrum Sensing-Based Approach

Spatial spectrum sensing is an approach to accessing the spectrum that relies on sensing rather than knowing the positions of interfering nodes [13]. The term “spatial” refers to the ability of this method to capture the spatial characteristics of large-scale networks. For a secondary user seeking to access a spectrum occupied by primary users, the process begins with evaluating the interference power. As shown in Fig. 3(b), if the test statistic for the interference power exceeds a predefined energy detection threshold, it indicates that the channel is currently busy. In this case, the secondary user transmits with a low probability. Conversely, if the interference power is below the threshold, the secondary user transmits with a high probability. It is essential to ensure that the false-alarm and miss-detection probabilities are controlled following the Neyman-Pearson criterion. When spatial spectrum sensing is implemented in STINs, NTN-related devices act as secondary users to monitor the busy status of the shared channel. By selecting an appropriate energy detection threshold, we can manage the interference generated by NTN toward TN.

III-C Joint Spatial Spectrum Sensing-Based Approach

Joint spatial spectrum sensing involves two different types of network nodes working together to perform spatial spectrum sensing. As shown in Fig. 3(c), if both nodes determine that the channel is available for transmission, the secondary user will likely access the channel with a high probability. Conversely, if they disagree on availability, the secondary user will have a lower probability of accessing the channel. This approach is particularly suitable for NTN uplink and TN downlink SS. The reason is that the satellite, acting as the NTN receiver, may experience significant interference from the TN, especially when there is a high density of ground base stations in specific areas. As a result, the spatial spectrum sensing performed by the satellite protects its communication from the severe interference generated by the TN. At the same time, the spatial spectrum sensing done by the NTN user helps manage the interference produced by the NTN towards the TN.

III-D Dynamic Spectrum Access-Based Approach

In STINs, users can sense and access the available spectrum as long as they are not causing significant interference to other users or the spectrum is idle, as shown in Fig. 3(d). Dynamic spectrum access (DSA) could be used to achieve this objective. Since the spectrum resource is scarce in STINs, DSA is a widely adopted SS approach that takes advantage of the temporal holes of the spectrum, and the spectrum efficiency and system rate can thus be significantly promoted. Since NTNs have high mobility, STIN users can adopt DSA strategies to access the NTN spectrum by integrating artificial intelligence (AI) tools. Similarly, NTNs can sense the spectrum licensed to TN users and adopt DSA to provide services for users. Hence, the DSA protocol is essential for the DSA-based approach in SS and has been designed to guarantee the performance of STINs in existing studies. The NTN and TN nodes can exchange spectrum information using a dedicated SS approach.

III-E Game Theory-Based Approach

Unlike centralized control, users tend to make their own decisions and compete with each other to access the shared spectrum. Generally, all users need to interact with each other to decide the optimal SS pattern. Recognized as a powerful theoretical approach, game theory models the interacting users as rational players aiming to maximize their own benefits by selling and buying spectrum resources, as shown in Fig. 3(e). Hence, the spectrum allocation and access scheme of the TN/NTN nodes and the users can be described by adopting precise game models, and the optimal spectrum utilization pattern can be obtained by deriving the Nash equilibrium. It can be noted that, since STINs are heterogeneous with multiple hierarchies, the internal coordination of various users in SS often exhibits considerably high complexity. In this sense, the game theory approach is considered one of the most competitive approaches for SS in STINs.

IV Performance Metrics of SS in STINs

In this section, we introduce key metrics for evaluating the performance of SS in STINs from different perspectives.

IV-A User Throughput and System Capacity of STINs

User throughput and system capacity refer to the amount of data that the user and the entire system can transmit and receive over a certain period, respectively. As a key performance metric, user throughput can be obtained by dividing a user’s data flow by the period, demonstrating the efficiency of STIN in continuously providing wireless services. The system capacity is the sum of user rates and a system-level performance metric, which denotes the ability of STIN to provide services in some geographical regions. Since STINs have dynamic architectures, a high and stable expected system capacity implies that STINs can provide wireless services with nice qualities.

IV-B Area Spectrum Efficiency of NTN and/or TN

Area spectrum efficiency (ASE) refers to the spectrum efficiency of a unit area. Since STINs consist of various network components with different densities, ASE can be used to evaluate the average performance of a large-scale network from a system-level perspective. For a specific network segment, the ASE can be calculated by multiplying the density of the network components by the average spectrum efficiency of the network. By determining the ASE of STINs, we can also assess their system capacity per unit of bandwidth.

IV-C Interference Intensity between NTN and TN

Interference intensity refers to the power strength per unit of bandwidth caused by secondary users to primary users, which denotes the negative impact of secondary users when accessing the shared spectrum in STINs. In SS, the users receive interference from the TN nodes during NTN downlink transmission, and the TN nodes receive interference from the users during NTN uplink transmission. Hence, a smaller interference intensity implies a higher efficiency and capacity of the STINs. Since the total interference power grows as the bandwidth increases, the STIN performance can be guaranteed by maintaining a satisfactory interference intensity.

IV-D Spectrum Access Probability of NTN User

Spectrum access probability (SAP) refers to the likelihood that a secondary user will access the shared spectrum when they want to transmit packets. This metric indicates how fairly the spectrum is utilized among secondary users. When a channel is detected as busy, the SAP is low, prompting secondary users to either attempt to re-access the spectrum later or to utilize their reserved spectrum. For secondary users, such as NTN users or satellites within STINs, we can analyze or simulate the SAP when they employ an SS approach. Note that SAP is typically included in the average data rate calculations.

IV-E End-to-End Latency in STINs

End-to-end (E2E) latency refers to the time a user receives a data packet across the network once the packet is triggered. This metric is essential in the SS of STINs, as it reflects the timeliness and fairness of packet transmission. In STINs, secondary users may need to delay their transmissions if they cannot access the spectrum in the current time slot. Additionally, when dealing with satellite-to-ground communication links, it is essential to consider propagation delay, which can add several milliseconds to the E2E latency. Generally, a smaller E2E latency indicates a quicker packet transmission.

IV-F Energy Efficiency of NTN and/or TN

Energy efficiency (EE) refers to the number of bits a user can transmit by consuming per unit of power. As a key performance metric, EE is obtained by dividing data throughput by the amount of power consumption, which denotes energy utilization in transmitting signals. Since NTN nodes and users in STINs often face limited energy supply restrictions, it is crucial to promote their EE to transmit more data and provide better QoS. Meanwhile, increasing EE also enables us to better utilize scarce spectrum resources. Thus, a good guarantee of EE is essential for the overall system performance of STINs.

V Case Study: SS in STINs Using Protection Zone-based Approach

In this section, we quantitatively study the effectiveness of SS in STINs by adopting the protection zone-based approach and compare it with STINs without SS. The simulation scenarios are first depicted in Fig. 4. The NTN downlink and TN downlink framework is considered, where the NTN users are the primary users and the TN users are the secondary users. A protection zone-based approach is adopted, so that the BSs are located at least $R_{\text{p}}$ away from NTN users. For the case of SS in Fig. 4(a), NTN users are primary users each having a protection zone with radius $R_{\text{p}}$ . An NTN user will use the reserved spectrum if there is a BS inside the protection zone and will use the shared spectrum if there is no BS inside. If SS is not adopted in STINs as shown in Fig. 4(b), NTN users and TN users will experience interference from the NTN spectrum and TN spectrum, respectively.

The simulation results are shown in Fig. 5. Both the system capacity and the average data rate of STINs without SS remain almost the same for different reserved NTN bandwidths $B_{\text{R}}$ in STINs with SS. Specifically, as shown in Fig. 5(a), the capacity of TN with SS first surpasses that without SS when $50\leq B_{\text{R}}\leq 110$ , but is exceeded by the latter subsequently. The capacity of NTN with SS is always higher than that without SS, and the sum capacity of STINs with SS continues to increase. In Fig. 5(b), the average data rate of the users in STINs with SS is compared to that of the users in STINs without SS. With an increase of $B_{\text{R}}$ , the data rate of both NTN users and TN users using shared spectrum decreases because more bandwidth is allocated to the reserved spectrum. However, the data rate of the NTN user using reserved spectrum increases. Therefore, it can be concluded that a win-win solution can be obtained by using SS within a certain range of reserved NTN bandwidth.

VI Future Opportunities of SS in STINs

In this section, we briefly discuss future applications of SS in STINs. An overview of these potential applications is presented in Fig. 6, and detailed descriptions of each of them are then given as follows.

VI-A SS for Multi-Tier NTN and TN

Due to the limited availability of orbital positions, satellites are deployed at various altitudes. Different types of satellites have varying antenna patterns and transmit powers, and the number of satellites in different tiers also varies. As a result, the single-tier STIN will evolve into a multi-tier heterogeneous STIN (HetSTIN). However, this multi-tier HetSTIN faces a significant challenge: the conflict between limited spectrum resources and the increasing demand for satellite services. Effectively sharing the spectrum among the multi-tier NTN and multi-tier TN is crucial for successfully integrating satellite and terrestrial networks. Addressing the complexities of interference management and the imbalanced communication resources that arise from the heterogeneous nature of multi-tier HetSTINs is essential in the design of SS.

VI-B SS for Multi-Connectivity-Based Cooperative STINs

Multi-connectivity in STINs indicates that a user can simultaneously connect to multiple network nodes, including BSs or access points (APs) in TNs and satellites in NTNs. Considering the complexity of the environment, cooperative communication, which includes relaying, joint transmission or reception, etc., can be utilized to connect different network tiers in STINs and promote more resilient and adaptable communications. However, the bandwidth allocated to a user, which may have multiple connection links, can be restricted due to the scarcity of available spectrum resources. Hence, allocating spectrum resources among multiple users, each possessing multiple links, presents significant research opportunities. In addition, the organization of cooperative network nodes, whether centralized or decentralized, requires further exploration and investigation.

VI-C Unlicensed SS in STINs

Unlicensed frequency bands offer a significant amount of high-bandwidth and low-frequency spectrum resources. However, users operating in these unlicensed bands, such as Wi-Fi, typically use the carrier sense multiple access with collision detection (CSMA/CD) medium access control (MAC) protocol to communicate in close proximity. As a result, these unlicensed bands are often underutilized in remote areas. To address this issue, satellite communications can opportunistically access unlicensed bands to maximize the use of available spectrum. It is essential to design modified MAC protocols that effectively achieve unlicensed SS in STINs. Additionally, the physical layer must adapt its frame structure to accommodate satellite mobility, the varying fading channels encountered in space and space-to-ground links, and the challenges of synchronization due to the extended propagation delays between satellites and ground users.

VI-D Multiple Access-Enabled Cognitive SS in STINs

Multiple access techniques can be incorporated into STINs due to their shared goals of improving spectrum utilization efficiency. Multiple access allows multiple users to be served simultaneously in the same frequency band. Using this characteristic, multiple access and cognitive SS can be connected by sensing the real-time radio environment. The cognitive radio can sense the surroundings and adjust the transmission parameters accordingly. With these sensing results, secondary users can assess the unutilized spectrum without interfering with primary users. In addition, some other benefits come with this fusion as well, including increased network capacity, more efficient and flexible resource allocation, and decreased interference. By continuously monitoring the radio environment and adjusting transmission parameters, more reliable network performance can be ensured, and a more equitable distribution of spectrum resources among users can be promoted.

VI-E SS for Different Satellite Constellations

As satellite constellations have expanded to mega size in recent years, the number of different satellite constellations built by different satellite companies is also increasing. Satellite constellations can also share their proprietary spectrum resources to increase the system’s capacity. Due to the mobile nature of the satellites, the NTN architecture is always dynamic. It changes rapidly, leading to the significant challenge of maintaining reliable and high-quality services under the restriction of limited and congested spectrum resources. Different satellite constellations face their transmission conditions as service providers. They require compatible spectrum access strategies. SS techniques enable the high-efficiency utilization of spectrum resources while introducing controllable interference. Therefore, identifying efficient SS configurations is crucial for various satellite constellations to mitigate interference and maintain service quality. Meanwhile, to meet the constellation dynamics, the designed SS strategy should also be adaptive to available satellites and spectrum resources from time to time in providing wireless services.

VI-F SS for Space-Air-Ground Integrated Networks (SAGINs)

Due to the rapid growth of aircraft and unmanned aerial vehicles (UAVs), STINs can be significantly extended by incorporating these users as the aerial network nodes to evolve towards the architecture of space-air-ground integrated networks (SAGINs). Compared to satellites, aerial nodes are more flexible in SAGIN architectures and much closer to the users. Therefore, SS is also a critical issue for aerial nodes, as these nodes can promptly monitor the spectrum dynamics, detect spatial and temporal holes, and provide corresponding wireless services for SAGIN users under well-designed SS strategies. Meanwhile, aerial nodes can act as interims between space and ground by forwarding messages, offloading, and balancing tasks among different regions. These tasks also require SS strategy designs to coordinate the utilization of spectrum resources among various SAGIN users.

VII Conclusions

This article investigated the frameworks, approaches, performance metrics, and future research opportunities for SS in STINs. We introduced four general SS frameworks in STINs, including different uplink and downlink combinations for NTN and TN, and discussed their characteristics. Then, state-of-the-art approaches for SS to reuse spectrum resources in space, time, and frequency domains in STINs were comprehensively reviewed. More importantly, related metrics were presented to study the SS techniques further and evaluate the performance of the system and the individual user. A case study on the protection zone-based approach for SS in STINs was carried out, and the simulation results showed that a win-win solution could be obtained for improved system capacity. Finally, we provided promising opportunities and insightful future research on SS in STINs.

References

[1] H. Yao, L. Wang, X. Wang, Z. Lu, and Y. Liu, “The space-terrestrial integrated network: An overview,” IEEE Commun. Mag., vol. 56, no. 9, pp. 178–185, 2018.
[2] H.-C. Chao, D. E. Comer, and O. Kao, “Space and terrestrial integrated networks: Emerging research advances, prospects, and challenges,” IEEE Netw., vol. 33, no. 1, pp. 6–7, 2019.
[3] Z. Li, S. Han, M. Peng, C. Li, and W. Meng, “Dynamic multiple access based on RSMA and spectrum sharing for integrated satellite-terrestrial networks,” IEEE Trans. Wireless Commun., vol. 23, no. 6, pp. 5393–5408, 2023.
[4] Z. Wei, L. Wang, Z. Gao, H. Wu, N. Zhang, K. Han, and Z. Feng, “Spectrum sharing between high altitude platform network and terrestrial network: Modeling and performance analysis,” IEEE Trans. Commun., vol. 71, no. 6, pp. 3736–3751, 2023.
[5] Q. Chen, W. Meng, S. Han, C. Li, and T. Q. Quek, “Coverage analysis of sagin with sectorized beam pattern under shadowed-Rician fading channels,” IEEE Trans. Commun., vol. 71, no. 8, pp. 4988–5004, 2023.
[6] C. Zhang, C. Jiang, L. Kuang, J. Jin, Y. He, and Z. Han, “Spatial spectrum sharing for satellite and terrestrial communication networks,” IEEE Trans. Aerosp. Electron. Syst., vol. 55, no. 3, pp. 1075–1089, 2019.
[7] M. J. Anjum, T. Anees, F. Tariq, M. Shaheen, S. Amjad, F. Iftikhar, and F. Ahmad, “Space-air-ground integrated network for disaster management: Systematic literature review,” Appl. Comput. Intell. Soft Comput., vol. 2023, no. 1, p. 6037882, 2023.
[8] K. Lin, D. Wang, L. Hu, M. S. Hossain, and G. Muhammad, “Virtualized QoS-driven spectrum allocation in space-terrestrial integrated networks,” IEEE Netw., vol. 33, no. 1, pp. 58–63, 2019.
[9] B. Shang, V. Marojevic, Y. Yi, A. S. Abdalla, and L. Liu, “Spectrum sharing for UAV communications: Spatial spectrum sensing and open issues,” IEEE Veh. Technol. Mag., vol. 15, no. 2, pp. 104–112, 2020.
[10] F. Tang, L. Chen, X. Li, L. T. Yang, and L. Fu, “Intelligent spectrum assignment based on dynamical cooperation for 5G-satellite integrated networks,” IEEE Trans. Cogn. Commun. Netw., vol. 6, no. 2, pp. 523–533, 2020.
[11] F. Li, K.-Y. Lam, H.-H. Chen, and N. Zhao, “Spectral efficiency enhancement in satellite mobile communications: A game-theoretical approach,” IEEE Wireless Commun., vol. 27, no. 1, pp. 200–205, 2019.
[12] W. Li, L. Jia, Q. Chen, and Y. Chen, “A game theory-based distributed downlink spectrum sharing method in large-scale hybrid satellite constellations,” IEEE Trans. Commun., vol. 72, no. 8, pp. 4620–4632, 2024.
[13] B. Shang, L. Liu, H. Chen, C. J. Zhang, S. Pudlewski, E. S. Bentley, and J. D. Ashdown, “Spatial spectrum sensing in uplink two-tier user-centric deployed hetnets,” IEEE Transactions on Wireless Communications, vol. 19, no. 12, pp. 7957–7972, 2020.

Biographies

Bodong Shang ([email protected]) received his M.S. degree from Xidian University, China, and his Ph.D. degree from the Department of Electrical and Computer Engineering at Virginia Tech, Blacksburg, USA, in 2021, and he was a Postdoctoral Research Associate at Carnegie Mellon University, Pittsburgh, USA. He held a research internship position at Nokia Bell Labs, USA. He is an Assistant Professor at Eastern Institute for Advanced Study, Eastern Institute of Technology (EIT), Ningbo, China. His research interests include space-air-ground-sea integrated networks, non-terrestrial networks, and space information networks.

Zheng Wang ([email protected]) received his Ph.D. degree from the Department of Electrical and Computer Engineering at George Mason University, Fairfax, USA, in 2020, and he was a Postdoctoral Research Associate at Virginia Tech, Blacksburg, USA. He is a Research Associate Professor at Ningbo Institute of Digital Twin, Eastern Institute of Technology (EIT), Ningbo, China. His research areas include satellite-terrestrial integrated networks, game theory, and physical layer communication.

Xiangyu Li ([email protected]) received an M.S. degree in Electrical and Computer Engineering from Georgia Institute of Technology, Atlanta, USA, in 2023, and he is currently pursuing a Ph.D. degree with Shanghai Jiao Tong University (SJTU), Shanghai, China, at the Eastern Institute of Technology (EIT), Ningbo and SJTU Joint PhD Program. His research interests include space-air-ground integrated networks, non-terrestrial networks, and performance analysis of wireless systems.

Chungang Yang ([email protected]) Chungang Yang is a full professor at Xidian University, leading the GUIDE: Game, Utility, Artificial Intelligent Design for Emerging Communications research team. His research interests include artificial intelligence 6G wireless mobile networks, intent-driven networks, space-terrestrial networks, and game theory for emerging communication networks.

Chao Ren ([email protected]) Chao Ren received a Ph.D. from Xidian University, China, in 2017. He is an Associate Professor at the University of Science and Technology in Beijing, China. He served as a Review Editor for Frontiers in Computer Science, the Co-Chairs for Ucom23-24 WKSP, a Guest Editor for Space-Integrated-Ground Information Network, and a TPC Member for IEEE ICC and Globecom.

Haijun Zhang ([email protected]) Haijun Zhang is a Full Professor and Vice Dean at the University of Science and Technology in Beijing, China. He serves/served as an Editor of IEEE Transactions on Information Forensics and Security, IEEE Transactions on Communications, IEEE Transactions on Network Science and Engineering, and IEEE Transactions on Vehicular Technology. He received the IEEE CSIM Technical Committee Best Journal Paper Award in 2018, the IEEE ComSoc Young Author Best Paper Award in 2017, and the IEEE ComSoc Asia-Pacific Best Young Researcher Award in 2019. He is a Fellow of IEEE.