The ever-increasing demand for ubiquitous and differentiated services at anytime and anywhere emphasizes the necessity of aerospace integrated networks (AINs) which consist of multi-tiers, e.g., high-altitude platforms (HAPs) tier, unmanned aerial vehicles (UAVs) tier, low earth orbit (LEO) satellites tier, etc. for wide coverage and high capacity. However, the inherent heterogeneity, time-variability, and differentiated mobility characteristics between terrestrial networks and AINs constraint the materialization of the aforementioned widely expected potentials of AINs. Therefore, it is crucial to make a profound understanding of AINsimpacts on pivotal models and methodologies for enabling 6G, thus providing technology reference for AINs empowering 6G. To understand the latest development and ultimately open new research niches on this significant topic, this survey is the pioneer paper to serve as a systematical and comprehensive overview in both single-tier scenarios and combining-tiers scenarios in AINs. We start with a profound discussion about the state-of-the-art potentially promising methodologies and models in the era of AINs empowering 6G from the perspective of the system architecture and networking design and enabling technologies of AINs, to enable flexible and scalable management and control of the AINs and ensure network stability. Furthermore, we make an in-depth literature overview across the network dynamics modeling, theoretical performance analysis model, and system optimization to enhance the system performance and guarantee user-on-demand services. Additionally, we highlight the ongoing research challenges and future directions with focusing on the development trend of ultra-dense satellite constellations in 6G.

Space-air-ground networks play important roles in both fifth generation (5G) and sixth generation (6G) techniques. Low earth orbit (LEO) satellites and high altitude platforms (HAPs) are key components in space-air-ground networks to provide access services for the massive mobile and Internet of Things (IoT) users, especially in remote areas short of ground base station coverage. LEO satellite networks provide global coverage, while HAPs provide terrestrial users with closer, stable massive access service. In this work, we consider the cooperation of LEO satellites and HAPs for the massive access and data backhaul of remote area users. The problem is formulated to maximize the revenue in LEO satellites, which is in the form of mixed integer nonlinear programming. Since finding the optimal solution by exhaustive search is extremely complicated with a large scale of network, we propose a satellite-oriented restricted three-sided matching algorithm to deal with the matching among users, HAPs, and satellites. Furthermore, to tackle the dynamic connections between satellites and HAPs caused by the periodic motion of satellites, we present a two-tier matching algorithm, composed of the Gale-Shapley-based matching algorithm between users and HAPs, and the random path to pairwise-stable matching algorithm between HAPs and satellites. Numerical results show the effectiveness of the proposed algorithms.

Satellite networks have been regarded as a promising architecture for supporting the Internet of remote things (IoRT) due to their advantages of wide coverage and high communication capacity in remote areas, which further promotes the development of the satellites for IoRT networks (SIoRTNs). The effectiveness of multi-dimensional resource collaboration has significant impacts on the IoRT data downloading performance. However, the environment ’ s dynamics, e.g., channel conditions and solar infeed process, are unknown in practical scenarios, which poses daunting challenges in making efficient utilization of multi-dimensional resources. Motivated by this fact, we model the joint resource scheduling and IoRT data scheduling problem with the aim of maximizing the amount of the IoRT data of the overall network by applying the model-free reinforcement learning framework. To overcome the limitations of traditional reinforcement learning algorithms, we propose several feature functions by investigating the natural attributes of the multi-dimensional resources of the SIoRTNs, and further exploit the concept of function approximation to approximate the expected downloaded IoRT data given the network state. Furthermore, we propose a state-action-reward-state-action (SARSA) based actor-critic reinforcement learning (SACRL) resource allocation strategy to achieve the optimal resource allocation and IoRT data scheduling with casual information at LEO satellites. Simulations validate the convergence property and the effectiveness of the proposed SACRL algorithm in terms of the amount of the downloaded IoRT data. Particularly, we investigate the impact of typical network parameters on network performance to further provide guidance for future SIoRTN system design.

Low earth orbit (LEO) satellite networks will play important roles in the sixth generation (6G) communication system. Software defined network technique is a novel approach introduced to the LEO satellite networks to improve the resource flexibility and efficiency, forming the software defined LEO satellite networks (SDLSNs). How to efficiently allocate the resources of SDLSN to provide services for the terrestrial users is a key issue. Hence, in this work, we explore the service provision for SDLSN via virtual network functions (VNFs) orchestration on the software defined time-evolving graph. In view of the scarce, intermittent and unstable satellite-to-satellite (S2S) links, the problem is formulated to minimize the S2S resource consumption while satisfying the terrestrial tasks, which is in the form of integer linear programming. Since the problem is intractable by exhaustive search, we design a branch-and-price algorithm based on the coupling of Dantzig-Wolfe decomposition, column generation, and branch-and-bound to efficiently acquire the optimal solution. Further, to obtain a faster solution for practical usage, we further design an approximation algorithm for the subproblem and leverage the beam search to accelerate the pruning for the search tree. Finally, extensive simulations are conducted and the numerical results validate the effectiveness of the proposed schemes.

The network control plays a vital role in the mega satellite constellation (MSC) to coordinate massive network nodes to ensure the effectiveness and reliability of operations and services for future space wireless communications networks. One of the critical issues in satellite network control is how to design an optimal network control structure (ONCS) by configuring the least number of controllers to achieve efficient control interaction within a limited number of hops. Considering the wide coverage, rising capacity, and no geographical constraints of space platforms, this paper contributes to designing the ONCS by constructing an optimal space control network (SCN) to improve the temporal effectiveness of network control. Specifically, we formulate the optimal SCN construction problem from the perspective of satellite coverage factors, and apply geometric topology analysis to derive both the conditions for constructing the optimal SCN and the formulaic conclusions for SCN and MSC configurations (i.e., scale and structure). From numerical results, we investigate the tradeoff between network scale, the number of controllers, and control delays in several satellite network control scenarios, to provide guidelines for the MSC control. We also design the optimal SCN for an existing MSC system to demonstrate the effectiveness of the proposed ONCS.