Advancing Multi-Connectivity in Satellite-Terrestrial Integrated Networks: Architectures, Challenges, and Applications

TL;DR

The paper investigates multi-connectivity in satellite-terrestrial networks to enhance coverage, reliability, and efficiency in NTN-TN integration. It introduces three deployment architectures—MS, SS-SBS, and MS-MBS—and analyzes their operation, use cases, and trade-offs. It details system design challenges including CF/multi-tier satellite networking, coordinated beamforming, CSI estimation, and synchronization, with a case study showing coverage gains over single-connectivity. It concludes with future research directions on interference management, massive access, and spectrum sharing to guide real-world deployment of MC in STINs.

Abstract

Multi-connectivity (MC) in satellite-terrestrial integrated networks (STINs), included in the Third-Generation Partnership Project (3GPP) standards, is regarded as a promising technology for future networks, especially the non-terrestrial network (NTN). The significant advantages of MC in improving coverage, communication, and sensing through satellite-terrestrial collaboration have sparked widespread interest. This article introduces three fundamental deployment architectures of MC systems in STINs, including multi-satellite, single-satellite single-base-station, and multi-satellite multi-base-station configurations. Considering the emerging but still evolving satellite networking, we explore system design challenges such as satellite networking schemes, such as cell-free and multi-tier satellite networks. Subsequently, key technical challenges severely influencing the quality of mutual communications, including beamforming, channel estimation, and synchronization, are discussed. Furthermore, typical applications such as coverage enhancement, traffic offloading, collaborative sensing, and low-altitude communication are demonstrated, followed by a case study comparing coverage performance in MC and single-connectivity (SC) configurations. Several essential future research directions for MC in STINs are presented to facilitate further exploration.

Figures (6)

• Figure 1: An overview of MC in STINs.
• Figure 2: Deployment architectures of MC in STINs.
• Figure 3: Coverage probability versus SINR thresholds with different Nakagami-$m$ parameters. The satellites are distributed with the density $\lambda_S=1\times 10^{-5}$/km$^2$ at an altitude of $H_S=500$ km. The transmit power, antenna gain, and pathloss exponent are $\rho_d=50$ dBm, $G_t=30$ dBi, and $\alpha_N\approx 2.0$, respectively. The UEs are distributed on the earth surface with the density $\lambda_U=4\times 10^{-6}$/km$^2$, and the receive antenna gain is $G_r=10$ dBi. The cases of beamforming and non-beamforming are compared.
• Figure 4: Discrepancy between CDF of the signal received in the ideal channel and that in the estimated channels, with different numbers of pilots. The simulation parameters are the same as those in Fig. \ref{['fig:BF']}, except that $m=2$.
• Figure 5: A sketch of application scenarios of MC in STINs.