Cost-effective hybrid satellite-terrestrial networks : Optimal beamforming with nonlinear PA and large-scale CSIT

Chengxiao Liu, Wei Feng, Yunfei Chen, Cheng‐Xiang Wang, Ning GearXiv 2019 · arXiv

Abstract

In hybrid satellite-terrestrial networks (HSTNs), spectrum sharing is crucial to alleviate the “spectrum scarcity” problem. Therein, the transmit beams should be carefully designed to mitigate the inter-satellite-terrestrial interference. Different from previous studies, this work considers the impact of both nonlinear power amplifier (PA) and large-scale channel state information at the transmitter (CSIT) on beamforming. These phenomena are usually inevitable in a cost-effective practical HSTN. Based on the Saleh model of PA nonlinearity and the large-scale multi-beam satellite channel parameters, we formulate a beamforming optimization problem to maximize the achievable rate of the satellite system while ensuring that the inter-satellite-terrestrial interference is below a given threshold. The optimal amplitude and phase of desired beams are derived in a decoupled manner. Simulation results demonstrate the superiority of the proposed beamforming scheme.

Introduction: The Challenge of Spectrum Sharing in HSTNs

As the demand for wireless connectivity grows, Hybrid Satellite-Terrestrial Networks (HSTNs) have emerged as a promising solution to "spectrum scarcity" by allowing satellite and terrestrial networks to share the same frequency bands. However, this coexistence introduces significant challenges, primarily the management of inter-system interference. To mitigate this, satellites employ beamforming—a technique to shape signal transmission toward intended targets while minimizing leakage to terrestrial users.

System model of a Hybrid Satellite-Terrestrial Network (HSTN) illustrating interference links and nonlinear power amplifiers.

Most existing research assumes idealized conditions: perfect, real-time Channel State Information at the Transmitter (CSIT) and perfectly linear hardware. In reality, satellites often only have access to "large-scale" CSIT (slowly-varying data like location and path loss) due to the high cost and latency of real-time updates. Furthermore, to maintain cost-effectiveness, satellites utilize Power Amplifiers (PAs) that are inherently nonlinear, causing signal distortion when pushed toward saturation. This paper addresses the design of optimal beams under these realistic constraints.

Methodology: Decoupling Complexity

The researchers modeled PA nonlinearity using the Saleh model, which describes how input amplitude and phase are distorted at the output. They also employed a composite multi-beam satellite channel model that separates stable large-scale parameters from volatile small-scale atmospheric phase shifts.

The resulting optimization problem is "non-convex," making it difficult to solve directly. The authors' approach involves three key steps:

  1. Decoupling: Proving that the optimal phase and amplitude of the beams can be determined independently.
  2. Feasible Region Reduction: Narrowing the search space based on the physical saturation points of the PAs.
  3. Variable Substitution: Recasting the non-convex problem into a convex optimization problem, allowing for efficient solution using standard mathematical tools.

Key Insights: Navigating the Saturation Trap

A primary finding is the "saturation trap." Traditional beamformers often maximize gain by increasing power to antennas with the best channels. With nonlinear PAs, this strategy backfires; once an amplifier hits its saturation point, additional power only increases distortion, actually decreasing overall performance.

The authors also demonstrated that Large-Scale CSIT is sufficient. By correctly accounting for nonlinearity, designing for "average" channel conditions proves highly effective, removing the need for costly millisecond-by-millisecond updates.

Results and Performance

Simulations with a 16-antenna satellite system showed that the proposed algorithm consistently outperforms Maximum Ratio Transmission (MRT) and other state-of-the-art hybrid beamforming methods.

Achievable rate comparison showing the proposed algorithm's robustness against PA saturation at high input power limits.

As shown in the results, the performance gap widens as the input power limit increases. While competing methods "choke" or plateau as they hit PA saturation, the proposed algorithm adapts its beam weights to remain within the hardware's efficient operating range. Additionally, the algorithm successfully maintains interference to terrestrial users below critical thresholds (e.g., -107 dBm) while maximizing satellite throughput.

Conclusion and Future Directions

This work provides a robust framework for beamforming in practical, cost-constrained HSTNs. While the current study focuses on a single interfering link, future research could extend this to multi-user scenarios and more complex amplifier models that account for memory effects. Nonetheless, the ability to achieve high data rates using large-scale CSIT and nonlinear hardware marks a significant step toward viable, large-scale hybrid networks.

Summary updated

Abstract

This thesis explores a new form of urban public place—multifunctional venues called ‘pay-per-minute cafes’, ‘public living rooms’, or ‘anti-cafes’. Charged by the minute and provided with free wifi and access to kitchen facilities, visitors of such spaces are entitled to use them however they like, as they are designed to accommodate various social, cultural, home-from-home and work activities. The first venue of this kind, Ziferblat, opened in 2011 in Moscow as a social experiment seeking to build ‘social media in real life’, turn customers into participants and overcome the limitations of the functionalist urban planning separating home, work and leisure from each other. In the next couple of years, Ziferblat’s look-alikes have spread overseas to Europe, Asia and North America; meanwhile, Ziferblat itself has developed into an international franchise with 18 branches, five of which are located in the UK. Using this phenomenon as a lens on two emerging urban trends—the post-functionalist city and the post-digital city—this thesis investigates the ‘who’, ‘why’, and ‘how’ of placemaking in the context of current debates on sociability, diversity and social inclusion in the urban public space, participatory culture and sharing economy, and neoliberal urban policy. This critical case study, drawing on large-scale media analysis, 48 qualitative interviews and over 160 hours of ethnography conducted in four Ziferblat branches in Moscow, London and Manchester, employs an interdisciplinary theoretical framework, combining urban sociology, human geography, cultural anthropology, media studies, and consumer and service research to examine the intricate connections and contradictions between the social and the spatial, the global and the local, the social and the commercial, the public and the private, the physical and the digital. The research findings are of potential interest to academics, practitioners, social activists, urban planners and policymakers dealing with the issues of placemaking and community building in the city. In hybrid satellite-terrestrial networks (HSTNs), spectrum sharing is crucial to alleviate the “spectrum scarcity” problem. Therein, the transmit beams should be carefully designed to mitigate the inter-satellite-terrestrial interference. Different from previous studies, this work considers the impact of both nonlinear power amplifier (PA) and large-scale channel state information at the transmitter (CSIT) on beamforming. These phenomena are usually inevitable in a cost-effective practical HSTN. Based on the Saleh model of PA nonlinearity and the large-scale multi-beam satellite channel parameters, we formulate a beamforming optimization problem to maximize the achievable rate of the satellite system while ensuring that the inter-satelliteterrestrial interference is below a given threshold. The optimal amplitude and phase of desired beams are derived in a decoupled manner. Simulation results demonstrate the superiority of the proposed beamforming scheme.
BibTeX
@misc{liu2019cost,
  author = {Chengxiao Liu and Wei Feng and Yunfei Chen and Cheng‐Xiang Wang and Ning Ge},
  title = {{Cost-effective hybrid satellite-terrestrial networks : Optimal beamforming with nonlinear PA and large-scale CSIT}},
  year = {2019},
  eprint = {1905.01654},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/1905.01654},
}
Figure