Commercial Dishes Can Be My Ladder: Sustainable and Collaborative Data Offloading in LEO Satellite Networks

Abstract

Low Earth Orbit (LEO) satellite networks, characterized by their high data throughput and low latency, have gained significant interest from both industry and academia. Routing data efficiently within these networks is essential for maintaining a high quality of service. However, current routing strategies, such as bent-pipe and inter-satellite link (ISL) routing, have their unique challenges. The bent-pipe strategy requires a dense deployment of dedicated ground stations, while the ISLbased strategy can negatively impact satellite battery lifespan due to increased traffic load, leading to sustainability issues.

In this paper, we propose sustainable collaborative offloading, a framework that orchestrates groups of existing commercial resources like ground stations and 5G base stations for data offloading. This orchestration enhances total capacity, overcoming the limitations of a single resource. We propose the collaborator group set construction algorithm to construct candidate groups and the collaborator selection and total payment algorithm to select offloading targets and determine payments no less than the costs. Extensive real-world-based simulations show that our solution significantly improves energy consumption, satellite service life, and end-to-end latency.

低地球轨道（LEO）卫星网络因其高数据吞吐量和低延迟的特性，引起了工业界和学术界的广泛关注。网络中数据的有效路由对于维持高质量服务至关重要。然而，现有的路由策略，如“弯管”（bent-pipe）转发和星间链路（ISL）路由，各有其独特的挑战。“弯管”策略需要密集部署专用的地面站，而基于星间链路的策略会因增加的通信负载而对卫星电池寿命产生负面影响，从而引发可持续性问题。

本文提出了一种名为“可持续协同卸载”的框架，该框架能够整合并调度如地面站和5G基站等现有商用资源群组以进行数据卸载。这种协同调度机制能够增强系统总容量，克服了单一资源能力受限的问题。我们设计了协作者群组集合构建算法，用以生成候选的资源群组；同时提出了协作者选择与总支付算法，用于选定最终的卸载目标并确定不低于其成本的支付费用。基于真实世界数据的大量仿真实验表明，我们提出的解决方案在能耗、卫星服务寿命以及端到端延迟方面均有显著改善。

Introduction

Low Earth Orbit (LEO) satellite networks have attracted great interest from industry and academia due to their high throughput, low latency, and relatively low launch cost characteristics due to technological advancement [1]. LEO satellite networks are also believed to be an important component in the 6G network [2] since they can offer larger service coverage at less cost compared to terrestrial networks, providing truly ubiquitous Internet access services on Earth [3].

An efficient routing strategy is significant for delivering high-quality services to users. LEO satellite networks currently use two main data routing strategies to reach ground destinations [4]: 1) bent-pipe routing and 2) inter-satellite link (ISL)based routing. Each strategy, however, has unique challenges. In the bent-pipe strategy, each LEO satellite relays data between locations within the same coverage area, which requires a dense deployment of ground stations [5], [6]. Conversely, the ISL-based strategy routes data between satellites using ISLs, which increases the traffic load on LEO satellites and can negatively impact their battery lifespan, leading to congestion and sustainability issues [7].

In this paper, we argue that using solely these strategies is not the best for routing data in LEO satellite networks. We systematically study and answer the question: is there a cost-effective strategy that can utilize existing resources to reduce the need for new dedicated ground stations while jointly enhancing the sustainability and quality of service (QoS) in LEO satellite networks?

One potential solution is to utilize existing commercial shared resources, such as commercial ground stations and 5G base stations, rather than constructing new dedicated ground stations. There exist numerous commercial shared resources, such as AWS Ground Station and Azure Orbital Ground Station [8], [9], and 5G base stations equipped with 5G New Radio [10], that are capable of receiving data from satellites and offering Ground Station as a Service (GSaaS). As the satellite market expands, there will be more commercial shared resources offering GSaaS, similar to the Infrastructure as a Service model in cloud computing. For brevity, we refer to these shared resources as commercial dishes in this paper. Utilizing these commercial dishes can significantly reduce the deployment costs for LEO satellite operators.

However, existing GSaaS offerings have high variation in terms of QoS, making it difficult to select the one that can maximize gains while minimizing costs. Furthermore, different commercial dishes have different Service Level Agreements with varying QoS and uptime guarantees, and a single commercial dish may not satisfy the requirements. Hence, selecting the group of commercial dishes under the satellite coverage that offers the most gains while maintaining high QoS is also a challenge.

To address these challenges, we propose a sustainable and cost-effective offloading framework named Sustainable Collaborative Offloading (SusCO). In SusCO, LEO satellite networks can offload data to a group of commercial dishes, mitigating congestion in satellite hops and prolonging LEO satellite service life without the need to construct new dedicated ground stations. The SusCO orchestrates the collaboration among commercial dishes to enhance the total capacity [5] such as higher bandwidth, lower end-to-end latency, and receiving more offloading data.

Fig. 1 illustrates how SusCO works. The user terminal generates and transmits data to the LEO satellite A, where the data can be routed on the original path (green dash line) but the LEO satellite C on the original path has a low battery level, where the energy consumption has a more negative impact on the battery lifespan for low battery levels [11]. On the other hand, the data can be offloaded from the LEO satellite B to a group of collaborative commercial dishes (red dash arrows) before the LEO satellite C and then further routed to the destination through terrestrial networks, which can reduce congestion [12] and significantly prolong the battery lifespan for the LEO satellite C. However, some questions must be carefully addressed before the LEO satellite networks can fully utilize the existing commercial dishes in SusCO:

1) How do we quantify the improvement in energy consumption, end-to-end latency, and satellite service life from offloading data to commercial dishes?

2) How do we select a group of collaborative commercial dishes to enhance the total capacity while maintaining high service QoS?

3) How do we design an incentive scheme for commercial dish and LEO satellite operators to participate in SusCO?

Existing studies focused on in-space data routing to address the sustainability issues [7], [11], [13] and improve QoS [14]. Different from their work, we focus on utilizing existing commercial dishes and provide incentives to compensate the cost for commercial dish providers. Some other studies focused on the integration of space and terrestrial networks [15]–

[18], while we consider the sustainability issues for LEO satellite networks and incentives for the existing commercial dish providers in terrestrial networks to participate.

To the best of our knowledge, we are the first to systematically study the problem of offloading data from LEO satellites to existing commercial dishes while considering orchestrating collaboration, maintaining high service QoS, and incentives for dish providers and LEO satellite operators. The contributions of this paper are summarized below:

• We propose a novel sustainable and cost-effective offloading framework named Sustainable Collaborative Offloading (SusCO), which orchestrates the collaboration between commercial dishes to receive offloading data from LEO satellite networks.

• We propose two algorithms, collaborator group set construction algorithm and collaborator selection and total payment algorithm, to construct candidate groups of commercial dishes for higher capacity and determine payments for incentives, respectively.

• We perform extensive simulations based on real-world settings. The results show that our SusCO reduces 11.36% to 26.07% more energy consumption, 11.15% to 26.75% more life consumption, and 8.37% to 32.77% more end-to-end latency on average, compared to the other three state-of-the-art schemes.

The remainder of this paper is organized as follows: we discuss background and motivation in Section II and SusCO framework and models in Section III. The problem formulation and solution are presented in Section IV, followed by the performance evaluation in Section VI. The related work is discussed in Section VII, future work is discussed in Section VIII, and the conclusion is in Section IX.

由于技术进步，低地球轨道（LEO）卫星网络因其高吞吐量、低延迟和相对较低的发射成本等特性，已吸引了工业界和学术界的极大兴趣[1]。LEO卫星网络也被认为是6G网络的重要组成部分[2]，因为与地面网络相比，它能以更低的成本提供更广泛的服务覆盖，从而在地球上实现真正无处不在的互联网接入服务[3]。

高效的路由策略对于向用户提供高质量服务至关重要。当前，LEO卫星网络主要采用两种数据路由策略到达地面目的地[4]：

1）“弯管”（bent-pipe）路由

2）基于星间链路（ISL）的路由

然而，每种策略都面临其独特的挑战。在“弯管”策略中，每颗LEO卫星仅在其覆盖区域内的不同位置间中继数据，这要求密集部署地面站[5], [6]。相反，基于ISL的策略利用星间链路在卫星之间路由数据，这增加了LEO卫星的通信负载，可能对其电池寿命产生负面影响，从而导致拥塞和可持续性问题[7]。

在本文中，我们主张，在LEO卫星网络中仅依赖上述任一策略并非数据路由的最佳选择。我们系统性地研究并回答了以下问题：是否存在一种经济高效的策略，能够 利用现有资源来减少对新建专用地面站的需求 ，同时共同提升LEO卫星网络的可持续性与服务质量（QoS）？

一个潜在的解决方案是利用现有的商用共享资源，例如商用地面站和5G基站，而非建设新的专用地面站。目前已存在大量商用共享资源，如AWS地面站（AWS Ground Station）和Azure轨道地面站（Azure Orbital Ground Station）[8], [9]，以及配备了5G新空口（New Radio）技术的5G基站[10]，它们均有能力从卫星接收数据并提供“地面站即服务”（GSaaS）。随着卫星市场的扩张，未来将出现更多提供GSaaS的商用共享资源，这类似于云计算中的“基础设施即服务”（IaaS）模式。为简洁起见，本文将这些共享资源统称为“商用地面设施”（commercial dishes）。利用这些商用地面设施可以显著降低LEO卫星运营商的部署成本。

然而，现有的GSaaS服务在服务质量（QoS）方面存在巨大差异，这使得选择能够最大化收益同时最小化成本的服务变得困难。此外，不同的商用地面设施拥有不同的服务等级协议（SLA），其QoS和正常运行时间保证各不相同，单个商用地面设施可能无法满足需求。因此，如何在卫星覆盖范围内选择一组能够提供最大增益同时保持高QoS的商用地面设施组合，也是一个挑战。

为应对这些挑战，我们提出了一个名为“可持续协同卸载”（Sustainable Collaborative Offloading, SusCO）的可持续且经济高效的卸载框架。在SusCO框架中，LEO卫星网络可以将数据卸载到一个商用地面设施群组，从而在无需建设新的专用地面站的情况下，缓解卫星链路的拥塞并延长LEO卫星的服务寿命。SusCO框架能够协同调度这些商用地面设施间的合作，以增强系统的总容量[5]，例如获得更高的带宽、更低的点到点延迟以及接收更多的卸载数据。

图1阐释了SusCO的工作原理。用户终端生成数据并将其传输至LEO卫星A。数据可以沿原始路径（绿色虚线）传输，但原始路径上的LEO卫星C处于低电量水平，此时能量消耗对其电池寿命的负面影响会更大[11]。另一方面，数据可以在到达LEO卫星C之前，从LEO卫星B卸载到一个协同工作的商用地面设施群组（红色虚线箭头），然后通过地面网络路由至最终目的地。

这一方式能够减少拥塞[12]，并显著延长LEO卫星C的电池寿命。然而，在LEO卫星网络能够于SusCO框架中充分利用现有商用地面设施之前，必须审慎解决以下几个问题：

1) 我们如何量化将数据卸载至商用地面设施在能耗、端到端延迟和卫星服务寿命方面的改进？ 2) 我们如何选择一个协同工作的商用地面设施群组，以在提升总容量的同时保持高质量的服务（QoS）？ 3) 我们如何为商用地面设施和LEO卫星运营商设计一种激励机制，以鼓励他们参与SusCO框架？

现有研究主要集中于利用空间内部数据路由来解决可持续性问题[7], [11], [13]和提升服务质量[14]。与他们的工作不同，我们专注于利用现有的商用地面设施，并提供激励措施来补偿商用地面设施提供商的成本。其他一些研究则聚焦于天地一体化网络[15]–[18]，而我们则同时考虑了LEO卫星网络的可持续性问题以及激励地面网络中现有商用地面设施提供商参与的机制。

据我们所知，我们是首个系统性研究 将数据从LEO卫星卸载至现有商用地面设施 问题的团队，并同时考虑了协同调度、维持高质量服务以及对地面设施提供商和LEO卫星运营商的激励机制。本文的贡献总结如下：

• 我们提出了一个名为“可持续协同卸载”（SusCO）的新型、可持续且经济高效的卸载框架，该框架通过协同调度商用地面设施间的合作来接收来自LEO卫星网络的卸载数据
• 我们提出了两种算法：协作者群组集合构建算法和协作者选择与总支付算法，分别用于构建旨在提升容量的候选商用地面设施群组，以及确定用于激励的支付费用
• 我们基于真实世界设定进行了广泛的仿真实验。结果表明，与其他三种先进方案相比，我们的SusCO框架在能耗方面平均多降低11.36%至26.07%，在电池寿命消耗方面平均多降低11.15%至26.75%，在端到端延迟方面平均多降低8.37%至32.77%

本文的其余部分组织如下：第二节讨论背景与动机，第三节介绍SusCO框架与模型。第四节阐述问题的数学构建与求解方法，随后第六节进行性能评估。第七节和第八节分别探讨了相关工作与未来工作，第九节为全文总结。

Background and Motivation

A. Why Solely Using Bent-Pipe Is Not Practical?

In the bent-pipe strategy, each LEO satellite relays data from one location to another location which requires the two location points to be under the coverage of the same LEO satellite [4]. This requires a dense ground station deployment, which is not a practical routing strategy for global coverage.

LEO satellite networks require more ground stations than the other satellite network family members, Medium Earth Orbit (MEO) and Geosynchronous (GEO) satellite networks, for bent-pipe routing. For example, LEO satellites with altitudes of 550 km to 1, 000 km can only cover 1.70% to 3.53% of the total Earth surface area at any location [19]. We analyze the number of ground stations required for the bent-pipe routing strategy to achieve global coverage. As shown in Fig. 2, LEO satellite networks require more than a magnitude number of ground stations than MEO and GEO satellite networks in different minimum elevation angles, which are the minimum possible angles between the horizon and the communicating satellite for a ground station.

Considering the high cost (i.e., construction, licensing application, and maintenance costs) of a ground station [6], the bent-pipe routing strategy is not practical for many LEO satellite network operators, especially those at the very beginning stage with a limited budget. A more cost-effective way to access ground stations is necessary to help these LEO satellite network operators enter the market.

在“弯管”策略中，每颗LEO卫星将数据从一个地点中继到另一个地点，这要求收发两个地点必须同时处于同一颗LEO卫星的覆盖范围之下[4]。该要求需要密集的地面站部署，因此对于实现全球覆盖而言，这并非一个切合实际的路由策略。

在“弯管”路由模式下，相比于中地球轨道（MEO）和地球同步轨道（GEO）卫星网络，LEO卫星网络需要更多的地面站。例如，高度在550公里至1000公里的LEO卫星，在任何位置点上仅能覆盖地球总表面积的1.70%到3.53%[19]。我们分析了“弯管”路由策略为实现全球覆盖所需的地面站数量。如图2所示，在不同的最低仰角（地面站与通信卫星之间的最小允许角度）下，LEO卫星网络所需的地面站数量比MEO和GEO网络要高出一个数量级以上。

考虑到单个地面站的高昂成本（即建设、许可申请及维护成本）[6]，“弯管”路由策略对于许多LEO卫星网络运营商而言是不切实际的，特别是对于那些预算有限的初创运营商。因此，有必要寻求一种更具成本效益的方式来接入地面站，以帮助这些LEO卫星网络运营商进入市场。

B. Why Solely Using ISL Is Not Sustainable?

Routing data solely through inter-satellite links (ISLs) can decrease satellite battery levels and negatively impact the battery lifespan as it increases traffic load and energy consumption on the satellites. Each satellite hop on the path needs to receive, process, and forward the data to the next hop.

The negative impact on battery lifespan is more severe in LEO satellite networks due to more frequent eclipse events. We analyze the battery level changes over 8 hours for LEO, MEO, and GEO satellites based on the energy-related models from [11]. As shown in Fig. 3, LEO satellites encounter eclipse events more than MEO and GEO satellites do due to their low altitudes, resulting in low battery levels several times.

Considering the negative impact on battery lifespan from the energy consumption at low battery levels [11], solely using ISLs for data routing is not sustainable. Having a cost-effective way to access ground stations is a promising solution to offload data to terrestrial networks, reducing the traffic load and prolonging the battery lifespan for LEO satellites.

完全依赖星间链路（ISL）进行数据路由会增加卫星的通信负载和能量消耗，从而导致卫星电池电量下降，并对其电池寿命产生负面影响。在传输路径上的每一跳卫星都需要接收、处理数据，并将其转发到下一跳。

由于更频繁的“日食期”（eclipse events），这种对电池寿命的负面影响在LEO卫星网络中尤为严重。我们基于[11]中的能量相关模型，分析了LEO、MEO和GEO卫星在8小时内的电池电量变化。如图3所示，由于轨道高度低，LEO卫星遭遇日食期的频率远超MEO和GEO卫星，导致其电池电量多次降至低点。

考虑到在低电量水平下，能量消耗对电池寿命的负面冲击会加剧[11]，仅使用ISL进行数据路由是不可持续的。因此，寻求一种经济高效的地面站接入方式，将数据卸载到地面网络，从而减少LEO卫星的通信负载并延长其电池寿命，是一个极具前景的解决方案。

C. Ground Station as a Service Still Faces Challenges

Ground Station as a Service (GSaaS) sells commercial ground station resources, i.e., the use of ground stations for receiving or sending data from or to satellites, in a pay-as-you-go manner, where satellite operators only pay for what they have used. Fig. 4 illustrates an overview of GSaaS. LEO satellite operators can schedule ground station resources from GSaaS providers based on their current requirements, i.e., peak or off-peak hours. Then, the providers can allocate corresponding resources adaptively to provide the capabilities that meet the requirements. Next, the data are offloaded to the scheduled ground stations, and the satellite operators pay for the resources they have used.

The pay-as-you-go manner can greatly reduce the burden of initial investment on ground stations for satellite operators, especially those at the beginning stage [6]. Furthermore, using GSaaS with pay-as-you-go can avoid the issue of excessive or insufficient ground stations since satellite operators usually need to plan for ground station construction ahead of time. With these advantages, GSaaS has gained much attention from the industry. For example, NASA plans to switch to GSaaS rather than building dedicated ground stations since the costs of maintaining and operating them consume a significant portion of its budget [20].

However, GSaaS still faces unique challenges. While GSaaS offers flexible pay-as-you-go services, selecting the GSaaS provider that best suits the requirements can be challenging. Using GSaaS from a single provider may lead to budget inefficiency if the resources are overpriced, or not satisfying the requirements if the quality of service (QoS) does not meet the expectation. Using GSaaS from diversified providers is ideal but selecting a group of GSaaS providers is non-trivial since each provider offers services with different QoS, which can result in insufficient or excessive resources. Hence, to fully leverage the potential of GSaaS, we propose SusCO framework for LEO satellite data offloading by adaptively orchestrating a group of commercial ground stations, while also ensuring service QoS.

“地面站即服务”（Ground Station as a Service, GSaaS）以“按需付费”（pay-as-you-go）的方式销售商用地面站资源（即使用地面站接收或发送卫星数据的服务），卫星运营商只需为他们实际使用的资源付费。图4展示了GSaaS的概览。LEO卫星运营商可以根据其当前需求（如高峰或低谷时段）向GSaaS提供商调度地面站资源。随后，提供商可以自适应地分配相应资源，以提供满足需求的服务能力。接下来，数据被卸载到已调度的地面站，卫星运营商则按其使用量支付费用。

这种“按需付费”模式可以极大减轻卫星运营商，特别是初创运营商，在地面站上的初期投资负担[6]。此外，由于运营商通常需要提前规划地面站的建设，使用“按需付费”的GSaaS可以避免地面站资源建设过量或不足的问题。凭借这些优势，GSaaS已获得业界的广泛关注。例如，美国国家航空航天局（NASA）正计划转向GSaaS，而非建设专用的地面站，因为维护和运营这些设施的成本占据了其预算的很大一部分[20]。

然而，GSaaS仍面临其独特的挑战。尽管GSaaS提供灵活的“按需付费”服务，但选择最符合需求的GSaaS提供商可能极具挑战性。使用单一提供商的GSaaS，可能会因资源定价过高而导致预算效率低下，或者因服务质量（QoS）未达到预期而无法满足需求。采用来自不同提供商的GSaaS是理想的选择，但遴选一个合适的GSaaS提供商组合是一个“非平凡的”（non-trivial）问题，因为每个提供商的服务质量各不相同，这可能导致资源配置不足或过剩。

因此，为充分发挥GSaaS的潜力，我们提出了用于LEO卫星数据卸载的SusCO框架，通过自适应地协同调度一组商用地面站，来完成数据卸载任务，同时确保服务的QoS

TL; DR

背景很简单：

现在常用的是基于 ISL 和 Bent-Pipe 的数据传递路径，这两个的共同点是都需要依赖GS。

1. 仅依靠 ISL / Bent-Pipe 是不行的:
   1. ISL: 电量 / 能耗 / 质量...
   2. Bent-Pipe: 需要极为密集的GS
2. 现在的局面是, GS 作为 sat-ground 的“几乎唯一”途径, 形成了 GSaaS 局面
   1. 对于一些刚起步的小型卫星运营商, 大量部署GS是成本过高的
   2. 不同的GS, 采用的协议/信令等不同, 从这个角度而言我们希望 GS 是“天下大同”的
   3. 但是一方面, 这样会引发 “一家独大” 局面
   4. 另一方面, 单一家的话, 有些独特的需求可能没法满足 (QoS / resources / ...)

因此, 有没有一种协作框架, 能够既实现"统一", 也能满足"需求的多样性", 最好还可以再减少"垄断"呢?

本文提出 SusCO, 将数据从LEO卫星卸载至现有商用地面设施:

1. 我们如何选择一个协同工作的商用地面设施群组，以在提升总容量的同时保持高质量的服务？
2. 我们如何为商用地面设施和LEO卫星运营商设计一种激励机制，以鼓励他们参与SusCO框架？