Phoenix Design Overview

In this section, we introduce the key idea and overview of our sunlight-aware energy-efficient task scheduling strategy.

A. Observation: Sunlight-Sufficient Space Edges

Based on the SPS knowledge introduced in §II-B, we know that the key to optimizing energy usage and prolonging lifetime for SECs is to reduce the battery energy consumption caused by in-orbit task processing. To this end, our PHOENIX design leverages an important observation: in an LEO satellite constellation, typically there exist many “sunlight-sufficient” satellites which spend most of their orbital time in sunlight.

To quantitatively introduce and understand this phenomenon, we define a metric called sunlit ratio, which is calculated by the ratio of the period a satellite is in sunlight to its total operation period. Fig. 3a plots the CDF of sunlit ratio of four state-of-the-art LEO satellite constellations which differ in their orbital altitudes and inclinations. We calculate the sunlit ratio based on their public real-world satellite trajectories [25] during May 2023. We obtain three important observations. First, we find that the sunlit ratio is approximately higher than 60%, indicating that a large fraction of satellites are exposed to the sun for a long time on their orbits. Second, interestingly we observe that some satellites in certain orbits can achieve near100% sunlit ratio with sufficient sunlit supplement. Fig.3b plots an example to explain this observation. Suppose that radius of the earth is R. The height of satellite orbit is denoted as h and the angle between sunlight and orbital plane is denoted as θ. Then, the distance between the orbit and the earth’s core is R + h and the vertical component of distance perpendicular to sunlight is (R+h)·sinθ. If the length of the vertical component is longer than the radius of the earth, the whole orbit can be exposed to the sunlight and the sunlit ratio can reach 100%.

基于第二节B部分介绍的卫星电源系统（SPS）知识，我们知道，为SEC系统优化能源使用和延长寿命的关键在于减少由在轨任务处理引起的电池能耗。为此，我们的 PHOENIX 设计利用了一个重要观察：在一个LEO卫星星座中，通常存在许多“光照充足”的卫星，它们大部分的轨道时间都处于阳光照射下

为了量化地介绍并理解这一现象，我们定义了一个名为光照比率 (sunlit ratio) 的指标，其计算方式为卫星处于光照下的时长与其总运行周期的比值。

图3a 绘制了四种先进LEO卫星星座（它们的轨道高度和倾角各不相同）的光照比率累积分布函数（CDF）。我们基于它们在2023年5月的公开真实世界卫星轨迹 [25] 计算了光照比率，并从中获得了三个重要观察。首先，我们发现光照比率普遍高于60%，这表明大部分卫星在其轨道上有很长时间都暴露在阳光下。其次，有趣的是，我们观察到某些轨道上的一些卫星可以达到接近100%的光照比率，获得充足的日照补充。

图3b 用一个示例解释了这一观察。假设地球半径为 R，卫星轨道高度为 h，太阳光与轨道平面之间的夹角为 θ。那么，轨道与地心的距离为 R+h，该距离在垂直于太阳光方向上的分量为 (R+h)⋅sinθ。如果这个垂直分量的长度大于地球半径 R，那么整个轨道都可以暴露在阳光下，光照比率可达到100%

B. Basic Idea: Sunlight-Aware SEC Task Scheduling

Inspired by the above crucial observations, PHOENIX exploits a key idea to accomplish energy-efficient SEC: dynamically offloading in-orbit tasks to appropriate power-sufficient nodes (e.g., sunlight-sufficient satellites with near-100% sunlit ratio) without exceeding various task deadlines. Specifically, PHOENIX judiciously schedules SEC tasks based on the following options (also illustrated in Fig. 4) to minimize the battery energy consumption while meeting various requirements of task completion time:

受上述关键观察的启发，PHOENIX 利用一个核心思想来实现高能效的SEC： 在不超过各种任务截止时间的前提下，动态地将在轨任务卸载到功率充足的节点（例如，光照比率接近100%的“光照充足”卫星）。 具体而言，为了在满足各种任务完成时间要求的同时最小化电池能耗，PHOENIX 基于以下几种选项（如图4所示）来审慎地调度SEC任务：

• Real-time local processing (Fig. 4a). Once in-orbit data is acquired on a satellite, processing it locally and immediately no matter where the current satellite is.

• Delayed local processing (Fig. 4b). Once in-orbit data is acquired, the satellite delays the data processing until a specific time point (e.g., when the satellite leaves eclipse).

• Offloading tasks to nearby sunlit edges (Fig. 4c). Instead of being processed locally, space data collected in orbit is transferred to another sunlight-sufficient satellite in the SEC network for energy-efficient processing.

• Offloading tasks to available ground stations (Fig. 4d). Space data is transferred to a ground station with sufficient power and computation capability through the SEC network.

Essentially, the above scheduling options represent different preferences on saving battery energy consumption and guaranteeing task completion time. For example, immediate local data processing may achieve shorter mission completion time, but possibly at the cost of higher battery energy consumption. Transferring the entire in-orbit task to the ground may save energy for a satellite edge, but offloading high-volume space data can easily overwhelm the limited downlink and involve unacceptable latency. PHOENIX dynamically schedules various SEC tasks under different computation, network and sunlight conditions, and further makes proper scheduling decisions.

• 实时本地处理 (图4a) 一旦在卫星上获取到在轨数据，无论当前卫星位于何处，都立即在本地进行处理
• 延迟本地处理 (图4b) 获取在轨数据后，卫星将数据处理延迟到某个特定时间点（例如，当卫星离开地影区时）
• 卸载任务至附近的“光照边缘”(图4c) 将在轨道上收集的空间数据不进行本地处理，而是传输到SEC网络中的另一颗“光照充足”的卫星上进行高能效处理
• 卸载任务至可用的地面站 (图4d) 通过SEC网络，将空间数据传输到拥有充足电力和计算能力的地面站

本质上，上述调度选项代表了在节省电池能耗和保障任务完成时间之间的不同权衡。例如，立即本地处理数据可能会缩短任务完成时间，但可能以更高的电池能耗为代价。将整个在轨任务传输到地面可能会为卫星边缘节点节省能源，但卸载海量空间数据很容易使有限的下行链路不堪重负，并引入不可接受的延迟。PHOENIX 能够根据不同的计算、网络和光照条件动态调度各种SEC任务，并作出恰当的决策。

C. System Overview

Architecture. Fig. 5 shows the system overview of PHOENIX, which combines: (i) a swarm of computational space edges constructing an SEC network, and (ii) terrestrial infrastructures such as geo-distributed ground stations and a mission control center. Each satellite is equipped with high-speed ISLs and GSLs for inter-satellite and ground-satellite communication. In addition, an on-board intelligent processor is deployed on each satellite for in-orbit computing. Based on this baseline SEC architecture, PHOENIX accomplishes energy-efficient SEC task scheduling by incorporating two new components as follows.

• Centralized mission center controller. The controller is a centralized coordinator, which consists of three modules: task publisher, sunlight predictor and orbit assignment manager. The task publisher distributes the tasks to satellites. The sunlight predictor predicts the sunlight states of satellites based on their trajectories and distributes the sunlight information to satellites for offloading decisions. The orbit assignment manager pre-allocates alternative orbit sets for satellites to avoid resource competition between orbits.

• Distributed on-board task manager. The manager on each satellite consists of two modules: local processing manager and offloading manager. The offloading manager decides where to process the task based on the pre-allocated orbit set. The local processing manager arranges the task execution time if a task is decided to be processed locally.

架构 图5 展示了 PHOENIX 的系统概览，它结合了：

(i) 由大量计算型空间边缘节点构成的SEC网络

(ii) 地理上分布的地面站和任务控制中心等地面基础设施

每颗卫星都配备了用于星间和星地通信的高速ISL和GSL。此外，每颗卫星上都部署了用于在轨计算的星上智能处理器。基于此基准SEC架构，PHOENIX 通过集成以下两个新组件来完成高能效的SEC任务调度。

• 集中式任务中心控制器: 该控制器是一个集中式协调器，由三个模块组成

  • 任务发布器 && 光照预测器 && 轨道分配管理器
  • 任务发布器: 负责向卫星分发任务
  • 光照预测器: 基于卫星轨迹预测其光照状态，并将光照信息分发给卫星以辅助卸载决策
  • 轨道分配管理器: 为卫星预先分配备选的轨道集合，以避免轨道间的资源竞争

• 分布式星上任务管理器:

  • 每颗卫星上的管理器包含两个模块： 本地处理管理器 和 卸载管理器
  • 卸载管理器: 根据预分配的轨道集决定任务的处理位置
  • 如果任务被决定在本地处理，本地处理管理器则负责安排任务的执行时间

---

Workflow. SEC tasks are scheduled by the following steps:

• Task publication and orbit assignment. The mission center receives tasks from customers, such as requirements of using SEC for disaster discovery, climate monitoring, maritime search and rescue. The common feature of these tasks is that they all require a constellation of satellites persistently capture information (e.g., high-resolution images) of a certain region of interest (RoI). Then, the mission center predicts the sunlight states of satellites and publishes the RoI along with sunlight information to all satellites. Based on the sunlight and task information, the mission center pre-assigns an orbit subset for each orbit and distributes the assignment decision.

• Task offloading. When a satellite flies over RoI, it captures images and selects offloading destination for each task, either a ground station or a satellite (including itself, i.e., local processing). The satellite first checks whether the task can be offloaded to a ground station before deadline. If not, it tries to arrange the task locally and judges whether it can finish the task under sunlight phase as well as before deadline. If both of the above conditions can not be satisfied, it offloads the task to other satellite.

• Processing time arrangement. The final destination satellite arranges the task processing time (the ground station can process the task immediately after offloading) and sends the result back to the mission center when the task is finished.

工作流程 SEC 任务的调度遵循以下步骤：

• 任务发布与轨道分配

  • 任务中心从客户处接收任务，例如使用SEC进行灾害发现、气候监测、海上搜救等需求
  • 这些任务的共同特点是，它们都要求一个卫星星座对某个感兴趣区域（Region of Interest, RoI） 持续捕获信息（如高分辨率图像）
  • 然后，任务中心预测卫星的光照状态，并 将RoI信息与光照信息一同发布给所有卫星
  • 基于光照和任务信息，任务中心为每个轨道预分配一个轨道子集，并分发该分配决策

• 任务卸载

  • 当一颗卫星飞越RoI时，它会捕获图像，并为每个任务选择卸载目的地——可以是地面站，也可以是另一颗卫星（包括其自身，即本地处理）
  • 卫星首先检查任务是否能在截止时间前卸载至地面站。如果不能，它会尝试在本地安排任务，并判断是否能在光照阶段且在截止时间前完成该任务
  • 如果以上两个条件都无法满足，它将任务卸载至其他卫星

• 处理时间安排

  • 最终的目的地卫星负责安排任务的处理时间（地面站则在接收到任务后立即处理），并在任务完成后将结果回传至任务中心