Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Microservice Deployment in Space Computing Power Networks via Robust Reinforcement Learning

Zhiyong Yu, Yuning Jiang, Xin Liu, Yuanming Shi, Chunxiao Jiang, Linling Kuang

TL;DR

The paper tackles real-time remote sensing inference over Space-CPN by decomposing monolithic onboard tasks into microservices and optimizing their deployment across a heterogeneous LEO constellation. It introduces a two-stage robust reinforcement-learning framework (MSRARL) that first places core microservices and then optimizes light microservice deployment under a box uncertainty set for regional data demand, formulating the problem as a robust POMDP. By decomposing the problem and applying adversarial RL, the approach achieves sub-optimal but robust solutions that minimize resource use while preserving QoS, with formal equilibrium analysis guaranteeing a minimax-Nash correspondence. Simulations on a representative LEO mesh demonstrate improved latency, reduced resource consumption, and resilience to demand fluctuations compared to Vanilla RL and heuristic baselines. The work advances onboard, scalable, and robust inference in space-based edge networks, enabling real-time RS tasks under stringent constraints.

Abstract

With the growing demand for Earth observation, it is important to provide reliable real-time remote sensing inference services to meet the low-latency requirements. The Space Computing Power Network (Space-CPN) offers a promising solution by providing onboard computing and extensive coverage capabilities for real-time inference. This paper presents a remote sensing artificial intelligence applications deployment framework designed for Low Earth Orbit satellite constellations to achieve real-time inference performance. The framework employs the microservice architecture, decomposing monolithic inference tasks into reusable, independent modules to address high latency and resource heterogeneity. This distributed approach enables optimized microservice deployment, minimizing resource utilization while meeting quality of service and functional requirements. We introduce Robust Optimization to the deployment problem to address data uncertainty. Additionally, we model the Robust Optimization problem as a Partially Observable Markov Decision Process and propose a robust reinforcement learning algorithm to handle the semi-infinite Quality of Service constraints. Our approach yields sub-optimal solutions that minimize accuracy loss while maintaining acceptable computational costs. Simulation results demonstrate the effectiveness of our framework.

Microservice Deployment in Space Computing Power Networks via Robust Reinforcement Learning

TL;DR

The paper tackles real-time remote sensing inference over Space-CPN by decomposing monolithic onboard tasks into microservices and optimizing their deployment across a heterogeneous LEO constellation. It introduces a two-stage robust reinforcement-learning framework (MSRARL) that first places core microservices and then optimizes light microservice deployment under a box uncertainty set for regional data demand, formulating the problem as a robust POMDP. By decomposing the problem and applying adversarial RL, the approach achieves sub-optimal but robust solutions that minimize resource use while preserving QoS, with formal equilibrium analysis guaranteeing a minimax-Nash correspondence. Simulations on a representative LEO mesh demonstrate improved latency, reduced resource consumption, and resilience to demand fluctuations compared to Vanilla RL and heuristic baselines. The work advances onboard, scalable, and robust inference in space-based edge networks, enabling real-time RS tasks under stringent constraints.

Abstract

With the growing demand for Earth observation, it is important to provide reliable real-time remote sensing inference services to meet the low-latency requirements. The Space Computing Power Network (Space-CPN) offers a promising solution by providing onboard computing and extensive coverage capabilities for real-time inference. This paper presents a remote sensing artificial intelligence applications deployment framework designed for Low Earth Orbit satellite constellations to achieve real-time inference performance. The framework employs the microservice architecture, decomposing monolithic inference tasks into reusable, independent modules to address high latency and resource heterogeneity. This distributed approach enables optimized microservice deployment, minimizing resource utilization while meeting quality of service and functional requirements. We introduce Robust Optimization to the deployment problem to address data uncertainty. Additionally, we model the Robust Optimization problem as a Partially Observable Markov Decision Process and propose a robust reinforcement learning algorithm to handle the semi-infinite Quality of Service constraints. Our approach yields sub-optimal solutions that minimize accuracy loss while maintaining acceptable computational costs. Simulation results demonstrate the effectiveness of our framework.
Paper Structure (35 sections, 7 theorems, 44 equations, 11 figures, 2 tables, 2 algorithms)

This paper contains 35 sections, 7 theorems, 44 equations, 11 figures, 2 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. Contributions
  3. Organization
  4. Preliminaries and Notations
  5. System Model
  6. Microservice Model
  7. LEO Constellation Model
  8. Latency Model
  9. Microservice Cost Model
  10. Deterministic Constraint
  11. Function Completeness Constraint
  12. Parallel Computing Constraint
  13. QoS Constraint
  14. Core Microservices Resource Constraint
  15. Light Microservices Resource Constraint
  16. ...and 20 more sections

Key Result

Lemma 1

The adversarial game in Section POMDP can be transferred into a two-player zero-sum game.

Figures (11)

  • Figure 1: On-board inference process of remote sensing.
  • Figure 2: Remote sensing onboard satellite edge AI inference with microservice architecture.
  • Figure 3: Microservice deployment problem specification.
  • Figure 4: Example of data amount uncertainty.
  • Figure 5: Multi-Step Robust Adversarial Reinforcement Learning framework.
  • ...and 6 more figures

Theorems & Definitions (8)

  • Definition 1: Minimax Equilibrium
  • Lemma 1
  • Lemma 2
  • Lemma 3
  • Lemma 4
  • Lemma 5
  • Lemma 6
  • Theorem 1