Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

MLFCIL: A Multi-Level Forgetting Mitigation Framework for Federated Class-Incremental Learning in LEO Satellites

Heng Zhang, Xiaohong Deng, Sijing Duan, Wu Ouyang, KM Mahfujul, Yiqin Deng, Zhigang Chen

Abstract

Low-Earth-orbit (LEO) satellite constellations are increasingly performing on-board computing. However, the continuous emergence of new classes under strict memory and communication constraints poses major challenges for collaborative training. Federated class-incremental learning (FCIL) enables distributed incremental learning without sharing raw data, but faces three LEO-specific challenges: non-independent and identically distributed data heterogeneity caused by orbital dynamics, amplified catastrophic forgetting during aggregation, and the need to balance stability and plasticity under limited resources. To tackle these challenges, we propose MLFCIL, a multi-level forgetting mitigation framework that decomposes catastrophic forgetting into three sources and addresses them at different levels: class-reweighted loss to reduce local bias, knowledge distillation with feature replay and prototype-guided drift compensation to preserve cross-task knowledge, and class-aware aggregation to mitigate forgetting during federation. In addition, we design a dual-granularity coordination strategy that combines round-level adaptive loss balancing with step-level gradient projection to further enhance the stability-plasticity trade-off. Experiments on the NWPU-RESISC45 dataset show that MLFCIL significantly outperforms baselines in both accuracy and forgetting mitigation, while introducing minimal resource overhead.

MLFCIL: A Multi-Level Forgetting Mitigation Framework for Federated Class-Incremental Learning in LEO Satellites

Abstract

Low-Earth-orbit (LEO) satellite constellations are increasingly performing on-board computing. However, the continuous emergence of new classes under strict memory and communication constraints poses major challenges for collaborative training. Federated class-incremental learning (FCIL) enables distributed incremental learning without sharing raw data, but faces three LEO-specific challenges: non-independent and identically distributed data heterogeneity caused by orbital dynamics, amplified catastrophic forgetting during aggregation, and the need to balance stability and plasticity under limited resources. To tackle these challenges, we propose MLFCIL, a multi-level forgetting mitigation framework that decomposes catastrophic forgetting into three sources and addresses them at different levels: class-reweighted loss to reduce local bias, knowledge distillation with feature replay and prototype-guided drift compensation to preserve cross-task knowledge, and class-aware aggregation to mitigate forgetting during federation. In addition, we design a dual-granularity coordination strategy that combines round-level adaptive loss balancing with step-level gradient projection to further enhance the stability-plasticity trade-off. Experiments on the NWPU-RESISC45 dataset show that MLFCIL significantly outperforms baselines in both accuracy and forgetting mitigation, while introducing minimal resource overhead.

Paper Structure

This paper contains 40 sections, 24 equations, 13 figures, 6 tables, 2 algorithms.

Table of Contents

  1. Introduction
  2. Related Work
  3. Non-IID Data Heterogeneity in FL
  4. Catastrophic Forgetting in CIL
  5. FCIL in LEO satellites
  6. System Scenario and Problem Formulation
  7. System Scenario
  8. Problem Formulation
  9. Optimization Objective
  10. Methodology Design
  11. Overview of MLFCIL
  12. Local Training
  13. Class-Balanced Training (Level I)
  14. Knowledge Preservation (Level II)
  15. Multi-Component Loss Function
  16. ...and 25 more sections

Figures (13)

  • Figure 1: FCIL scenario in LEO: edge satellites perform on-board class-incremental training and synchronize with a regional aggregation satellite via inter-satellite links (ISL), without reliance on ground stations.
  • Figure 2: System scenario of MLFCIL.
  • Figure 3: Overview of the MLFCIL.
  • Figure 4: Prototype-guided feature drift compensation. (a) Feature drift problem. (b) After drift compensation.
  • Figure 5: Stability-plasticity gradient projection. (a) Before projection (conflict). (b) After projection (resolved).
  • ...and 8 more figures