Forecasting-Driven Stable Successor Matching for UAV-Assisted Continuous Edge Services
Forecasting-Driven Stable Successor Matching for UAV-Assisted Continuous Edge Services
Authors: Houyi Qi, Minghui Liwang, Yuhan Su, Xianbin Wang Date: 2026-05-04 Paper ID: openalex:2605.02180
Summary
This paper addresses the problem of service continuity in UAV-assisted edge networks (UENs) by proposing a proactive successor matching framework named Fresco. The approach utilizes an LSTM model to forecast potential service disruption risks caused by UAV mobility and energy depletion, triggering proactive service-context synchronization. Evaluation demonstrates that Fresco significantly reduces service interruptions compared to reactive baselines while maintaining low resource reservation overhead.
Key Contributions
- Introduces Fresco, a forecasting-driven service scheduling framework that proactive schedules successor UAVs to mitigate service interruptions in mobile edge networks.
- Employs an LSTM-based module for predicting mission-specific disruption risks based on network observations.
- Develops an online risk-aware successor matching scheme that optimizes standby UAV selection under energy, latency, and resource constraints with minimal overhead.
Open Questions & Future Work
Key Concepts
- proactive-successor-matching: A framework design pattern for mobile edge computing that proactively reserves and synchronizes resources on standby nodes based on predicted disruption risks.
Archivist Review
I have approved the core architectural concept of ‘Proactive Successor Matching’, as it represents a shift in service scheduling strategy for edge networks beyond simple reactive handovers. The open question regarding predictive robustness was approved because it addresses a fundamental reliability bottleneck in proactive resource reservation. I rejected ‘Fresco’ as it is a specific framework name and a routine application of existing predictive methods.
Approved Concepts
- Proactive Successor Matching: This shift from reactive handover to proactive service reservation in edge networks is a distinct architectural pattern that is likely to see further development in mobile computing.
Approved Open Questions
- Robust Predictive Risk Assessment: Predictive accuracy is the bottleneck for the entire proactive reservation process; if predictions are inaccurate, the system incurs unnecessary reservation overhead or fails to prepare for actual failures.
Rejected Candidates
- [concept] Fresco (
fresco) - paper_local: This is a specific framework name rather than a reusable concept; it is a composition of existing techniques rather than a novel mechanism itself.
Links
Metadata & Links
- url
- https://arxiv.org/abs/2605.02180
- paper_id
- 2605.02180
- paper_source
- openalex
- domain
- nlp
- tags
-
- architectures
-
- datasets
-
- concept_slugs
- proactive-successor-matching
- dataset_slugs
-
- skill
- TimeSeriesSkill
- created_at
- 2026-05-07T06:04:22Z