Enabling secure and intelligent network services management for edge-to-cloud applications
No Thumbnail Available
Authors
Meeting name
Sponsors
Date
Journal Title
Format
Thesis
Subject
Abstract
[EMBARGOED UNTIL 12/01/2026] Modern edge-to-cloud applications such as autonomous drones, precision agriculture, and secure surveillance depend on intelligent network services to ensure performance, trust, and resilience in real-world deployments. These systems require low-latency, highthroughput, and adaptive routing strategies capable of operating under dynamic network and environmental conditions, especially in mission-critical or adversarial contexts. This thesis presents architectures, frameworks, and field-tested solutions that enable secure and intelligent edge-to-cloud operations. First, it introduces methods to improve security in Collaborative Drone Systems (CDS) through trust enforcement and real-time anomaly detection. This includes DronetNetSec for trust scoring, Arculus for context-aware Zero Trust enforcement, and two Federated Learning-enabled IDS frameworks: one for detecting model-level inconsistencies and another for identifying anomalies in network communication. These approaches are validated through simulations and AERPAW testbed experiments. Second, it presents TIGER, a programmable network orchestration framework for managing edge-to-cloud video traffic. TIGER supports emergency-aware routing and dynamic congestion control using Software-Defined Networking (SDN) and P4-based policies to ensure reliable transmission under variable demand and network stress. Third, it proposes scalable, intelligent data processing pipelines for distributed sensing and analytics. FlyNet enables network-aware video offloading; a modular multi-sensor framework supports autonomous field data collection; and FieldVision leverages multiagent reinforcement learning to coordinate drone missions for computer vision tasks in agriculture. These systems are evaluated using UAV data and edge workloads on platforms such as Chameleon and AERPAW.
Table of Contents
PubMed ID
Degree
Ph. D.