Back to search

University of Crete (UOC)

A network-aware application partitioning framework for edge-cloud collaborative extended reality with soft mesh deformations

Abstract

dc:description

Collaborative Extended Reality (XR) requires responsive interaction, consistent shared state, and efficient use of limited client and network resources. These requirements become crucial when 3D scenes contain soft-mesh objects, since their deformation increases computation, network synchronization traffic, and sensitivity to latency, jitter, and packet loss. This thesis presents a novel network-aware application partitioning framework method for edge-cloud collaborative XR with soft mesh deformations. Instead of continuously streaming vertices or particles, the framework represents soft-mesh interaction through particle-based control primitives and ordered interaction events. It combines a novel compact deformable-control synchronization with relay-supported authority management, per-object epochs for handover, separated pose and event streams, and optional edge assistance. A Geometric Algebra (GA)-based interpolation method reconstructs rigid pose and control streams, including controller and constraint-target transforms that drive deformable particles, from lower-rate network updates. The framework is local-first, with remote physics as the directly supported offloading path, while remote rendering remains within the architectural scope. The results of the system report a significant reduction in bandwidth for the pose/control streams of GA with comparable experience quality, lower cost of interpolation on the receive-side and feasible remote-physics operation in scenes with up to 10,000 physics objects and 100+ concurrent-user workloads. Overall, the thesis contributes a new integrated bandwidth-aware system framework method for interactive collaborative XR with deformable content.

Degree

thesis:*
Grantor dc:publisher
University of Crete (UOC)
Year dc:date
2026

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Lydatakis, Nikolaos
  • Λυδατάκης, Νικόλαος

Subjects

dc:subject × 26

Rights

Language dc:language
eng

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/63026

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Lydatakis, Nikolaos; Λυδατάκης, Νικόλαος. A network-aware application partitioning framework for edge-cloud collaborative extended reality with soft mesh deformations. University of Crete (UOC), 2026. http://hdl.handle.net/10442/hedi/63026