Work

Resource Allocation for Latency Reduction in Geographically Distributed Real-Time Simulation

Year: 2026
Type: conference-paper
Abstract: This paper introduces a new latency reduction approach for geographically distributed real-time simulation (GDRTS) where different models run in a distributed architecture coupled by a communication network. Delay is a major issue and can influence the stability of the GDRTS. Different methods exist to reduce or compensate delay. They are specifically tailored for given models and are not usable in a generic way. Instead, since fog or edge computing uses resource allocation, this work exploits these methods to minimize latency in GDRTS. Models are allocated to resources that have minimal latency between each other, thus reducing communication latency which is the largest contribution to delay in GDRTS. This paper builds on top of an existing resource allocation approach by introducing a more realistic latency model, an additional constraint to limit the values of latency, and a margin to cover other delay contributions from the resources. A case study demonstrates that resource allocation fulfills timestep requirements and reduces the latency. The method contributed by this paper achieves the reduction of the major contribution to delay in GDRTS setup
Language: English
Open Access status: hybrid
OpenAlex ID: W7224770442
FWCI:
Cites: 11
Cited by:
Related to: 10