Efficient Service Provisioning in the Musical Metaverse Using a Custom Network Simulator

This thesis addresses the critical challenge of efficiently allocating network resources to support the demanding requirements of the Musical Metaverse. Leveraging the principles of Network Virtualization and Virtual Network Embedding (VNE), this work proposes a novel framework for optimizing the placement and routing of musical metaverse services. The unique characteristics of the Musical Metaverse, including ultra-low latency communication, stringent Quality of Service (QoS) requirements, and the need for precise synchronization in multi-user interactions, necessitate a specialized approach beyond traditional VNE solutions. To this end, the thesis introduces the Musical Metaverse Optimization (MusMOPT) model, which extends classical VNE to account for these specific demands. Furthermore, a dedicated simulation environment, SiMusMet, is developed to rigorously evaluate the proposed models and algorithms. SiMusMet provides a comprehensive platform for configuring network topologies, simulating various scenarios, and assessing performance metrics relevant to the Musical Metaverse. The thesis also details the design of a heuristic placement and routing algorithm, incorporating techniques such as community detection and probabilistic iterative refinement, to efficiently map musical metaverse service graphs onto underlying cloud network infrastructures. Through extensive simulations and analysis, this research demonstrates the effectiveness of the proposed MusMOPT model, and heuristic algorithm in achieving near-optimal resource utilization and ensuring a high-quality user experience within the Musical Metaverse.