Modeling of Resource Allocation Mechanisms in Distributed Computing Systems using Petri Nets and Stochastic Activity Networks (SAN): a Review and Reo-based Suggestion
It tackles resource allocation efficiency and deadlock prevention in distributed systems, but is incremental as it builds on established modeling techniques.
This paper reviews existing Petri Net and Stochastic Activity Network models for resource allocation in distributed computing systems and proposes a new approach based on Reo coordination language to address identified limitations.
Resource allocation is crucial in the distributed systems. It is a key step in designing the mechanisms of systems for determining the resource allocation mechanism, it is important for obtaining the desired efficiency in the system, plus it is vital for predicting and preventing Deadlocks. Various models of Petri Net (Stochastic PN, Colored PN, Generalized PN, etc.) are used for modeling, simulation, execution, and solving the problems of resource allocation. SAN models are used for modeling the problems pertinent to resource allocation. First, we shall address the basic concepts pertinent to these models and the resource allocation problem (introduction chapter), then, some applications of the Petri Net and SAN models in the distributed computational systems or systems based on them shall be studied. Finally, the issues and findings will be concluded.