A New Function-based Framework for Classification and Evaluation of Mutual Exclusion Algorithms

This paper presents a new function-based framework for mutual exclusion algorithms indistributed systems. In the traditional classification mutual exclusion algorithms were dividedin to two groups: Token-based and Permission-based. Recently, some new algorithms areproposed in order to increase fault tolerance, minimize message complexity and decreasesynchronization delay. Although the studies in this field up to now can compare and evaluatethe algorithms, this paper takes a step further and proposes a new function-based frameworkas a brief introduction to the algorithms in the four groups as follows: Token-based,Permission-based, Hybrid and K-mutual exclusion. In addition, because of being dispersaland obscure performance criteria, introduces four parameters which can be used to comparevarious distributed mutual exclusion algorithms such as message complexity, synchronizationdelay, decision theory and nodes configuration. Hope the proposed framework provides asuitable context for technical and clear evaluation of existing and future methods.