1998•International Journal of Modelling and SimulationRequires access

A General Mutual Exclusion Primitive: Simulation And Evaluation

M.H. Samadzadeh, R.R. Avutu

Open publisher page 1 citations

Abstract

The need for control of concurrent processes is a consequence of problems associated with their unmanaged execution. The mechanisms that enable such control are generally called synchronization or mutual exclusion primitives. When processes try to access shared data simultaneously, mutual exclusion is to be enforced. Over the past three decades, the mutual exclusion phenomenon has emerged as one of the foremost paradigms of the difficulties associated with parallel and distributed programming, and several solutions for it have been offered. In this paper a new primitive for mutual exclusion is proposed. This primitive was implemented and evaluated in a simulated setting. It was shown to satisfy the requirements for any mutual exclusion scheme. The new primitive compares favourably with other well-known and efficient n-process mutual exclusion primitives in terms of the worst-case waiting time, the number of variables used, and understandability.

About this research paper

What this paper is about

The need for control of concurrent processes is a consequence of problems associated with their unmanaged execution. The mechanisms that enable such control are generally called synchronization or mutual exclusion primitives. When processes try to access shared data simultaneously, mutual exclusion is to be enforced. Over the past three decades, the mutual exclusion phenomenon has emerged as one of the foremost paradigms of the difficulties associated with parallel and distributed programming, and several solutions for it have been offered. In this paper a new primitive for mutual exclusion is proposed. This primitive was implemented and evaluated in a simulated setting. It was shown to satisfy the requirements for any mutual exclusion scheme. The new primitive compares favourably with other well-known and efficient n-process mutual exclusion primitives in terms of the worst-case waiting time, the number of variables used, and understandability.

Why it matters

OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.

Key contribution

A contribution statement is not available in the OpenAlex record.

Method / approach

Method details are not available in the OpenAlex metadata.

Main findings

Findings are not separately available in the OpenAlex metadata.

Limitations

Limitations are not available in the OpenAlex metadata.

Applications

Application details are not available in the OpenAlex metadata.

Available abstract

The need for control of concurrent processes is a consequence of problems associated with their unmanaged execution. The mechanisms that enable such control are generally called synchronization or mutual exclusion primitives. When processes try to access shared data simultaneously, mutual exclusion is to be enforced. Over the past three decades, the mutual exclusion phenomenon has emerged as one of the foremost paradigms of the difficulties associated with parallel and distributed programming, and several solutions for it have been offered. In this paper a new primitive for mutual exclusion is proposed. This primitive was implemented and evaluated in a simulated setting. It was shown to satisfy the requirements for any mutual exclusion scheme. The new primitive compares favourably with other well-known and efficient n-process mutual exclusion primitives in terms of the worst-case waiting time, the number of variables used, and understandability.

Key concepts: Mutual exclusion, Computer science, Synchronization (alternating current), Mutual information, Suzuki-Kasami algorithm, Process (computing), Distributed computing, Theoretical computer science

Related papers

Back to paper searchBrowse research topicsOriginal source
A General Mutual Exclusion Primitive: Simulation And Evaluation — Research Paper | ScholarLens