Date of Award

5-2018

Culminating Project Type

Starred Paper

Degree Name

Computer Science: M.S.

Department

Computer Science and Information Technology

College

School of Science and Engineering

First Advisor

Jie Hu Meichsner

Second Advisor

Dennis Guster

Third Advisor

Mehdi Mekni

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

Keywords and Subject Headings

Fibonacci Heap, Parallel Data Structure, Scalable Data Structure, Multi-threading

Abstract

With the advancement of multiple processors, the sequential algorithms are being investigated and gradually substituted for its concurrent equivalent to effectively exploit the parallel architecture. Parallel algorithms speed up the performance by dividing the task into a number of processes (or threads) that can be scheduled and executed simultaneously in independent processing units. Various well-known basic algorithms and data-structures have been explored for its efficient parallel counterparts and have been published as popular libraries. However, advanced data-structures and algorithms have not seen similar investigation mainly because they have many optimization steps mostly backed by many states and finding safe and efficient parallel implementation isn’t an easy endeavor.

Safety concerns for shared-memory parallel implementation are of utmost importance as it provides a basis for consistency of any data structure and algorithm. There are well-known tools like locks, semaphores, atomic operations and so on that assist towards safe parallel implementation but using them effectively and in well-defined synchronization are key factors in the overall performance of any data-structures and algorithms.

This paper explores an advanced data structure, Fibonacci Heap, and its operations to evaluate its implementation using two different synchronization mechanisms: Coarse-grained and Fine-grained. The analysis in this paper shows that a fine-grained synchronized Fibonacci Heap implementation with certainly relaxed semantics is more scalable with growing number of concurrency in comparison to the coarse-grained synchronized Fibonacci Heap implementation.

Share

COinS
 
 

To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.