Comparison of Priority Queue algorithms for Hierarchical Scheduling Framework

The Time Event Queue (TEQ) is a datastructure that is part of the implementation of a Hierarhcial Scheduling Framework (HSF). It's main function is to store future task scheduling events (represented as absolute time values) in a sorted manner, thus, it implements a Priority Queue (PQ). A PQ is a queue with elements that are sorted by their priority. The two main operations on a PQ is insert and delete-min. The first mentioned operation will insert an item based on its priority. The second operation will extract an item with highest (or lowest) priority. As of now, the PQ structure is implemented as a sorted linked list with a median pointer and where binary search is used for insertions. The efficiency of the current implementation is not up to standard so there is a requirement that the current implementation should be replaced or optimized. An investigation should be made so that other PQ implementations are considered that are well suited for hard real-time systems. When choosing a suited PQ for the hard real-time scheduler HSF, considerations such as a low number of queue items in combination with good worst case performance is of importance. This paper will motivate and choose a group of suited PQ algorithms and present empirical tests of each of these algorithms. Based on these results and our requirements, one algorithm will be chosen to be integrated with the HSF.