Performance Evaluation of Sorting Algorithms in Raspberry Pi and Personal Computer

Abstract

Nowadays, more and more data is going to be collected. Most of the collected data must be sorted to be analysed. Thus it is important to use efficient algorithms to save time and energy. This is especially important for battery-operated devices that collect and sort data directly. The lifetime of the battery can be remarkably extended by choosing effective sorting algorithms. Furthermore, processors with lower clock frequencies can be used, leading to lower battery consumption.

This thesis focuses on the time complexity of algorithms (bubble sort, insertion sort, quick sort and heap sort) executed on raspberry Pi and personal computer. In this thesis, a client and a server have been implemented. The client is a Graphical User Interface (GUI) that generates random numbers, which are sent to the server. On the server the random numbers are sorted and the sorting time is measured. After sorting, the random numbers will be sent back to the client where they are displayed on the GUI. The sorting time is also send to the GUI in order to display it.

It is to mention that the time consumption is measured with timing functions of the C library. An oscilloscope has been used to check if the timing function works properly. At the end, the performances of four sorting algorithms on Raspberry Pi and on the Personal Computer (PC) are compared.