Satellite positioning has been recognised as an fundamental underlying technology for growing number of modern technology and socio-economic systems, i.e. orientation in space, traffic routes analysis time and space stamping of financial transactions etc. The quality of services provided by those systems relies on the satellite positioning performance, especially positioning accuracy. This paper covers the analysis of (reference) position estimation procedures in the navigation application domain, discovering of potential weaknesses and suggesting the improvement. The algorithm analysis involves practical performance and rating of the algorithm performance quality using measured pseudorange and programmable GPS receiver. A program-specific GPS receiver is practically performed on a personal computer. It was used for processing the GPS receiver’s observations in the RINEX data format to form the input of the algorithm performance. GPS receiver’s observations are collected from IGS reference station located in Padua, Italy. Algorithm enhancements were analyzed by a parallel analysis of the improved and original algorithm. The introductory chapter introduces the world of global navigation satellite systems (GNSS) and formulates the objectives. Chapter one details a selected GNSS, Global Positioning System (GPS), and introduces and defines the problem of positioning in the navigation application domain. Selected GNSS is used when collecting experimental data. The second chapter describes solution finding methods of the problem defined by chapter one. GNSS user equipment based on program-specific radio is used when collecting and pre-processing experimental data. The description of the programmable radio and the GNSS user equipment used are given with the third chapter. Chapter Four provides a practical performance of the basic (benchmark) algorithm for position assessment in the navigation application domain, the reasons for its improvement and performance and description of the improvement. The fifth chapter evaluates and compares the quality of two derived algorithms, basic and improved, followed by conclusions.