The proposed Doctoral Thesis is dedicated to the study of chaos phenomenon applications in communication systems. The work provides an overview of the applications of chaotic signals in communication systems, shows the most promising applications and identifies the main concerns. Applications of chaotic sequences for spread-spectrum code-division systems and for multicarrier systems are explored. Properties, strengths and weaknesses of spread-spectrum code-division and multicarrier systems are analysed. The generation methods and correlation properties of the chaotic sequences are studied. The ability of application of analytical methods for performance evaluation of the DS-CDMA system based on chaotic sequences are examined. Chaotic sequence selection algorithms that decrease the sequence correlation level and improve the interference immunity of DS-CDMA system are proposed. The influence of system and transmission parameters (sequences generation algorithm and length, number of users and transmission mode) on the DS-CDMA system performance are evaluated. A PAPR reduction method for OFDM signals and a linear precoding method for OFDM system performance improvement in multipath propagation environment are proposed. Furthermore, the validation of the offered performance improvement methods is performed using MATLAB and Simulink simulations. The proposed systems can be used for highly secure sensor networks with simple signal processing and moderate data rate.