الفهرس 
ADMR CategoriesOnly 14 pages are availabe for public view
 |  | from | 133 |  |  |
| | | from | 133 | | |
Abstract
The automatic modulation recognition system is a system that identifies the type and the order of the modulation scheme that used to transmit the signal to the receiver without any prior knowledge of the characteristics of such a signal. These systems can play an important role due to its ability to integrate multiple receivers in a single device. It has many applications in several areas such as electronic surveillance systems, military communications, eavesdropping devicesand satellite systems.
There are two Principles can be used to develop these systems, one of them is the decision-theoretic and the other is the pattern recognition. In this thesis, the pattern recognition method has been used to identify the modulation order of the received modulated signal. This method relies on two basic processes, namely the features extraction and classification processes.
The Mel-frequency cepstral coefficient (MFCC) method is considered one of the efficient features extraction methods, where it considered the state of the art especially in voice recognition systems but it has a disadvantage that its coefficientsare not robust in noisy mediums.
To improve the robustness of MFCCs different discrete transformations are applied on the signal before applying MFCC, i.e., MFCCisapplied after different discrete transformations as Discrete Wavelet Transform (DWT), Discrete Sine Transform (DST) and Discrete Cosine Transform (DCT). Then a decision is taken about which discrete transform to extract distinct features of the signal and to improve MFCC robustness.For the classification process; which represents the most important parts in the automatic identification systems, the Support Vector Machine (SVM) andArtificial Neural Network (ANN) are used. The advantage of this system is to recognize the modulation order without the need of prior knowledge of any information of the signal on the side of the transmitter.
Different modulation techniques have been used to modulatethe signals depending on the application which are used for. The orderidentification of the modulation scheme is a bit of a challenge, especially in the case of multipath fading. This thesis presents a suggested modulation recognition technique over wireless channels with both Orthogonal Frequency Division Multiplexing (OFDM) and Multi-Carrier Code Division Multiple Access (MC-CDMA). A comparative study between OFDM and MC-CDMA about theirmodulation recognition rate is presented in the thesis to get the best combination of the features extraction method and classification method that can be used to obtain the best recognition rate.