Detection and Extraction of OFDM Parameters Using Difference of Gaussians

Signals type detection is very important in telecommunication. Telecommunication signals can be divided into two major groups: single-carrier signals and multi-carrier signals. The first step in extracting data in multicarrier communication signals is to detect signals and their subcarriers. OFDM signals are one of the most popular multi-carrier signals that are used widely. This paper will introduce a blind detection method for OFDM signals, subcarriers, and the central frequency of them based on the Difference of Gaussians (DoG) technique which is applied for blob detection in machine vision. Performance of our method is compared with high-resolution spectral estimation such as Capon, Borgiotti-Lagunas, and MUSIC. Results showed that it has less computational complexity than the others. Also, there is no need to learn parameters, so the response time of the system is appropriate. Furthermore, many tests have been done on real and artificial signals corrupted with noise and fading and the results showed our proposed method has better performance and cause the lower error in the severe condition like SNR=0.