Piecewise demodulation based on combined artificial neural network for quadrate frequency shift keying communication signals
Küçük Resim Yok
Tarih
2020
Yazarlar
Dergi Başlığı
Dergi ISSN
Cilt Başlığı
Yayıncı
Springer
Erişim Hakkı
info:eu-repo/semantics/closedAccess
Özet
The Quadrate Frequency-shift Keying (QFSK) modulation is one of the most widely used modulation methods for transmitting base band signal in the transition band in digital communication. It is a very common modulation type for fast and easy transmission of data, especially in wireless communication. In QFSK modulation, 4 separate carriers are used, and since each carrier represents 2 bits, the sending speed doubles according to classical digital modulation. In this study, firstly, with QFSK modulation, 8-bit information-bearing module signals have been obtained. Moreover, the theory has been developed on the demodulation of the QFSK module signal. The aim is to get the base band signal again. For this purpose, all data from 0–255 to 8-bit decimal length has been obtained as QFSK modulated. SNR = 5 dB–10 db–15 dB–20 dB noise has been added to the QFSK signal in order to examine the system performance during the transmission of the modulated signal. The generated signals are given to the developed demodulation system using different foldings. The matrix consisting of QFSK modulation data to be trained in the network of ANN in demodulation is divided into 4 parts. Each piece represents two bits of data. In the case of a column matrix, 4 parts have been applied as an introduction to the neural network model. In the output of ANN, the result matrix to be estimated has been formed. Trained with 5 dB noisy QFSK data of 10-layer network, application of other QFSK signal data to the network for testing, base band data of the signal has been obtained and 100% performance has been achieved. Each piece of modulation signal applied to the ANN network input is classified between 0–3 at the output. The modulation data that carries 8 bits are applied to the network in 4 steps and classified. 4 separate classification data from ANN output are converted to 2 bits of logic. 8-bit demodulation data is obtained from 4 steps. After the formation of the ANN network, the base band digital signal estimation is performed quickly in 4 steps through each byte QFSK modulation signal under different noise coming to the network.
Açıklama
Anahtar Kelimeler
Demodulation, Digital Communication, Digital Modulation, Piecewise ANN, QFSK Modulation
Kaynak
Lecture Notes on Data Engineering and Communications Technologies
WoS Q Değeri
N/A
Scopus Q Değeri
Q3
Cilt
43
