Acta Optica Sinica, Volume. 32, Issue 11, 1106001(2012)
Blind Optical Baseband Signals Detection Using Recurrent Neural Network Based on Continuous Multi-Valued Neurons
Spatial diversity reception can be used over wireless optical links to resist channel fading. A novel blind optical baseband signals detection direct algorithm is proposed by using recurrent neural network based on continuous multi-valued neurons (RNNCMVN) with amplitude and phase continuous activation (APCA). Considering the characteristics of mary phase shift keying (MPSK) signals, two types of continuous multi-valued activation functions are designed and the method of selecting their parameters is illustrated briefly. The new energy functions of synchronous and asynchronous mode of the RNNCMVN are derived and proved, respectively. Considering the characteristics of quadrature amplitude modulation (QAM) signals, the new continuous amplitude and phase multi-valued activation functions are designed and discussed, respectively. Finally, simulation results demonstrate the effectiveness of the proposed method.
Get Citation
Copy Citation Text
Ruan Xiukai, Zhang Yaoju. Blind Optical Baseband Signals Detection Using Recurrent Neural Network Based on Continuous Multi-Valued Neurons[J]. Acta Optica Sinica, 2012, 32(11): 1106001
Category: Fiber Optics and Optical Communications
Received: May. 9, 2012
Accepted: --
Published Online: Aug. 8, 2012
The Author Email: Xiukai Ruan (ruanxiukai@wzu.edu.cn)