A neural network-assisted iterative reception method for high-speed visible light communication

Abstract

The present invention provides a neural network-assisted iterative receiving method suitable for high-speed visible light communication. The purpose of the solution is to modify soft decision information through a neural network receiver to eliminate the influence of clipping distortion. The neural network BICM receiver includes: (1) the cost function adopts the cross entropy function; (2) the training method adopts the backpropagation algorithm; (3) the input layer includes the real part, the imaginary part and the corresponding Gaussian noise of the accepted symbol, and the output layer Get the probability value of each constellation point; (4) The hidden layer and output layer use tanh and softmax functions respectively. Compared with the existing clipping enhancement scheme, the present invention has more obvious suppression effect and reasonable algorithm complexity, and is beneficial to enhance the robustness of DCO‑OFDM against clipping distortion.

Description

technical field [0001] The present invention relates to the field of visible light communication, in particular to a neural network-assisted iterative receiving method suitable for high-speed visible light communication. Background technique [0002] As a supplement to the fifth generation communication system (5G), visible light communication has become a promising indoor short-range communication technology. Visible light communication uses intensity modulation direct detection to drive light-emitting diodes (LEDs), which can realize data communication and indoor lighting at the same time. In recent years, Optical Orthogonal Frequency Division Multiplexing (O-OFDM) has attracted widespread attention due to its higher spectrum utilization and anti-multipath effects. O-OFDM technology includes asymmetric clipped optical OFDM (ACO-OFDM), pulse amplitude modulated discrete multi-tone (PAM-DMT) and DC biased optical OFDM (DCO-OFDM) and other candidate schemes. Compared with o...

AI Technical Summary

Problems solved by technology

Nevertheless, traditional BICM receivers still suffer from severe distortion, the reason is that the clipping distortion does not take into account the maximum a posteriori soft demodulation process, still adopts the Gaussian distribution as the channel conditional probability, resulting in the logarithm of the mismatch likelihood ratio

For nonlinear distortion scenarios, research institutions have proposed such as maximum likelihood sequence detection BICM receiver (MLSD-BICM), Gaussian mixture model BICM receiver (GMM-BICM), etc., but the performance of the two is still far from the Shannon limit.

Images