Channel feature transfer method based on recurrent generative adversarial network
A channel characteristic and channel technology, applied in the field of channel characteristic migration based on cyclic generative adversarial network, can solve the problem of missing input signal and output signal sample, etc.
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0102] This embodiment implements a method for migrating channel characteristics based on a generative adversarial network. The modulation mode of the data used is DBPSK, and the channel types used are flat fading channels and white noise channels. The implementation process of this embodiment is summarized as follows:
[0103] (1) Signal data set construction. A signal data set is generated for DBPSK modulation and flat fading channel + white noise channel.
[0104] (2) Use the convolutional neural network to realize the one-dimensional signal channel characteristic transfer each module of the GAN network.
[0105] (3) Perform the training of the channel feature transfer GAN network.
[0106] (4) Use the trained GAN network to generate channel characteristic transfer simulation signals.
[0107] (5) The quality evaluation of the channel characteristic migration simulation signal is carried out.
[0108] Each implementation link is described in detail below.
[0109] (1) ...
Embodiment 2
[0148]This embodiment implements a channel characteristic migration method based on a generative adversarial network, where the modulation mode of the data used is FSK, and the channel type used is a frequency selective channel. The implementation process of this embodiment is summarized as follows:
[0149] (1) Signal data set construction. Generate signal data sets for FSK modulation and frequency selective channels.
[0150] (2) Use the convolutional neural network to realize the one-dimensional signal channel characteristic transfer each module of the GAN network.
[0151] (3) Perform the training of the channel feature transfer GAN network.
[0152] (4) Use the trained GAN network to generate channel characteristic transfer simulation signals.
[0153] (5) The quality evaluation of the channel characteristic migration simulation signal is carried out.
[0154] Wherein, the steps (2), (3), and (4) are implemented in the same manner as in Embodiment 1, and will not be d...
PUM
Login to View More Abstract
Description
Claims
Application Information
Login to View More 


