Autoencoder Neural Network Signal Transmission
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current wireless communication systems face inefficiencies in signal transmission and reception due to complex neural network configurations, which increase the complexity of configuring autoencoders for efficient data processing in wireless communication systems.
Innovation Solution
The method involves configuring a transmitter encoder neural network and a receiver decoder neural network using a sparsely-connected neural network structure, where each activation function receives only some input values, and the neural network configuration unit includes a first and second activation function that apply weights to input values through specific paths, reducing the complexity of autoencoder configuration and improving efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a complex neural network configuration is used for autoencoder, then signal transmission and reception efficiency is improved, but device complexity increases
Solution Approach 1:
The neural network is divided into multiple configuration units, each handling a specific subset of input values. Each configuration unit contains activation functions that process only certain input values, segmenting the overall complex task into manageable smaller units. This segmentation reduces the complexity of configuring the entire neural network while maintaining the efficiency benefits of deep learning for signal transmission and reception.
Solution Approach 2:
Instead of requiring all activation functions to process all input values (full action), the patent applies partial action where each activation function processes only a subset of input values. This partial action approach reduces the total number of connections and parameters that need to be configured, thereby reducing device complexity while still achieving effective signal processing through the collective action of all configuration units.
2Reliability
If each activation function receives all input values, then computational completeness is ensured, but configuration complexity increases
Solution Approach 1:
The input values are segmented and distributed to different activation functions within configuration units. Each activation function receives a specific subset of input values rather than all inputs, which reduces the configuration complexity while the collective segmentation across multiple units ensures computational completeness through parallel processing of different input subsets.
Solution Approach 2:
Multiple configuration units work together to provide universal coverage of all input values. While each individual activation function has a limited scope (receiving only some input values), the system as a whole achieves universality by having different configuration units process different portions of the input space, ensuring all computational requirements are met without increasing individual unit complexity.
Data Source
AI summary
The present specification provides a method for transmitting/receiving a signal in a wireless communication system by using an auto encoder. More specifically, the method performed by means of a transmission end comprises the steps of: encoding at least one input data block on the basis of a pre-trained transmission end encoder neural network; and transmitting a signal to a reception end on the basis of the encoded at least one input data block, wherein each of activation functions included in the transmission end encoder neural network receives only some of all input values that can be input into each of the activation functions.