Grant-Free User Detection Using Pilot and Data Symbols
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In grant-free multiple access systems, the base station cannot determine which IoT devices are actively transmitting data due to the lack of resource reservation, leading to inefficiencies in channel estimation and data detection performance.
Innovation Solution
A method and apparatus using deep learning-based active user detection (AUD) that utilizes both user data and pilot symbols for improved detection, employing a neural network to generate active user information and channel estimation, enabling effective data detection in non-orthogonal multiple access scenarios.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If grant-free multiple access is used to allow multiple IoT devices to transmit data simultaneously without resource reservation, then resource allocation efficiency is improved, but the base station cannot detect active users in advance leading to poor channel estimation and data detection performance
Solution Approach 1:
The patent applies preliminary action by performing active user detection using a neural network before channel estimation and data detection. The base station receives pilot signals from multiple users, uses a trained neural network to predict which users are active, and then performs channel estimation only for predicted active users. This preliminary detection step resolves the contradiction by enabling efficient resource usage while maintaining accurate active user identification.
Solution Approach 2:
The patent replaces traditional mechanical signal processing methods with a neural network-based machine learning system. Instead of using conventional correlation-based detection or combinatorial search methods, the invention uses a neural network that has been pre-trained with simulation data to predict active user patterns. This substitution provides better detection accuracy while maintaining the grant-free efficient resource allocation.
2Device complexity
If traditional pilot-only methods are used for active user detection, then the detection process is simple, but the activity error rate performance and data detection accuracy are insufficient
Solution Approach 1:
The patent merges multiple information sources to improve detection reliability. It combines pilot signals from multiple users, features extracted by the neural network (such as signal energy, correlation values, and temporal patterns), and prediction results from previous time slots. This merging of multiple data sources and processing steps significantly improves activity error rate performance and data detection accuracy compared to simple pilot-only correlation methods.
Data Source
AI summary
A method performed by a base station of a wireless communication system including: a step to transmit a codebook allocated to individual user terminals for data transmission; a step to receive a pilot and data from the user terminals; a step to generate an active user information based on the data; a step to detect an active user terminal among the user terminals based on the pilot and the active user information; a step to generate a channel estimation information related to the active user terminal; and a step to detect the data of the active user terminal based on the channel estimation information.