AI Signal Detection Receiver Using Dual-Domain Transformer Attention
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wireless communication systems face challenges in accurately detecting signals in the terahertz band due to severe path loss and atmospheric absorption, necessitating improved technologies for securing signal transmission distance and coverage, especially in 6G communication systems.
Innovation Solution
Implementing a transformer-based neural unit in signal receivers that includes a feature embedding block and multiple transformer blocks for calculating log likelihood ratios, utilizing multi-head attention modules in both spatial and resource element domains to enhance signal detection accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If transformer-based neural unit with multi-head attention modules is implemented, then signal detection accuracy is improved, but device complexity increases
Solution Approach 1:
The transformer-based neural unit is divided into multiple transformer blocks, each containing specialized multi-head attention modules (spatial domain MHA and resource element domain MHA). This segmentation allows the complex detection task to be broken down into manageable modules that can be processed independently, improving accuracy while managing complexity through structured organization.
Solution Approach 2:
The patent introduces multi-head attention modules that operate in multiple domains (spatial domain and resource element domain) rather than a single domain. This dimensional expansion allows the system to capture relationships from different perspectives simultaneously, significantly improving signal detection accuracy by considering both spatial correlations and resource element correlations.
2Productivity
If dual multi-head attention modules operating in spatial and resource element domains are used, then signal detection performance is enhanced, but computational complexity increases
Solution Approach 1:
The computational workload is segmented into two distinct multi-head attention modules: one for spatial domain processing and another for resource element domain processing. Each module focuses on specific types of correlations, dividing the overall computational task into specialized sub-tasks that can be optimized independently, thereby enhancing detection performance while managing computational complexity.
Solution Approach 2:
The transformer blocks are designed to perform multiple functions through their dual multi-head attention modules - they simultaneously process spatial relationships and resource element relationships within a unified architectural framework. This multi-functionality allows the same computational structure to handle different types of signal correlations without requiring separate processing paths.
Data Source
AI summary
The present disclosure relates to a 5G communication system or a 6G communication system for supporting higher data rates beyond a 4G communication system such as long-term evolution (LTE). A method performed by a receiver in a wireless communication system according to embodiments of the disclosure may include: estimating a channel, based on a received signal; embedding the channel; performing spatial domain attention calculation for acquiring a cross-covariance value of a spatial domain, based on the embedded channel; performing resource element attention calculation for acquiring a cross-covariance value of a resource element domain, based on the embedded channel; and calculating a log likelihood ratio (LLR), based on results of the spatial domain attention calculation and the resource element attention calculation.