Cascaded Hierarchical ML for PHY Layer Signal Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for coexistence and cooperation between different wireless communication systems, such as LTE and WiFi, are ineffective due to lack of intelligence for sharing bandwidth, leading to high interference and poor quality of service (QoS) and throughput, especially with the introduction of new systems like 5G NR, which lack coexistence mechanisms with existing systems.
Innovation Solution
A cascaded hierarchical machine learning (ML) technique is employed to detect and adapt to PHY and MAC layer parameters, using artificial intelligence (AI) and neural networks to learn and converge on a common MAC scheme, enabling efficient spectrum sharing and coexistence without prior knowledge of other systems' protocols.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional spectrum sensing algorithms are used, then detection of known signal types is achieved, but adaptability to new signal types and features is poor
Solution Approach 1:
The patent segments the spectrum sensing process into multiple hierarchical levels: coarse-grained detection for signal presence/absence and fine-grained detection for specific feature identification. This segmentation allows the system to maintain high detection accuracy for known signals while improving adaptability to new signal types through the coarse detection layer that can identify unknown signals for further analysis.
Solution Approach 2:
The patent implements dynamic adaptation mechanisms where the sensing system can learn and adapt to new signal types over time. The system dynamically adjusts its detection parameters and models based on observed signals, enabling it to maintain accuracy for known signals while becoming progressively better at detecting and adapting to new signal types in the environment.
2Reliability
If collaborative sensing between multiple nodes is implemented, then detection reliability is improved, but communication overhead and system complexity increase
Solution Approach 1:
The patent segments collaborative sensing into hierarchical levels where individual nodes perform local detection first, then share results through gateways. This segmentation reduces communication overhead by only transmitting essential detection information rather than raw data, while maintaining improved reliability through collaborative verification across multiple nodes.
Solution Approach 2:
The patent introduces gateway nodes as intermediaries that aggregate and process sensing information from multiple participant nodes. This intermediary layer reduces system complexity by centralizing coordination functions, filtering redundant information, and managing communication between nodes, thereby maintaining detection reliability without proportionally increasing overall system complexity.
3Adaptability or versatility
If machine learning techniques are used for feature detection, then adaptability to unknown signals is improved, but computational complexity and processing time increase
Solution Approach 1:
The patent segments machine learning processing into online and offline components. Offline training occurs during low-activity periods to build detection models without impacting real-time performance. Online inference uses pre-trained models for rapid detection. This segmentation enables the system to maintain high learning capability while minimizing processing time during critical detection operations.
Solution Approach 2:
The patent performs preliminary model training and feature extraction setup in advance during offline phases. By pre-processing and preparing detection models before they are needed for real-time sensing, the system accumulates learning capability over time without incurring processing delays during actual signal detection, thus resolving the contradiction between adaptability and processing time.
Data Source
AI summary
A method can be performed by a first node for determining a parameter of physical (PHY) layer circuitry of a second node. The method can include implementing a cascaded hierarchy of techniques to determine, based on an electrical signal from a second node, a parameter of the PHY layer circuitry of the second node, and causing an antenna of the first node to transmit an electromagnetic wave consistent with the determined parameter.