Sensing Signal Identification for Predictive Spectrum Sharing
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing wireless communication and sensing technologies fail to provide full awareness of future spectrum usage, leading to inefficient resource allocation and interference between different devices sharing the spectrum, particularly in decentralized networks with varying sensing and communication applications.
Innovation Solution
Implementing methods and devices for identifying and predicting sensing applications based on signal periodicity, frame structure, and other features to optimize resource allocation and minimize interference.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If devices share spectrum without sensing application identification, then spectrum availability is increased, but interference between devices increases
Solution Approach 1:
The system performs preliminary sensing application identification and future occupancy prediction before devices transmit data. By analyzing signal characteristics (periodicity, frame structure) and predicting future spectrum usage, devices can proactively schedule transmissions to avoid interference while maintaining spectrum sharing
Solution Approach 2:
The system continuously monitors spectrum occupancy and updates predictions based on observed patterns. Devices receive feedback about predicted future occupancy and adjust their transmission schedules accordingly, creating a closed-loop system that reduces interference while maintaining high spectrum utilization
2Productivity
If spectrum occupancy prediction is implemented, then resource allocation efficiency is improved, but computational complexity increases
Solution Approach 1:
The system uses lightweight machine learning models that can be trained offline and deployed as compact inference engines. These models process spectrum occupancy data with minimal computational resources, enabling prediction functionality on resource-constrained devices without requiring complex real-time computation
Solution Approach 2:
Computational intensive training of prediction models is performed offline in advance. During runtime, only lightweight inference is required, significantly reducing the computational burden on devices while maintaining high prediction accuracy for resource allocation decisions
3Measurement precision
If sensing signal transmission is continuous, then sensing accuracy is maintained, but power consumption increases
Solution Approach 1:
Devices transmit sensing signals periodically rather than continuously, synchronized with predicted future occupancy patterns. This periodic transmission maintains sufficient sensing accuracy for detecting presence and motion while dramatically reducing power consumption compared to continuous transmission
Solution Approach 2:
Future spectrum occupancy is predicted in advance using machine learning models. This prediction information is used to schedule sensing signal transmissions only when spectrum is predicted to be available, avoiding both continuous transmission and random access, thereby optimizing the trade-off between sensing accuracy and power consumption
Data Source
Figure 1~2
Figure 3~4
Figure 5
AI summary
Some embodiments in the present disclosure relate to sensing application identification. In particular, a wireless signal is obtained to be further processed. Then, presence of a sensing signal in the received signal is estimated, the sensing signal being a signal generated by a sensing application. Based on the estimation of the sensing signal, wireless reception, transmission or sensing is then performed. The sensing application detection and/or identification may be performed by a trained module such as machine learning based module. The wireless reception, transmission or sensing performed according to the result of the estimation may further include channel access, resource allocation, utilization of the detected sensing signal for own sensing purposes, or the like.