Multi-modal information assisted channel adaptive high-frequency difunctional waveform design method
By constructing a complex-value beam strabismus adaptive precoding network (CSP-Net) model, multimodal data of communication and perception channels in high-frequency band transmission systems are processed, and the optimized design of dual-function waveforms is realized, which solves the problem that dual-function waveform design in the prior art is difficult to take into account communication and perception requirements, reducing computing complexity and improving performance.
Patent Information
- Application Number
- CN202411866656.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-18
- Publication Date
- 2025-03-25
- Estimated Expiration
- 2044-12-18
AI Technical Summary
The dual-function waveform design of existing high-frequency transmission systems is difficult to take into account communication and perception needs, and fails to fully utilize the multimodal information of synesthesia channels. The integrated performance gain is limited, and the calculation complexity of the hybrid precoding process makes it difficult to deploy in actual roadside base stations.
Using a multimodal information-assisted channel adaptive high-frequency dual-function waveform design method, the complex-value beam strabismus adaptive precoding network (CSP-Net) model is constructed, and a deep neural network is used to process multimodal data of communication and perception channels, thereby achieving the optimized design of dual-function waveforms and reducing the computational complexity.
It significantly reduces the computing complexity of dual-function waveform design, improves the coordinated improvement of communication and perception performance, adapts to dynamic environments, has low cost and small hardware changes.