一种多波束卫星预编码方法及装置
By dynamically predicting the optimal modulus calculation frame size using a residual multilayer perceptron network architecture, the performance degradation caused by a fixed modulus calculation frame size is solved, achieving efficient nonlinear precoding in complex satellite channel environments and improving system performance.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- BEIJING UNIV OF POSTS & TELECOMM
- Filing Date
- 2025-10-27
- Publication Date
- 2026-07-17
AI Technical Summary
In existing nonlinear precoding techniques, the fixed size of the modulus calculation frame makes it impossible to adapt to dynamically changing satellite channel environments. This results in severe modulus loss at low signal-to-noise ratios and significant power loss at high signal-to-noise ratios, affecting the system's bit error rate performance.
A network architecture based on residual multilayer perceptron is adopted. The optimal size of the modulus calculation operation box is dynamically predicted through dedicated preprocessing of the input layer, dynamic adaptive mechanism of ResMLP layer and physical constraints of output layer. Accurate feedback is achieved with limited feedback resources through high-precision feedback index table.
It significantly improves the performance of nonlinear precoding in complex channel environments, effectively eliminates inter-beam interference, reduces noise impact, and improves the system's bit error rate performance.
Smart Images
Figure CN121441373B_ABST