Signal Interference Cancellation via Adaptive Gradient Tuning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current RF cancellers in 5G and 6G wireless systems lack the tuning speed required to effectively suppress self-interference, leading to inefficiencies in signal transmission and reception due to slow convergence to optimal settings.
Innovation Solution
The use of an effectiveness gradient, tunable aggressiveness factor, and forgetting factor to generate coefficients for the signal interference canceller, allowing for iterative tuning and improved cancellation effectiveness.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional RF cancellers use traditional tuning algorithms (DLS, LMS), then they can provide sufficient cancellation performance, but they converge to optimal settings slowly causing errors and inefficiencies
Solution Approach 1:
The patent changes the parameter update rule by introducing an effectiveness gradient-based step factor that adapts dynamically. The step factor is calculated as: step factor = (cancellation error) × (tunable coefficient step aggressiveness factor) × (time-based forgetting factor). This parameter change enables faster convergence by adjusting the step size based on real-time cancellation effectiveness measurements.
Solution Approach 2:
The patent implements a feedback mechanism where the cancellation error is continuously measured and used to adjust the step factor in real-time. The feedback loop includes: measuring cancellation error → calculating step factor based on error and effectiveness gradient → updating cancellation coefficients → applying new coefficients to the canceller. This closed-loop feedback accelerates convergence compared to traditional open-loop or slow-converging algorithms.
2Productivity
If the step factor is increased to speed up convergence, then tuning time is reduced, but cancellation precision may deteriorate due to overshooting optimal settings
Solution Approach 1:
The patent makes the step factor dynamic rather than static. The step factor changes over time based on the cancellation error magnitude and the effectiveness gradient. Initially, when error is large, the step factor is larger for fast convergence. As the system approaches optimal settings and error decreases, the step factor automatically reduces to prevent overshooting. This dynamic adjustment resolves the contradiction between tuning speed and precision.
Solution Approach 2:
The patent employs periodic measurement and adjustment cycles. The cancellation error is measured periodically, and the step factor is updated in each cycle based on the current error and effectiveness gradient. This periodic feedback allows the system to adapt the step size at different stages of convergence, maintaining both speed and precision through rhythmic adjustment cycles.
3Productivity
If traditional cancellation algorithms are used, then system complexity is kept low, but real-time suppression capability is insufficient for 5G and 6G wireless systems
Solution Approach 1:
The patent replaces traditional mechanical/conventional algorithms (DLS, LMS) with a gradient-based optimization approach that uses mathematical effectiveness gradients to directly compute optimal coefficients. This substitution of the optimization mechanism enables real-time performance required by 5G/6G systems while managing complexity through efficient computation rather than hardware complexity.
Data Source
AI summary
A system for canceling signal interference (SI) includes a transceiver configured to concurrently transmit signals and receive signals within a single frequency band, which causes signal interference between the transmitted and received signals. The SI canceller is configured to use a set of cancellation coefficients to generate a cancellation signal to mitigate the SI. The system is configured to iteratively change the cancellation coefficients by a step factor to produce tuned cancellation coefficients. The step factor is determined by a cancellation error gradient and one or more of: a tunable coefficient step aggressiveness factor; and a time-based forgetting factor; and cause the SI canceller to cancel the SI using the tuned cancellation coefficients.