Synchronous Bus Capture Timing for DPD Sample Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing RF transceiver systems face inefficiencies due to the computationally intensive task of aligning pre-distorted samples and digital feedback signals, which degrades DSP performance, especially in multiple-input multiple-output (MIMO) systems with multiple antennas, due to the delay in the feedback path and the need for correlation calculations.
Innovation Solution
A system that includes a sample counter module, programming interface module, and comparator module to generate a dynamic count value for initiating the storage of output samples, reducing the number of computing cycles required for alignment by determining the starting position and sample boundaries of the output samples in memory.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If correlation calculations are performed to align pre-distorted samples and digital feedback signals, then alignment accuracy is improved, but computational complexity and power consumption increase
Solution Approach 1:
The patent calculates and stores delay values in advance based on buffer positions before actual signal processing. By pre-computing the delay information and storing it in memory, the system avoids performing computationally intensive correlation calculations during real-time operation, thereby reducing power consumption while maintaining alignment accuracy.
Solution Approach 2:
The system pre-determines the starting position and sample boundaries of output samples in memory before correlation processing. This preliminary organization of data allows the DSP to perform alignment operations more efficiently without requiring extensive computational resources during the actual correlation calculation phase.
2Measurement precision
If correlation calculations are performed to align pre-distorted samples and digital feedback signals, then alignment accuracy is improved, but DSP performance deteriorates
Solution Approach 1:
Delay values are pre-calculated and stored based on buffer positions before the DSP needs to perform alignment operations. This eliminates the need for real-time correlation calculations during signal processing, significantly reducing the computational burden on the DSP and improving overall system productivity while maintaining alignment accuracy.
Solution Approach 2:
The patent extracts and stores delay information separately from the main signal processing flow. By isolating the delay calculation and storing it in memory, the system removes the computationally intensive correlation operations from the DSP's real-time workload, allowing the DSP to focus on other critical processing tasks and improving overall performance.
3Measurement precision
If feedback path delay is compensated through correlation, then alignment accuracy is improved, but the number of computing cycles increases
Solution Approach 1:
The system pre-calculates delay values corresponding to different buffer positions and stores them in memory before they are needed. When alignment is required, the system simply retrieves the pre-computed delay value based on the current buffer position, avoiding the need to perform time-consuming correlation calculations during operation and reducing the number of computing cycles required.
Data Source
AI summary
A system for storing pre-distorted output samples in a memory includes a sample counter, a programming interface module, and a comparator. The sample counter counts the pre-distorted output samples, generates a dynamic count value, receives a capture counter status signal, and generates a first count value. The programming interface module receives and outputs the first count value, an offset value, and a capture control signal, and generates a first interrupt signal. The comparator receives the first count value, the offset value, the dynamic count value, and the capture control signal, generates a final value, compares the final value with the dynamic count value, and generates a trigger signal when the final value equals the dynamic count value based on the capture control signal. The trigger signal initiates the storing of the pre-distorted output samples in the memory.