RFDAC Sampling Frequency Shift for LO Noise and Replica Suppression
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Communication systems face noise generation issues due to errors in local oscillator (LO) signal generation, leading to degraded in-band and out-of-band performance, particularly in modern cellular systems like 4G/LTE-LTE-CA and 5G/NR, caused by remodulation effects and replicas of baseband signals.
Innovation Solution
The implementation of an enhanced sampling frequency in the radio frequency digital to analog converter (RFDAC) system, which upsamples the baseband signal to twice the carrier frequency, reducing self-pulling effects and suppressing even replicas of the digital baseband signal, thereby improving quantization noise-floor and performance metrics such as EVM and ACLR.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If the baseband signal is sampled at the carrier frequency in a conventional RFDAC, then the system complexity is low, but the quantization noise-floor is high and replicas of the baseband signal are generated
Solution Approach 1:
The patent changes the sampling frequency parameter from the conventional carrier frequency (fc) to twice the carrier frequency (2fc). This parameter change directly improves the quantization noise-floor by 3 dB and suppresses even replicas of the baseband signal, while the complexity increase is managed through efficient implementation in the RFDAC circuitry
Solution Approach 2:
The patent applies excessive action by sampling at a frequency higher than the minimum required Nyquist rate. By sampling at 2fc instead of fc, the system oversamples the baseband signal, which provides the benefit of improved noise-floor and replica suppression while maintaining practical system complexity through optimized hardware design
2Reliability
If the sampling frequency is increased to suppress replicas and improve noise-floor, then the performance metrics (EVM, ACLR) are improved, but the device complexity and power consumption increase
Solution Approach 1:
The patent implements a specific parameter change to twice the carrier frequency (2fc) rather than arbitrarily increasing the sampling rate. This targeted parameter change achieves the desired performance improvement in EVM and ACLR while limiting the complexity increase to a manageable factor of two in the sampling frequency
Solution Approach 2:
The patent segments the RFDAC operation into distinct phases corresponding to the two sampling periods within one carrier cycle. This segmentation allows the system to process the baseband signal at twice the frequency while maintaining clear temporal separation of operations, thereby managing complexity through structured time-multiplexed implementation
Data Source
AI summary
A communication system using an increased sampling rate includes a baseband signal generator and a PLL that generates an LO signal and an inverted LO signal. An interpolator generates an interpolated/delayed data stream from a baseband original data stream, and a cell control circuit generate control signals based on the interpolated data stream and the baseband original data stream. An RFDAC generates an RF output signal from the baseband signal using an increased sampling rate, the LO signal, the inverted LO signal, and the control signals.