Signal Demodulation Phase Noise Compensation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In LTE systems, phase noise independent of Gaussian white noise affects signal demodulation, leading to reduced precision and efficiency, particularly at higher Modulation and Coding Scheme (MCS) levels, as existing methods either ignore phase noise or degrade receiver performance when accounting for it.
Innovation Solution
A method and apparatus that model the received signal with phase noise as a uniform distribution, creating constellation point probability models and likelihood probability ratios, and applying phase compensation using a complex exponential function to improve demodulation accuracy and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If phase noise is ignored in the demodulation process (first scheme), then the calculation is simple and fast, but the LLR precision deteriorates
Solution Approach 1:
The patent extracts phase noise as a separate factor from the received signal model. By explicitly modeling phase noise as an independent random variable θ with uniform distribution, the system separates phase noise effects from Gaussian noise, allowing for targeted compensation in the LLR calculation process while maintaining computational efficiency.
Solution Approach 2:
The patent changes the statistical distribution parameters by modeling phase noise as a uniform distribution over (-π, π) rather than assuming Gaussian distribution. This parameter change enables the system to accurately represent phase noise characteristics and improves LLR precision without significantly increasing computational complexity.
2Measurement precision
If phase noise is directly introduced to constellation points (second scheme), then LLR precision is improved, but receiver performance deteriorates and demodulation efficiency is lowered
Solution Approach 1:
The patent introduces an intermediary phase rotation factor e^(jθ) that acts as a mediator between the transmitted constellation points and the received signal. This intermediary allows the system to account for phase noise effects without directly modifying constellation points, thereby maintaining demodulation efficiency while improving precision.
Solution Approach 2:
The patent performs preliminary phase compensation by integrating over the phase noise distribution in the likelihood calculation before final LLR computation. This preliminary action removes phase noise effects early in the processing chain, preventing performance degradation while maintaining computational efficiency.
3Measurement precision
If phase noise is modeled with uniform distribution and integrated into likelihood calculation, then LLR precision is improved, but calculation complexity increases
Solution Approach 1:
The patent changes the integration variable and utilizes the uniform distribution property of phase noise to simplify the integral calculation. By transforming the likelihood function and exploiting the periodicity of the complex exponential function, the system reduces calculation complexity while maintaining precision.
Solution Approach 2:
The patent creates a simplified copy of the likelihood calculation that explicitly incorporates phase noise integration. This copied calculation structure reuses existing signal processing components while adding phase noise compensation, avoiding the need for completely new complex computation paths.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Disclosed are a method and apparatus for demodulating a signal. The method comprises: obtaining a received signal, wherein the received signal comprises a phase noise signal; establishing a likelihood probability ratio integral model on the basis of the received signal and a preset phase noise parameter, wherein the phase noise parameter represents the phase noise signal and is a random variable; performing phase rotation angle extraction and transformation processing and discretization processing on the likelihood probability ratio integral model to obtain a likelihood probability ratio discrete model, wherein the phase rotation angle represents the phase rotation angle obtained on the basis of the phase noise signal; and determining a likelihood probability ratio corresponding to the received signal on the basis of the likelihood probability ratio discrete model to obtain a demodulation result. In this way, a base station overcomes phase noise, and improves signal receiving performance, signal demodulation efficiency and signal demodulation accuracy by performing phase compensation and discrete calculation on a received signal to obtain the demodulation result.