LTE Channel Estimation Using Adaptive DCT Window Truncation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current channel estimation methods in LTE networks, particularly in the uplink receiver chain, face challenges with spectral leakage and distortion due to truncation in the discrete cosine transform (DCT) domain, which affects the accuracy of channel estimation and introduces errors in data flow.
Innovation Solution
A method and arrangement in a receiving communication node that applies a sloped window truncation function of adaptive length in the DCT domain, calculated based on channel state information such as RMS delay spread and SNR, to improve channel estimation accuracy by smoothing the channel response.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If truncation is applied in DCT domain to improve signal to noise ratio, then noise is reduced, but spectral leakage and distortion are introduced
Solution Approach 1:
A transition matrix is introduced as an intermediary between the DCT domain and frequency domain. This transition matrix enables smooth transformation and energy redistribution, acting as a mediator that connects the truncation operation with the final channel estimate while reducing spectral leakage effects
Solution Approach 2:
The patent transforms the channel estimate from DCT domain to frequency domain using a transition matrix, changing the domain parameter. This parameter change allows the system to exploit the noise reduction benefits of DCT truncation while converting back to the original frequency domain representation to minimize spectral leakage
2Ease of manufacture
If truncation is applied to keep only first taps in DCT domain, then processing is simplified and signal to noise ratio is improved, but channel energy is sacrificed
Solution Approach 1:
The patent applies a domain transformation parameter change by converting the truncated DCT domain channel estimate back to the frequency domain using a transition matrix. This parameter change allows the system to maintain processing simplicity while recovering channel energy that would otherwise be lost in truncation
3Device complexity
If conventional truncation is used, then implementation is simple, but accuracy of channel estimation deteriorates with larger delay spread
Solution Approach 1:
The transition matrix serves as an intermediary that connects the simple truncation operation with improved channel estimation accuracy. By introducing this intermediate transformation step, the system maintains implementation simplicity while achieving better estimation performance through energy redistribution
Solution Approach 2:
The patent changes the domain parameter from DCT domain to frequency domain through a transition matrix transformation. This parameter change enables the system to achieve better channel estimation accuracy for channels with larger delay spread while keeping the overall implementation relatively simple
Data Source
AI summary
The invention relates to a method in a receiving communication node for performing channel estimation on a signal received over a channel from a transmitting communication node. The receiving and transmitting communication nodes are comprised in a radio communications network. The receiving communication node receives the signal from the transmitting communication node, filters the signal by applying matched filter to the received signal in a matched filter process resulting in an estimated channel, transforms the estimated channel from a frequency domain to a discrete cosine transform domain in a discrete cosine transformation process, truncates the transformed estimated channel by applying a sloped window truncation function of an adaptive length onto the transformed estimated channel, wherein the adaptive length in the discrete cosine transform domain is calculated based on channel state information, and transforms the truncated transformed estimated channel from the discrete cosine transform domain to the frequency domain in an inverse discrete cosine transformation process resulting in a channel estimate of the channel estimation.