University of Oulu

Pilot sequence based IQ imbalance estimation and compensation

Saved in:
Author: Theisen Rodrigues Pinto, Enrique1
Organizations: 1University of Oulu, Faculty of Information Technology and Electrical Engineering, Communications Engineering
Format: ebook
Version: published version
Access: open
Online Access: PDF Full Text (PDF, 3.4 MB)
Pages: 82
Persistent link:
Language: English
Published: Oulu : E. Theisen Rodrigues Pinto, 2023
Publish Date: 2023-06-21
Thesis type: Master's thesis (tech)
Tutor: Juntti, Markku
Tapio, Visa
Reviewer: Juntti, Markku
Tapio, Visa


As modern radio access technologies strive to achieve progressively higher data rates and to become increasingly more reliable, minimizing the effects of hardware imperfections becomes a priority. One of those imperfections is in-phase quadrature imbalance (IQI), caused by amplitude and phase response differences between the I and Q branches of the IQ demodulation process. IQI has been shown to deteriorate bit error rates, possibly compromise positioning performance, amongst other effects. Minimizing IQI by tightening hardware manufacturing constraints is not always a commercially viable approach, thus, baseband processing for IQI compensation provides an alternative.

The thesis begins by presenting a study in IQI modeling for direct conversion receivers, we then derive a model for general imbalances and show that it reproduces the two most common models in the bibliography. We proceed by exploring some of the existing IQI compensation techniques and discussing their underlying assumptions, advantages, and possible relevant issues.

A novel pilot-sequence based approach for tackling IQI estimation and compensation is introduced in this thesis. The idea is to minimize the square Frobenius norm of the error between candidate covariance matrices, which are functions of the candidate IQI parameters, and the sample covariance matrices, obtained from measurements. This new method is first presented in a positioning context with flat fading channels, where IQI compensation is used to improve the positioning estimates mean square error. The technique is then adapted to orthogonal frequency division multiplexing (OFDM) systems,including an version that exploits the 5G New Radio reference signals to estimate the IQI coefficients. We further generalize the new approach to solve joint transmitter and receiver IQI estimation and discuss the implementation details and suggested optimization techniques.

The introduced methods are evaluated numerically in their corresponding chapters under a set of different conditions, such as varying signal-to-noise ratio, pilot sequence length, channel model, number of subcarriers, etc. Finally, the proposed compensation approach is compared to other well-established methods by evaluating the bit error rate curves of 5G transmissions. We consistently show that the proposed method is capable of outperforming these other methods if the SNR and pilot sequence length values are sufficiently high. In the positioning simulations, the proposed IQI compensation method was able to improve the root mean squared error (RMSE) of the position estimates by approximately 25 cm. In the OFDM scenario, with high SNR and a long pilot sequence, the new method produced estimates with mean squared error (MSE) about a million times smaller than those from a blind estimator. In bit error rate (BER) simulations, the new method was the only compensation technique capable of producing BER curves similar to the curves without IQI in all of the studied scenarios.

see all

Copyright information: © Enrique Theisen Rodrigues Pinto, 2023. This publication is copyrighted. You may download, display and print it for your own personal use. Commercial use is prohibited.