System level modeling and evaluation of advanced linear interference aware receivers

摘要

To cope with the growth of data traffic through mobile networks, efficient utilization of the available radio spectrum is needed. In densely deployed radio networks, User Equipments (UE) will experience high levels of interference which limits the achievable spectral efficiency. In this case, a way to improve the achievable performance is by mitigating interference at the UE side.Advanced linear interference aware receivers are linear receivers able to mitigate external co-channel interference. Optimum linear interference rejection is obtained with the Interference Rejection Combining (IRC) receiver which relies on the ideal knowledge of the interference covariance matrix. The IRC interference covariance matrix is the sum of all interference channel covariance matrices. In practical radio networks, like LTE-Advanced, the knowledge of interference channel covariance matrices might not always be available. However, the IRC interference covariance matrix estimation can be done with a data-based or reference-symbol-based interference covariance matrix estimation algorithm.In this thesis, the modeling and evaluation of advanced linear interference aware receivers for LTE-Advanced downlink are studied. In particular, the data-based and reference-symbol-based covariance matrix estimation algorithms are modeled by using the Wishart distribution. This modeling allows the evaluation of advanced linear receivers without explicit need for baseband signals. The evaluation is done with a system level simulator. Later, a comparison of performance between advanced linear interference aware receivers and 3GPP baseline linear receivers for multiple homogeneous and heterogeneous deployment scenarios is presented.Finally, it is shown that advanced linear interference aware receivers can provide spectral efficiency improvements specially to UEs located at cell borders.