Joint Estimation of I/Q Imbalance and Carrier Frequency Offset Using Virtual Carriers for OFDM Systems
Date Issued
2010
Date
2010
Author(s)
Chang, Chao-Wei
Abstract
Recently, the direct-conversion transceiver for orthogonal frequency division multiplexing (OFDM) systems become more appealing because of its significant advantages in cost, package size, and power consumption. However, the direct-conversion receivers are very sensitive to non-idealities at analog front-end, such as in-phase and quadrature-phase (I/Q) imbalance and carrier frequency offset (CFO). These non-idealities destroy the subcarrier orthogonality of OFDM systems. As a result the performance of OFDM systems will degrade severely if these non-idealities are not properly compensated. Therefore, it is an important issue to precisely estimate and compensate these mismatches.
In this thesis, the effect of I/Q imbalance and CFO in OFDM systems are first studies, and then the estimation and compensation of these parameters are considered. We propose a semi-blind method for joint estimation of I/Q imbalance and CFO. In many practical communication systems, some carriers are not used for data transmission and instead zeros are transmitted on these carriers. These carriers are known as virtual carriers. In this thesis, we will exploit the virtual carriers for joint estimation of I/Q imbalance and CFO. The method is semi-blind because we only need to know the locations of these virtual carriers. Using these virtual carriers, we can improve the bandwidth efficiency without usage of any training sequences or pilots. By minimizing the cost function, the I/Q imbalance and CFO parameters can be estimated accurately. It is shown by simulation results that the performance of the proposed methods is better than other existing methods that also use virtual carriers.
Subjects
OFDM
I/Q Imbalance
Carrier Frequency Offset
Virtual Carrier
Type
thesis
File(s)![Thumbnail Image]()
Loading...
Name
ntu-99-R97942037-1.pdf
Size
23.32 KB
Format
Adobe PDF
Checksum
(MD5):02a64941939dd593db48578106a5b0d0