Digital quadrature compensators scheme for analog imperfections of quadrature modulator in wireless communication systems

Physical imperfections are unavoidable noise existed in electronic devices. These imperfections are usually caused due to the tolerance of electronic components. Like other electronics devices, wireless communication systems are not immune to such imperfections. These imperfections cause high energy consumption as well as degradation in efficiency. In cases of wireless transmission devices these imperfections are called “analog imperfections”. Analog imperfections in transmitters are caused by inaccuracy of a component called quadrature modulator (QM). QM is responsible for creating 90 degree phase differences between real and imaginary parts of the input signal, however, in real application QM cannot create the exact required phase difference. This will result in the phase and gain imbalances in the transmitter. To solve this issue, a compensator block is added to the transmission systems. The compensator is responsible for compensating the imperfections. In order for the compensator to do its job, imperfections should be estimated. There are two methods available for channel estimation: to use no extra physical feedback, or to use an extra physical feedback. Adaptive algorithms are applied in the design of active noise control systems. Adaptive algorithms used to estimate the QM’s imperfections, operate either in online or offline mode. Online mode describes the condition in which the adaptive algorithm is operating in parallel with the transmission process. Offline on the other hand, is a mode where adaptive algorithms cannot estimate the imperfections in parallel with the transmission. This study proposes an efficient estimation and a pilot-free compensation scheme for frequency independent I/Q imbalances in the broadband direct up conversion transmitter (DCT). For such a transmitter, the main challenge is to have the online compensation while normal transmission is in order and no dedicated pilot signal is available. I/Q imperfections –or analog imperfections- have devastating effects on the efficiency of DCT. Two different imperfection exist in each branch of I and Q while there is only one feedback existing for the compensation. The issue of only one feedback availability degrades the accuracy of channel estimation causing the final estimated values to have higher level of error. It will also result in increasing the number of iteration required to estimate the channel values In the proposed schemes, the problem of channel estimation accuracy for an adaptive algorithm named “least mean square algorithm” (LMS) with one feedback is avoided via a two feedback scheme. The proposed scheme provides two identical error function for each branch of I and Q separately. Adding extra feedback increased the speed of convergence as well as accuracy compared to other studies. The normalize mean square error (NMSE) of adaptive algorithm in other studies was -68dB and -59dB whereas for the proposed two feedbacks method it is -79.8 dB. The NMSE error is calculated for 128 numbers of iterations as well as 3% gain and °3 phase imbalance in QM. Further the performance of the proposed scheme is compared with other studies in terms of accuracy.

Preview Text FK 2016 103 - IR.pdf Download (1MB) | Preview

Actions (login required)

View Item