The evolution of electronics allows nowadays the usage of the unused, free frequency bands at 10-100 GHz for commercial purpose. One of these systems is the WiGig (and the IEEE 802.11ad standard, which describes the WiGig) which operates at the yet rarely used 60 GHz frequency band to ensure high data transfer rate. The standard aims that the WiGig can alone substitute all wired and wireless transmission between indoor devices, within a 10-15 m radius.
The 802.11ad standard only specifies the behavior of the transmitter, so that is not obvious to choose the synchronization methods for processing and compensating the received data. At the beginning of my work, I describe the transmitter as well, which was implemented based on the standard. I create a coded, modulated data stream which can be sent to the analog frontend of the transmitter.
The model for the radio channel, which I applied in my work, contains different failures and distorting effects. The parameters of these errors have to be estimated before processing the received data at the receiver-side to compensate the influences of the imperfections. These considered errors are the following: timing error, frequency and phase offset. Furthermore I take into consideration, that the radio channel has linear distortion and contains additive gaussian noise as well. There are already many solution in the literature to estimate and compensate these errors. I investigate these methods in details and propose improvements to get more accurate estimations e.g. estimators with lower variance.
At the receiver side, the signal processing steps of the transmitter have to be performed in a reverse order. The data are demodulated first, then decoded so the originally transmitted binary data can be retried. If the received signal containing errors is not compensated before the demodulation, the bit error rate may drastically increase. In case the estimates of the errors are known, the received signal can be compensated through a multi-stage process to decrease the bit error rate.
The simulation environment of the transmission chain was implemented using Matlab, I created a modular framework. With this simulator I prepared first the transmitter-receiver connection, then the model of the transmission channel was expanded step by step. The receiver was extended with the estimation and the correction algorithms. Using this framework the errors can be easily handled, and the investigated methods can be compared quantitative as well.
After the simulations, the methods were tested in practice as well. The test environment for the measurements consisted of a suitable arbitrary signal generator and an high performance oscilloscope. The signal generator transmitted the loaded data samples, which were received and stored by the oscilloscope. The data was retrieved, processed and evaluated using the previously mentioned Matlab framework, also the estimation and compensation algorithms were verified.