Design of WOLA FTT based channelizer

In the thesis study, numerical channelizer systems with widespread use in communication and electronic intelligence systems are examined and a channellizer technique is designed. As the technology advances, increasing of analog digital converters's(ADC) sampling frequency has led to the sampling of wideband signals. In addition, decreasing cost and increasing performance of digital signal processors such as field programmable gate arrays (FPGAs) and digital signal processing (DSP) processors have facilitated processing of the sampled wideband signals. For this reason, the use of DSP processors with ADC is increasing day by day. The ability to sample and process wideband signals facilitates solving some of the needs of communications and electronic intelligence systems. In systems where more than one channel must be used, different radio frequency (RF) receivers must be used for each channel. This leads to increase cost and size of the system, and make the flexibility and stability worse.In order to avoid this situation, it is necessary to sample the wideband signals that are contained in all the channels that need to be processed.all the channels are obtained independently from each other by using channelizers that separate the sampled signal into channels. As the necessity to the channelizers has occurred, many channelizer techniques have been developed. Each technique has advantages and disadvantages over other techniques. For this reason, it is important to examine the channelizer techniques correctly and to use the most appropriate channelizer technique according to the needs. The channelizer techniques that are common in use are examined in this thesis and a selected channelizer technique is designed in the MATLAB Simulink platform and their advantages and disadvantages are examined in details. In the design phase, channelizer make compatible for use with Xilinx FPGA family. A fast hardware identification language (VHDL) code of the channelizer that was designed on MATLAB Simulink is developed and synthesized on the Xilinx ISE platform. So, the resource utilization of the channelizer that was designed in the FPGA can be calculated. In addition, the maximum number of channels that can be calculated using the WOLA channelizer in different FPGA's in the market are calculated and displayed in graph. Some channelizer techniques are intended to use fast Fourier transform (FFT) processing . However, the leakages of FFT bins are not compatible to channelizer systems. For this reason, solutions have been developed to reduce these leakages. The windowing processing that is applied to the signal before the FFT operation reduces the leakages of FFT bins. In the FFT process that is applied windowing, while the level of bin leakages to the bins which are far away is low, the level of bin leakages to the bins which are near is high. Several solutions have been developed to reduce the level of bin leakages to the bins which are near. One of these solutions is the WOLA technique. This technique allows FFT processing to be used more efficiently in channelizer. The WOLA FFT channelizer was used to design channelizer that has 32-channels. The channelizer performance is measured by giving diffrerent types of signals to channelizer input. As input to the channelizer, The input to the channelizer is the signal that containes FM channels with a bandwidth of 200 kHz. The sampling frequency of this signal is 21875 kHz. According to the output in this input, signals are placed in the channels as desired but the center frequencies of the signals are not set in the base band. The reason for this is that the center frequencies of the WOLA FFT channelizer have certain values. In addition, the single fm signal whose center frequency is set to the base band and the sinus signal whose center frequency is 683 kHz are given as inputs. According to the output in this input, signals are placed in the channels as desired and the channel signals are placed on the base band. It was also observed that the channels did not leak into each other. The sinus signal whose center frequency is 683 kHz are given as inputs. . According to the output in this input, the input signal does not exist in any channel. The passband bandwith of the channels which are 563 kHz and the stopband bandwith of the channels which are 120 kHz cause this situation.If the center frequency of the input signal is set to filter's stopband frequency, the input signal does not exist on any channel at the output of the channelizer. This indicates that the channelizer is not working properly for all possible input signals. So, WOLA channelizer has blind areas that can not be channelized. The size of these blind areas determines the transfer function of weighting process. The wider the transmission band of the weighting function, the narrower the blind region becomes. As described in previous studies, this thesis proved that WOLA FFT channelizers use lower resources than other channelizer types. However, the disadvantage of this channeler is that the WOLA FFT channelizer does not output the input signals between the two adjacent channels. As a solution to this problem, it is suggested that the FFT operation used in the channelizer be run twice as fast to twice the channel bandwidths. In this case, the signals between the two adjacent channels are also present in the two adjacent channels. In this case, alising can occur between the adjacent channels. So channel signal can be corrupt. This is one of the most important disadvantages of the WOLA channelizers. It is understood that the channel bandwidths of the WOLA FFT channelizer are the same each others and the channel bandwidths depend on the sampling frequency of the signal and the the number of channel. This indicates that this channelizer is not working properly for input signals that have different channels which are not same bandwidths. In addition, the ability to adjust the channels center frequency at the desired frequencies as opposed to the DDC-based channelizer adversely affects the flexibility of the channelizer. In order to adjust the channel center frequencies, a mixing module is added to the FFT output, which allows the channel signal to be set on the baseband. Generally speaking, while WOLA FFT channelizer provides advantages over other channeler types in terms of resource utilization, it has disadvantageous in terms of flexibility. WOLA FFT channelers can be used in multichannel systems which have a limited amount of DSP source and the channels that have same bandwidths.