This paper presents the results of test observations toward a point source, 4C39.25, for observation modes with various bandwidths and numbers of IF streams in order to examine a reliability of the Daejeon hardware correlator performance for correlating VLBI (Very Long Baseline Interferometry) data obtained with the several observation modes of the KVN (Korean VLBI Network). We used a DiFX software correlator (DiFX) as a reference, for investigating the output visibilities from the Daejeon corelator. It is found that the band shapes of the output visibilities from two correlators are similar to each other and the correlated flux density for each baseline obtained from the Daejeon hardware correlator is lower by 3 - 7% than that from the DiFX. The flux difference is attributed to the limitation of FPGA resources and the difference of fringe rotation algorithm of the Daejeon hardware correlator. The conversion factor, 0.93 ~ 0.97, is proposed for future correlation with the Daejeon hardware correlator.

This paper describes the development of algorithm for direct data transmission between Raw VLBI Data Buffer (RVDB) and Huge Capacity Data Server (HCDS) operated in Korea-Japan Correlation Center (KJCC). The transmitted data is the VLBI observation data, which is recorded at each radio telescope site, and the data transmitting rate is varying from 1 Gbps, in usual case, upto 8 Gbps. The developed algorithm for data transmission enables the direct data transmission between RVDB and HCDS through 10 Gbps optical network using VLBI Data Interchange Format (VDIF). Proposed method adopts the conventional UDP/IP protocol, but in order to prevent the loss of data during data transmission, the packet error monitoring and data re-transmission functions are newly designed. The VDIF specification and VDIFCP (VDIF Control Protocol) are used for the direct data transmission between RVDB and HCDS. To validate the developed algorithm for data transmission, we conducted the data transmission from RVDB to HCDS, and compared to the transmitted data with the original data bit by bit. We confirmed that the transmitted data is identical to the original data without any loss and it has been recovered well even if there were some packet losses.

In this paper, we propose a new re-quantization method after FFT processing to prevent the distortion of correlation result of VCS (VLBI Correlation Subsystem). The re-quantization is used to rearrange the data bit so as to reduce the data rate processed as 16-bit of FFT result of VCS. Having done this procedure, we found that the distorted spectrum of correlation result occurred in the delay tracking experiments by the re-quantization method introduced for initial design of VCS. In order to solve this, two kinds of re-quantization method, that is, the comparison and selection-type, are proposed. The first is to re-quantize the FFT result as a valid-bit by comparing with the input data after determining the adequate threshold. The second is manually to select the valid-bit of FFT result after finding the valid-field of data according to the bit-distribution of input data. We confirmed that the second is more effective compared with the first through the experimental result, and it will be implemented without so much modification of applied method in the condition of the limited resource of FPGA. The re-quantization is, however, carried out with 4-bit in the proposed second method for FFT result, and then the distortion of correlation result is also appeared. To fix this problem, the bit for re-quantization is extended to 8-bit. The proposed 8-bit selection-type is effectively verified so that the distortion of correlation result disappeared by applying to VCS in consequence of the simulation and correlation experiments.

In this paper, we describe the installation of VLBI Correlation Subsystem (VCS) main product and its performance at the Korea-Japan Correlation Center (KJCC). The VCS main product was installed at KJCC in August 2009. For the overall performance evaluation of VCS, playbacks, Raw VLBI Data Buffer (RVDB) system, and Data Archive (DA) system were installed together. The VCS main product was connected between RVDB and DA, and the correlation results were put into the DA to confirm the normal operation of VCS 16 station mode configuration. The evaluation test was first performed with 4 station mode, same as the factory test of VCS main product. Based on the results of 4 station mode, the same evaluation test was conducted for 16 station mode of VCS. We found that the correlation results of VCS were almost similarly compared to those of the Mitaka FX Correlator. Through the test results, we confirmed that the problems such as spectrum errors, delay parameter processing module and field programmable gate array errors in antenna unit, which were generated at the factory test of VCS main product, were clearly solved. And we verified the performance and connectivity of VCS by obtaining the expected correlation results and we also confirmed that the performance of VCS was sufficient for real VLBI observation data in both 4 and 16 station modes.

We present the performance test results of VLBI Correlation Subsystem (VCS) trial product which was being developed for 1 year from August 2007. It is a core component of Korea-Japan Joint VLBI Correlator (KJJVC). The aim for developing VCS trial product is to improve the performance of VCS main product to reduce the efforts and cost, and to solve the design problems by performing the preliminary test of the manufactured trial product. The function of VCS trial product is that it is able to process the 2 stations-1 baseline, 8 Gbps/station speed, 1.2 Gbps output speed with FX-type. VCS trial product consists of Read Data Control Board (RDC), Fourier Transform Board (FTB), and Correlation and Accumulation Board (CAB). Almost main functions are integrated in the FTB and CAB board. In order to confirm the performance of VCS trial product functions, the spectral analysis, delay compensation and correlation processing experiments were carried out by using simulation and real observation data. We found that the overflow problem of re-quantization after FFT processing was occurred in the delay compensation experiment. We confirmed that this problem was caused by valid bit-expression of the re-quantized data. To solve this problem, the novel method will be applied to VCS main product. The effectiveness of VCS trial product has been verified through the preliminary experimental results, but the overflow problem was occurred.