///////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////// Francesco Raffaelli 13th June 2018; The data contained in this folder was collected at the University of Bristol, School of Physics and presented in the article: "Generation of random numbers by measuring phase-fluctuations from a laser diode with a silicon-on-insulator chip" Francesco Raffaelli, Philip Sibson, Jake E. Kennard, Dylan H. Mahler, Mark G. Thompson and Jonathan C. F. Matthews. This readme file is written with the expectation the reader has read carefully and in detail the manuscript, to understand context and terminology. //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////// AUTOCORRELATION (Figure 4 in the paper): These are the data used to obtain the autocorrelations in the case of raw and hashed bit strings. Raw data: -raw_bits_500Msamples.txt -raw_bits_1250Msamples.txt -raw_bits_5000Msamples.txt Hashed Data: -hashed_bits_500Msamples.txt; -hashed_bits_1250Msamples.txt; -hashed_bits_5000Msamples.txt; These data have been converted into integer included between [0,256] and then normalised to 1. Then, the matlab function autocorr has been used to calculate the autocorrelations. STATISTICAL TESTS (Table II in the paper): Description of the files: - hashed_nist.txt is the file of hashed bits tested and reported in the paper. The presented P-values and proportions of passed tests were obtained with dividing the file into 1000 blocks composed of 1100000 bits each. - finalAnalysisReport_nist.txt has the report given in output from the NIST test suite. It can be used to read the P-values. BANDWIDTH (Figure 3 in the paper) The data for the spectra of the homodyne detector are reported in the following files: -spectrum_signal.csv Column 1 is the frequency in Hz units Column 2 is the spectral density in dB units -spectrum_noise.csv Column 1 is the frequency in Hz units Column 2 is the spectral density in dB units These data were taken with the same setup for laser and chip used for the generation of random numbers. EXPERIMENTAL CHARACTERISATION of the QCNR (Figure 2 in the paper): - quadratic_plot.txt Column 1: optical power in W units Column 2: Variance of the voltage signal in V^2 units. Description of the measurements acquisition: the chip was first configured in order to optimise the visibility of phase shifters 1 and 3 (Fig. 1). Then, for each power from 10 mA to 30 mA input current, the phase of phase-shifter 2 was varied. Each point in the file "quadratic_plot.txt" is obtained as the maximum of these fringes. - quadratic_plot_VOA.txt Column 1: optical power in W units Column 2: Variance of the voltage signal in V^2 units. Description of the measurements acquisition: The phases were chosen similarly as the previous measurements. In this case the input current of the laser was set at its maximum and the Variable Optical Attenuator was used to set the same power as in the measurement without the VOA. The QCNR is obtained from these two set of data as described in the paper. STABILITY OF THE DEVICE (Figure 6 in the paper) "std_dev_monitor.txt" contains the standard deviation of the voltage signal when controlling the phase shifter 2. Column 1 is the time between two consecutive measurements in second units Column 2 is the optical power in W units Column 3 is the variance in V^2 units The data with the controlled optical coupling were obtained by normalising the varaince values taking into account the decoupling happening during the measurements. "std_dev_monitor_free_run.txt" is the standard deviation of the voltage signal without any control. Column 1 is the time between two consecutive measurements in second units Column 2 is the standard deviation in V units*. *The square must be taken to compare these data with the data present in the file "std_dev_monitor.txt" and to obtain the plot in Fig. 6.