Real-time measurement of noise statistics

There has been much interest in the statistics of electromagnetic noise such as amplitude probability distribution (APD), average crossing rate (ACR), pulse duration distribution (PDD), and pulse spacing distribution (PSD). Traditionally, noise statistics have been measured using a digitizer and personal computer. In such measurements, the total number of samples for calculating these statistics is limited by the amount of RAM in the digitizer. To remove this limitation, we have developed new statistics measuring equipment that digitizes the detected noise at a sampling rate of 20 MHz and classifies the sampled data into 256 levels with an accuracy of 30 bits. The frequency distribution of the classified data is calculated in histogram measuring circuits. Therefore, the frequency distribution is measured in real time based on a total of up to 10/sup 9/ samples. This paper describes the statistical measurement method and reports the statistics of noise leaked by a microwave oven in the 2.4-GHz band measured by the developed equipment. It was confirmed that the noise statistics with an occurrence probability of less than 10/sup -9/ can be measured.