On the need to use the median signal filtering method to improve the metrological characteristics of the rubidium frequency standard during processing and transmitting large data arrays Anton Valov Vasiliy Rud Bonch-Bruevich Saint Petersburg State University of All-Russian Research Institute of Phytopathology Telecommunications Moscow Region, Russia Saint-Petersburg, Russia rudvas.spb@gmail.com tony.valov2015@yandex.ru Nikita Lukashev Vadim Davydov Bonch-Bruevich Saint Petersburg State University of Peter the Great Saint Petersburg Polytechnic University Telecommunications Saint-Petersburg, Russia Saint-Petersburg, Russia All-Russian Research Institute of Phytopatholog n-lukash@list.ru Moscow, Russia davydov_vadim66@mail. ru Abstract—The article describes the construction of the navigation signals emitted by different satellites during a rubidium – 87 quantum frequency standard for satellite temporary mismatch of 10 ns causes an additional error in infocommunication and navigation systems for various determining the location of an object of 10-15 meters. purposes. The necessity of improving the metrological The expanding of the tasks number for which the characteristics of the quantum frequency standard for long- satellite navigation systems are used, the accuracy is term transmission of large amounts of data is substantiated. A new algorithm is proposed for processing data of a large required of increase of determining the object position to 0.5 amount of data about the error signal using the median m. On the other hand, a development of scientific and filtering method to improve the short-term and long-term technological progress is changing the composition of the stability of the frequency standard. The results of experimental used electronic equipment. All this requires constant studies of the metrological characteristics of the quantum modernization of satellite navigation systems, including frequency standard are presented. The improvement in long- quantum frequency standards. term frequency stability was found to be 7%, which reduces The development and commissioning of new QFS the number of bit errors during long-term transmission of models is a very lengthy and expensive process. In most information by at least 3%. cases, there is no time and sufficient funds for its Keywords—quantum frequency standard, signals processing, implementation. Therefore, in most cases, for the specific information transfer, data arrays, Allan's deviations, median problems it is better to perform the modernization of the filtering method rubidium – 87 QFS and caesium - 133 QFS, which are in operation on satellite systems [18-20, 23-25, 32-35]. I. INTRODUCTION The process of modernizing frequency standards In the modern world, accurate measurement of time and includes various directions: changing weight and frequency is necessary to solve various problems in the dimensions, reducing energy consumption, improving fields of science and technology [1-7]. One of the most metrological characteristics. The quantum frequency complex of them is ensuring the uninterrupted operation of standards are characterized by the fact that modernization satellite communication systems, navigation, Earth surface may not be for its entire structure, but only for individual research and spacecraft control from orbit [1, 3, 8-17]. A nodes or blocks. In our work, we are considering one of special place among the devices for determining the such directions for improving the metrological characteristic frequency and time is occupied by quantum frequency of the rubidium-87 quantum frequency standard. standards (QFS) [8, 9, 18-22]. Currently, global satellite constellations for various purposes use only QFS [9.18-20, II. THE CONTROL ALGORITHM FOR FREQUENCY ADJUSTING OF THE QUANTUM STANDARD 23-25]. The main advantage of QFS over other devices is the use for stable operation of stabilization systems for the During the period of operation of rubidium – 87 QFS, frequency of laser radiation and optical elements [3, 8, 19, the structural schemes of its various QFS models did not 20, 25-29]. Other devices cannot provide the necessary undergo fundamental changes in comparison with their accuracy in determining the frequency and long-term classical representation [8, 9, 18-20, 30, 31]. In the design of stability of its nominal value in conditions of large the QFS, individual elements or blocks, as well as control overloads. A slight deviation of the frequency from the systems for various parameters, are mainly changed to nominal value leads to large errors, especially when improve the metrological characteristics of the standard. On transmitting large data streams. fig. 1 is shown the structural diagram of the rubidium – 87 One of the main problems of a satellite system is the QFS. mutual synchronization of satellite timelines to nanoseconds In this design the method developed by us for improving or less [9, 18-20, 30-32]. For example, the error of the the microwave signal parameters to improve the metrological characteristics of the standard is implemented. Copyright © 2020 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0) Data Science where 𝑐𝑜𝑠(𝜔1 − 𝜔2 )𝑡 is the difference and 𝑐𝑜𝑠(𝜔1 + 𝜔2 )𝑡 is the total frequency. All side components, if they appear in the spectrum of signals with frequencies ω1 and ω2 (for example, due to temperature drift in a quartz filter), are converted to signals with combination frequencies. These signals will create additional errors. Therefore, during each communication session of the satellite with the ground station, when comparing the time scales in the QFS on the rubidium Fig. 1. The structural diagram of the rubidium QFS: 1 - reflector; 2 - a atoms, 87 adjust the frequency. If for some reason the lamp with vapors of rubidium - 87; 3 - cell - filter; 4 - solenoid C - fields; 5 - radio frequency resonator; 6 - a cell with buffer gas; 8 - photodetector; 9 - communication session did not take place, then the satellite automatic frequency control system; 10 - highly stable crystal oscillator; 11 can be suspended in the navigation system. - frequency synthesizer. Therefore, it is extremely important to develop a method The operation of the rubidium – 87 QFS is based on the that, on the one hand, provides high accuracy of the output principle of tuning a highly stable voltage-controlled quartz frequency with its tuning in autonomous mode, regardless of oscillator 10 to the quantum-frequency transition of communication with the ground station. In the methods for rubidium–87 atoms [8, 18, 23, 24, 35]. To implement the QFS considered in [8, 18, 30, 31], the frequency tuning step noted frequency adjustment of the quartz oscillator, a is more than 1 Hz using a voltage setting of the voltage- microwave signal from a frequency synthesizer (FS) 11 is controlled crystal oscillator 10. supplied to a vacuum cell 6 filled with rubidium-87 atoms The frequency tuning of the crystal oscillator is and a buffer gas 11. When the frequency of the microwave controlled by the AFC system 9, which includes a crystal signal coincides with the quantum transition frequency of oscillator control unit (XOCU) and a control unit that the excited rubidium – 87 atoms, the signal detected by the converts the signal from the quantum discriminator and photodetector has maximum signal to noise ratio (S/N). If calculates the tuning code for further sending it to the the frequency of the microwave signal fmw leaves the value XOCU. of the frequency of the resonant transition, the S/N ratio In the previous version of the software for the QFS decreases and an error signal is generated by the electronic control device, simple accumulation of all received data at circuit 9. This signal is used to adjust the frequency of the different frequencies was used with further calculation of crystal oscillator 10. Therefore, one of the key points in the the value of the error signal of the microwave signal. In the operation of the QFS on rubidium atoms is 87, is the new version, it is proposed to use the median filtering formation of a microwave signal taking into account its method as one of the data filtering methods. various features. The process of generating a microwave The median filtering method uses the ordering of several signal is carried out in the midrange 11. It is necessary that elements received at the input of the control unit and the the output of the midrange provide high accuracy of the subsequent selection of a value equally spaced from the output frequency, high suppression of the side amplitude beginning and end of the ordered series of elements. components in the spectrum of the output signal, low III. THE RESULTS OF EXPEREMENTAL INVESTIGATIONS AND dependence of the frequency and amplitude of the output DICUSSION signal on temperature. During of the using of the optical light signals for The method of generating a microwave signal in a recording of resonance conditions on a photodetector, the frequency synthesizer discussed in detail in [1, 2, 20] has important characteristic is the spectral density Sφ [8]. The one significant drawback. The spectrum of the output signal value of Sφ has a significant effect on the S/N ratio. In fig. 2 with a frequency of 5.3125 MHz, which is obtained at the shows the spectral densities of phase noises for two designs output of one of the FS balanced mixers (BM) 11 [20], of the QFS (the previously used design and a new). In new contains lateral amplitude components. If one of the side design was used the method of improving the parameters of components coincides with the frequency of any Zeeman the microwave signal. transition, then this will lead to transitions of atoms at these levels and an error in establishing the actual value of the frequency of the output signal of the QFS. To suppress the side components, a quartz filter is used, which has a high temperature dependence. For reliable operation of the quartz filter, high temperature stabilization is required. This is extremely difficult to achieve, especially in conditions of a long flight of the satellite. It should also be noted that the frequency of the microwave signal necessary for the operation of the QFS, which corresponds to the frequency difference between the two ultra-thin sublevels F = 2 and F = 1 [8], is formed at the output of the balanced midrange mixer 11. The BM operation can be described by the equation: 1 1 𝑈𝑜𝑢𝑡 = 𝑐𝑜𝑠(𝜔1 − 𝜔2 )𝑡 + 𝑐𝑜𝑠(𝜔1 + 𝜔2 )𝑡, (1) Fig. 2. The phase noise amplitude of error signal. (а) – corresponds to the 2 2 earlier value of the error signal with simple accumulation, (б) – error signal using median filtration. VI International Conference on "Information Technology and Nanotechnology" (ITNT-2020) 103 Data Science An analysis of the obtained results shows a significant The resulting improvements in short-term and long-term decrease in phase noise amplitude when using the median frequency stability can improve the reliability of satellite filtering method. transmission systems of large amounts of information. The spectral density of noise in the frequency range of the tuning of the resonant frequency of the quantum REFERENCES transition is presented on fig. 3. [1] A.A. Petrov, V.E. Shabanov, D.V. Zalyotov, A.L. Bulyanitsa and D.V. Shapovalov, “Modernization of the frequency synthesizer of cesium atomic clock,” IEEE International Conference on Electrical Engineering and Photonics, EExPolytech, vol. 8564389 pp. 52-55, 2018. [2] R. Davydov, V. Antonov and A. Moroz, “Parameter Control System for a Nuclear Power Plant Based on Fiber-Optic Sensors and Communication Lines,” IEEE International Conference on Electrical Engineering and Photonics (EExPolytech), vol. 8906791, pp. 295- 297, 2019. [3] V.V. Semenov, N.F. Nikiforov and S.V. Ermak, “Calculation of stationary magnetic resonance signal in optically oriented atoms induced by a sequence of radio pulses,” Soviet journal of Communications Technology and Electronics, vol. 36, no. 4, pp. 59- 63, 1991. [4] V.A. Lenets, M.Yu. Tarasenko, N.S. Rodygina and A.V. Moroz, “New method for testing of antenna phased array in X frequency Fig. 3. The phase noise spectral power density of error signal: a - the range,” Journal of Physics: Conference Series, vol. 1038, no. 1, previous version of the software; b - the new version. 012037, 2018. [5] R. Davydov, V. Antonov, S. Makeev, Y. Batov, V. Dudkin and N. An analysis of the obtained results of experimental Myazin, “New high-speed system for controlling the parameters of a investigations in fig. 3 showed that the use of the method nuclear reactor in a nuclear power plant,” E3S Web Conf., vol. 140, developed by us, as well as the use of a microcontroller for 02001, 2019. control in the QFS, reduces the power of phase noise in the [6] N.S. Myazin, V.I. Dudkin, N.M. Grebenikova, R.V. Davydov and A.S. Podstrigaev, “Fiber - optical system for governance and control output spectrum of the signal. of work for nuclear power stations of low power,” Lecture Notes in All this made it possible to improve the short-term Computer Science, vol. 11660, pp. 744-756, 2019. frequency stability - Allan's deviation, as well as long-term [7] R.V. Davydov and V.I. Antonov, “Research and modeling of laser frequency stability. The results of the investigations of ablation by ultra-short laser pulses for metal targets,” Journal of Allan's deviation are presented on fig. 4. Physics: Conference Series, vol. 1368, no. 2, 022007, 2019. [8] F. Riehle, “Frequency standard. Basic and applications,” Wiley-VCH Verlag CmbH Co. KGaA, New-York, 2008. [9] S.V. Sokolov, V.V. Kamenskiy, S.V. Kovalev and E.N. Tishenko, “Using of inter-satellite measurements for high-precision assessment of the navigation parameters of an object,” Measurement Techniques, vol. 60, no. 1, pp. 115-119, 2017. [10] L.I. Lebedev, Yu.V. Yashakov, T.S. Utesheva, V.P. Gromov, A.V. Borusyak and V.E. Turpalov, “Complex analysis and monitoring of the environment based on Earth sensing data,” Computer Optics, vol. 43, no. 2, pp. 282-295, 2019. DOI: 10.18287/2412-6179-2019-43-2- 282-295. [11] A.D. Tarasov and O.B. Milder, “Gradation trajectories as an analog of gradation curves in the metric CIE Lab space: discrete approach,” Computer Optics, vol. 43, no. 1, pp. 132-136, 2019. DOI: 10.18287/ 2412-6179-2019-43-1-132-136. Fig. 4. The Allan deviation: (a) – previously used QFS construction, (b) – [12] N.G. Fedotov, A.A. Syemov and A.V. Moiseev, “Theoretical developed QFS construction. foundations of hypertrace-transform: scanning techniques, mathematical apparatus and experimental verification,” Computer The analysis of the obtained results (fig. 4) shows that Optics, vol. 42, no. 2, pp. 273-282, 2018. DOI: 10.18287/2412-6179- 2018-42-2-273-282. the implemented technical solutions and the developed new [13] N.I. Chervyakov, P.A. Lykianov and A.R. Orazaev, “Tho methods of method for improving the parameters of the microwave adaptive filering of impulse noise in images,” Computer Optics, vol. excitation signal during modernization of the standard 42, no. 4, pp. 667-678, 2018. DOI: 10.18287/2412-6179-2018-42-4- design made it possible to improve Allan deviation by 7 %. 667-678. [14] P.A. Filimonov, M.L. Belov, Yu.V. Fedotov, S.E. Ivanov and V.A. IV. CONCLUSION Gorodnichev, “An algoritm for sementation of aerosol inhomogeneities,” Computer Optics, vol. 42, no. 6, pp. 1062-1067, The obtained results have shown that the use of a new 2018. DOI: 10.18287/2412-6179-2018-42-6-1062-1067. method of filtering data in the automatic frequency [15] D.G. Asatryan, “Image blur estimation using gradient field analysis,” adjustment system of the microwave excitation signal Computer Optics, vol. 41, no. 6, pp. 957-962, 2017. DOI: 10.18287/2412-6179-2017-41-6-957-962. reduces one of the most important disturbing factors [16] D.G. Vorotnikova and D.L. Golovashkin, “Difference solutions of the (spectral noise density) that affects short-term frequency wave equation on GPU with reuse of pairwise sums of the differential stability. template,” Computer Optics, vol. 41, no. 1, pp. 134-138, 2017. DOI: The experimental investigations of the metrological 10.18287/2412-6179-2017-41-1-134-138. characteristics of the rubidium – 87 QFS are showed the [17] M.V. Gashnikov and N.I. Glumov, “Onboard processing of improvement of the long-term frequency stability on 7 %. hyperspectral data in the remote sensing systems based on VI International Conference on "Information Technology and Nanotechnology" (ITNT-2020) 104 Data Science hierarchical compression,” Computer Optics, vol. 40, no. 4, pp. 543- Computer Optics, vol. 43, no. 4, pp. 692-698, 2019. DOI: 10.18287/ 551, 2016. DOI: 10.18287/2412-6179-2016-40-4-543-551. 2412-6179-2019-43-4-692-698. [18] V.N. Baryshev, D.S. Kupalov, A.V. Novoselov, M.S. Alennikov, A.I. [27] E.S. Kozlova and V.V. Kotlyar, “Specified parameters of sellmeyer Boyko, V.G. Pal’chikov and A.Yu. Blinov, “Small-sized quantum model for silica glass,” Computer Optics, vol. 38, no. 1, pp. 51-56, frequency standard on a rubidium gas cell with pulsed optical 2014. pumping and microwave excitation according to the Ramsey [28] R.V. Skidanov, V.A. Blank and A.A. Morozov, “Study of an imaging scheme,” Measurement Techniques, vol. 59, no. 12 , pp. 1218-1224, spectrometer based on a diffraction lens,” Computer Optics, vol. 39, 2016. no. 2, pp. 218-223, 2015. DOI: 10.18287/0134-2452-2015-39-2-218- [19] A.A. Petrov, “New scheme of the microwave signal formation for 223. quantum frequency standard on the atoms of caesium-133,” Journal [29] M.S. Kuzmin, V.V. Davydov and S.A. Rogov, “On the use of a of Physics: Conference Series, vol. 769, no. 1, 012065, 2016. multi-raster input of one-dimensional signals in two-dimensional [20] A.A. Petrov, “Improvement frequency stability of caesium atomic optical correlators,” Computer Optics, vol. 43, no. 3, pp. 391-396, clock for satellite communication system,” Lecture Notes in 2019. DOI: 10.18287/2412-6179-2019-43-3-391-396. Computer Science (including subseries Lecture Notes in Artificial [30] G.P. Pashev, “Optimal algorithm for synchronizing the quantum Intelligence and Lecture Notes in Bioinformatics), vol. 9247, pp. clock timeline,” Measurement Techniques, vol. 59, no. 6 , pp. 1005- 739-744, 2015. 1012, 2016. [21] G. Fokin, A. Kireev and A.H.A. Al-Odliari, “TDOA positioning [31] S.I. Donchenko, A.N. Shipunov, O.V. Denisenko and I.Yu. Blinov, accuracy performance evaluation for arc sensor configuration,” “Current status and development prospects of the fundamental and Proceedings Systems of Signals Generating and Processing in the metrological means of the GLONASS system,” Measurement Field of on Board Communications, vol. 2018, pp. 1-5, 2018. Techniques, vol. 61, no. 1 , pp. 13-19, 2018. [22] A. Koucheryavy, A. Vladyko and R. Kirichek, “State of the art and [32] A.A. Petrov and N.M. Grebenikova, “Some Directions of Quantum research challenges for public flying ubiquitous sensor network,” Frequency Standard Modernization for Telecommunication Systems,” Lecture Notes in Computer Science (including subseries Lecture Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9247, pp. 299-308, 2015. vol. 11118 LNCS, pp. 641-648, 2018. [23] N.A. Lukashev, A.A. Petrov, V. V. Davydov, N.M. Grebenikova and [33] A.A. Petrov and N.S. Myazin, “Rubidium atomic clock with A.P. Valov, “Improving performance of quantum frequency standard improved metrological characteristics for satellite communication with laser pumping,” Proceedings of 18th International conference of system,” Lecture Notes in Computer Science (including subseries Laser Optics ICLO-2018, vol. 271, 8435889, 2018. Lecture Notes in Artificial Intelligence and Lecture Notes in [24] A.P. Valov, V.V. Davydov and V.Yu. Rud, “The method of Bioinformatics), vol. 10531 LNCS, pp. 561-568, 2017. improving the parameters of the microwave excitation signal in the [34] A.A. Petrov, N.M. Grebenikova, N.A. Lukashev, N.V. Ivanova, N.S. rubidium - 87 quantum frequency standard,” Journal of Physics: Rodygina and A.V. Moroz, “Features of magnetic field stabilization Conference Series, vol. 1410, no. 1, 012246, 2019. in caesium atomic clock for satellite navigation system,” Journal of [25] A.A. Petrov, V.A. Vologdin and D.V. Zalyotov, “Dependence of Physics: Conference Series, vol. 1038, no. 1, 012032, 2018. microwave – excitation signal parameters on frequence stability [35] N.A. Lukashev, V.V. Davydov, A.P. Glinushkin, V.Yu. Rud and V.S. caesium atomic clock,” Journal of Physics: Conference Series, vol. Lukyantsev, “Microwave low mass-dimensional frequency standard 643, no. 1, 012087, 2015. on Hg-199 ions,” Journal of Physics: Conference Series, vol. 1410, [26] A.A. Zhirnov and O.B. Kudrjashova, “Peculiarities of data processing no. 1, 012211, 2019. for optical measurements of disperse parameters of bimodal media,” VI International Conference on "Information Technology and Nanotechnology" (ITNT-2020) 105