Shinn-Yan Lin

Sagnac Effect and Diurnal Correction on Two-Way Satellite Time Transfer

Precision of two-way satellite time and frequency transfer (TWSTFT) highly depends on the residual nonreciprocity delays; one of them is caused by the Sagnac effect. In this paper, we employ the satellite ephemeris data to compute the precise Sagnac effect and discuss the possible additive effect related to the Sagnac effect. We also describe a method to calculate the diurnal correction based on the time derivative of the Sagnac effect. Through the analysis of the link between the National Institute of Information and Communications Technology in Japan and the Telecommunication Laboratories in Taiwan, we demonstrate that the correction was useful to improve the TWSTFT data. The time deviation at an 8-h averaging time is reduced from 346 to 156 ps after the correction.

First international two-way satellite time and frequency transfer experiment employing dual pseudo-random noise codes

Two-way satellite time and frequency transfer (TWSTFT) is one of the main techniques used to compare atomic time scales over long distances. To both improve the precision of TWSTFT and decrease the satellite link fee, a new software-defined modem with dual pseudo-random noise (DPN) codes has been developed. In this paper, we demonstrate the first international DPN-based TWSTFT experiment over a period of 6 months. The results of DPN exhibit excellent performance, which is competitive with the Global Positioning System (GPS) precise point positioning (PPP) technique in the short-term and consistent with the conventional TWSTFT in the long-term. Time deviations of less than 75 ps are achieved for averaging times from 1 s to 1 d. Moreover, the DPN data has less diurnal variation than that of the conventional TWSTFT. Because the DPN-based system has advantages of higher precision and lower bandwidth cost, it is one of the most promising methods to improve international time-transfer links.

Remote frequency control via IEEE 1588

A new scheme of remote frequency control based on the IEEE 1588 standards, a precision time synchronization protocol (PTP) is proposed in this paper. A remote oven-controlled crystal oscillator (OCXO) is steered by integrating the PTP system, fuzzy controller and D/A converter, such that its frequency can follow a primary cesium atomic clock used as the master clock of the PTP system. Experimental results show that for averaging times of one day, the frequency stability of the OCXO can be improved from a few parts in 109 to 1012.

Improving TWSTFT short-term stability by network time transfer

Two Way Satellite Time and Frequency Transfer (TWSTFT) is one of the major techniques to compare the atomic time scales between timing laboratories. As more and more TWSTFT measurements are performed, large numbers of point-to-point two-way time transfer links have grown to be a complex network. For a future improvement of the TWSTFT performance, it is getting important to reduce measurement noise of the TWSTFT results. One of the methods is using the TWSTFT network time transfer. The Asia-Pacific network is an exceptional case of simultaneous TWSTFT measurements. Some indirect links through relay stations show better short-term stability than the direct link because the measurement noise may be neutralized in a simultaneous measurement network. In this paper, the authors propose a practicable method to improve the short-term stability by combing the direct and indirect links in the network. Through the comparisons of time deviation (TDEV), the results of network time transfer show clear improved short-term stabilities. For the links based on H-maser, the average gain of TDEV at 1h averaging times is 20%. As TWSTFT short-term stability can be improved by network time transfer, the network may allow larger number of simultaneously transmitting stations.

The comparison between TWSTFT and GPS time transfer result of PTB-TL LINK

In this paper we compare the PTB-TL time transfer link data of TWSTFT and GPS carrier phase solutions. We investigated in detail the double difference between the time transfer data of TWSTFT and GPS IGSC/TAIPPP. Our results show the PTB-TL link has a conspicuously diurnal effect. The periodicity is almost exact 1 day and the peak-to-peak amplitude is about 1∼1.5 ns. We also found a phase drift which is about 5 ns between TWSTFT and GPS data during a period of 130 days.

Analysis of the Asia-Pacific TWSTFT network

Two way satellite time and frequency transfer (TWSTFT) is an important method to compare the atomic time scales between time keeping laboratories. The aim of this paper is to analyze the Asia-Pacific TWSTFT network, in which the measurements are performed synchronously by using multi-channel modems. As a result of the synchronous measurements, the clock drift and the majority path instabilities could be neutralized in the calculation of the closure links. This paper discusses what kinds of noise could be cancelled and what kinds of noise would still be kept in the result of the closure links by taking all three stations at the same time into account. From the re-examination of TWSTFT equations, we deduced that the time difference between two laboratories could be obtained via relay stations without debasing the stability in the synchronous TWSTFT measurements. The results show that the noises of the closure links in the Asia-Pacific TWSTFT network are very low. The smallest standard deviation of triangle closure is only 121 ps for the data over four months. The closing errors are closed to zero after removing calibration values. The analysis of modified Allan deviation confirms that the stability of data via relay station is in competition with the direct TWSTFT link. This paper also shows that the stability of TWSTFT could be improved by the weighted mean of the redundant TWSTFT links.

Design of GPS Remote Calibration System

A new design of GPS remote time and frequency calibration system is presented. Its hardware was composed with a time interval counter card, a GPS receiving modular and a controller. The counter card was manufactured with modeled number GT200 by Guide Technology, Inc. The Trimble ThunderBolt GPS disciplined clock was used as the GPS receiving modular. A personal computer with Redhat Linux operation system was utilized for the controller. The software is written by C language which controlled the GPS modular, the counter card, and data collecting. The characteristic of the counter card and GPS receiving modular were performed. A series of test were executed to valuate this system. The short baseline common clock test with hydrogen maser was studied for the system noise level. The direct phase comparison in laboratory, GPS short baseline test in laboratory, and GPS field test within 10 km between hydrogen maser and cesium clock were also experimented two methods were applied to evaluate the system performance and there were GPS all-in-view and phase difference comparison. The later one could illustrate the short-term frequency stability. The former method could demonstrate the longer term stability and time characteristic of the clock at remote site. The accuracy and stability of the experimental result of the short baseline common clock with hydrogen maser reached to a few parts of 10-13 and the TDEV values were lower than the ITU-T G.811 requirements. The designed system with the capability could be applied for the remote time/frequency calibration and for the monitoring the performance of the primary reference clocks (PRC) of the digital telecommunication network, too.

Progress in the link calibration for UTC time transfer

The dominant part of the total uncertainty budget in the [UTC-UTC(k)] is from the type B uncertainty (uB) in time link calibration. In fact, 60% of the type A uncertainty (ua) in UTC time links are <; 1 ns while the 93% of uB ≥ 5ns. In 2011, the BIPM launched a pilot experiment to reduce the uB from 5 ns at present to ≤ 2 ns and has developed a standard calibration scheme, namely METODE, associated with an integrated portable GNSS calibrator together with the reference receivers stationary at BIPM. Similar setups are made in TL. Using these systems, we performed a series experiments at BIPM, OP and TL to study the attainable uncertainty of METODE. The results are compared to that of the traditional differential receiver calibrations.

Multidevice Time Measurement System via a PTPD Network

The precision time synchronization protocol daemon (PTPd) is a software package that realizes a PTP. In this paper, a traceable PTPd measurement system is examined. The key component of the proposed time measurement system is the PTPd. It runs in a Linux kernel-based network. The PTPd operation mode is free running. A stable clock (i.e., an embedded system oscillator in our experiments) plays the role of a master clock. Measured clocks (i.e., desktop system time in our experiments) play the roles of slave clocks. From our results, we find that the system clock of an embedded system and desktop computers all accord with the demand of the PTPd, and the PTPd can be used as a software time measurement tool in the submicrosecond range.

Long-term inconsistency between TWSTFT and GPS time transfer results in PTB-TL and NICT-TL time links

In this paper, we compared the TWSTFT and GPSPPP data for the PTB-TL, PTB-NICT, and NICT-TL time transfer links. We investigated the double difference between the time transfer data of TWSTFT and GPSPPP from 2005 to 2010. The results show two time transfer techniques had a conspicuously inconsistency and the patterns of double difference results were similar in PTB-TL and PTB-NICT links. The double difference results of all direct links exhibited the annual varying patterns, whose peak-to-peak phase difference in 1-day moving average are 4.5 ns both for the PTB-TL and PTB-NICT links, and 2 ns for NICT-TL link.