CC BY
71
NEW GENERATION MOBILE NETWORKS SYNCHRONIZATION
Griazev A.N.,
Director General Central Science Research Telecommunication Institute, Russia, Moscow,
Melnik S.V.,
Ph. D., head of department, Central Science Research Telecommunication Institute, Russia, Moscow, [email protected]
Petrov D.A.,
Ph.D., analytic, Univercity of Jyvaskyla, Finland, Jyvaskyla, [email protected]
Smirnov N.I.,
Doctor of science, professor, Moscow Technical University of Communications and Informatics, Russia, Moscow, [email protected]
Keywords: 4G, 5G, GPS, GLONASS, mobile networks, service, synchronization, time server
In the article solved time and frequency synchronization problem for new generation (NG) mobile networks. Distributed network with PTP (PrecisionTimeProtocol) was created For this solution. We present distributed algorithm for time synchronization at the NG mobile networks. This algorithm has more effective results for application timing, then present algorithm with one Grand master device and time network distribution with border clock units. So we needn't change existing network infrastructure equipment and just it is only need to add a few cheap, but high stable timing sources with GPS/GLONASS synchronization on board.
Described the method for time and frequency synchronization of new generation mobile networks by the PTP usage. PTP did not guarantee high accuracy timing itself. If we use centralized timing mechanism, we put Grand Master at the global mobile switching centre and send time labels through transit routers with Boundary Clock inside for retiming. This is similar then in previous time core network synchronization, but for new generation networks there is new more effective timing method. There are two other cases for time and frequency synchronization of new generation mobile networks. One of them is the GPS/GLONASS usage receivers on each base station, controller and mobile switching centre. And the second one is the providing time and frequency synchronization by the time servers installed at the mobile radio network controllers (RNC). This last case has the most effective characters, because provide enough quality with minimum extra costs and same time has more reliability then centralized timing mechanism.
For citation:
Griazev A.N., Melnik S.V., Petrov D.A., Smirnov N.I. New generation mobile networks synchronization // T-Comm. 2015. No.2. Pp. 94-96.
1. time and frequency synchronization
By GPS
Today the most popular method for time and frequency synchronization is the GPS receiver's usage on each base station. But for QoS guarantee we need not the simple GPS receiver, but the measurement one. Measurement GPS receiver is too expensive tool.
At the World Mobile Congress in Barcelona 2012 there was the round table discussion about perspective timing in the new generation mobile networks. Most of the mobile LTE operators representative persons confirmed, that measurement GPS receivers usage an each base station is too expansive. Also it is very expansive to make reengineering of the existing core networks for Boundary Clock installing for centralized timing mechanism. It was consolidate opinion of the European and Asian companies.
Now time Central Science Research Telecommunication Institute provides the development of the Russian standard for time and frequency synchronization in telecommunication networks. The fundamental basis of this job is usage GLONASS as the main timing source for service timing.
In a network with non 1588 aware switches, PDV and Asymmetry can be significant. Slave clock recovery is a challenge. Tests have shown Frequency recovery is easier but Phase/Time recovery is a challenge.
Boundary Clocks reduce PDV accumulation by:
• Terminating the PTP flow and recovering the reference timing,
■ Generating a new PTP flow using the local time reference, (locked to the recovered time).
• There is no direct transfer of PDV from input to output.
A Boundary Clock is in effect a back-to-back Slave+Master. IEEE 1588 standardized back-to-back architecture Slave+Master with Boundary Clock at the transit routers:
• Boundary Clocks (BCs) are back-to-back Slave-Masters.
• BCs recover and re-generate the 1588v2 clocking.
• With a network of BCs, PDV contribution (per hop) is only from BC and link.
■ PDV experienced by Slave is minimised.
ii. Timing mechanism with one central grand master device
Centralized timing mechanism based on IEEE 1588 Master - Slave hierarchy. In the core network with packet switching traffic we have Propagation Delay Various (PDV) effect. Each transit router adds contribution to the propagation delay and this delay is random. It is depends on the traffic intensive, microcontroller mechanism in the routers port, temperature, quality of connectors and other parameters. PDV accumulation shown at the Figure I.
PDV Accumulation
HIB« I »tun m _____ —M* . . —Ml
h ^ »^SÈ B^fife Jt^si
EÏÏ' 1 W
Test results of the BC's Output Noise, Noise Transfer and Noise Tolerance shown at the Figure 2.
PDV final reduce making by Transparent Clocks (TC). TC device calculating the time a PTP packet (in nsec) and insert the value into the CorrectionField. Using the CorrectionField, the Slave or terminating BC can effectively remove the PDV introduced by the TC. This architecture needs to reengineering all core networks in Operator Company. Most of 4G LTE Operators Company understood that it needs really big investment.
In Russia it means, that all synchronization and network transit equipment should be local produced. For Russian operators companies it can cost very big money.
Fortunately there is another way to made service timing, which will be described after.
III. General requirements for time
and frequency synchronization at the mobile networks
Fig. I. Propagation Delay Various accumulation effect
Fig. 2. Test results of the BC's Output Noise, Noise Transfer and Noise Tolerance IEEE 1588
New generation telecommunication networks must have high precision timing for provide services in Mobile Backhaul, High Frequency Trading and support Power Smart grid.
Key timing levels shown in the Table I.
A new radio canal signal modulation based on basis pseudo orthogonal not only in time line, but also pseudo orthogonal in frequency line will be used in the new mobile networks 5G [6]. For mobile backhauls will be used united frequency bands similar as at the cognitive radio technology. For signal processing will be need high precision synchronization about 100 ns.
T-Comm #2-2015
Table I
Key timing levels
Standard Frequency Time (Physical level) Time (application level)
CDMA2000 SO 10' (± 3 - 10) mks 100 ms
TD-SCDMA 50 10' ± 1,5 mks 100 ms
GSM 50 10' Not difinded 100 ms
WCDMA 50 10' Not defended 100 mks
WCDMA/HSPA+ 50 10' Not defended 10 mks
LTE (FDD) 50 10' Not defended 10 mks
LTE (TDD) 50 10' ± 1,5 mks small cell (<3km), ±5 big cell(>3km) 10 mks
LTE A MBSFN 50 10' (± 1 - 32) mks 10 mks
LTE-A CoMP (Network MIMO) 50 10* +/- 500 ns 10 mks
LTE-A ICIC (HetNet Coordination) 50 10* ± 5 mks 10 mks
WiMAX (TDD) 50 10* (±1-8) mks 10 mks
IV. Perspective service synchronization
for mobile networks based on time servers with GPS/GLONASS usage
a get
Fig. 3. Test results of Output Noise, Noise Transfer and Noise Tolerance the distributed timing solutions
If we use GPS/GLONASS receivers for timing, we can provide high stable internal clock inside each device. It is very expensive. I case usage centralized timing with PTP Master-Slave hierarchy we need to make deep reengineering at the core network to include BC and TC devices. It is also takes huge investments. But there is another way. I case we use high stable clock (time server) as Master device and install it in the R.NC device. It is also very expensive if we'll use full
functional Grand Master device, but if we can use small not expensive time server with high stable internal clock and GPS/GLONASS receiver inside - it is the best solution. In STC KOMSET we design such time server. So we have distributed timing architecture, which is not expensive and more надежная with the same time precision characters. It is approved and certified in Russia as measurement device and it already used in the new LTE BSS solutions. Test results of Output Noise, Noise Transfer and Noise Tolerance the distributed timing solutions shown at the Figure 3,
This distributed synchronization architecture will help operators companies build new generation mobile networks SG ready with minimum additional expenses and maximum reliability. In future when new GLONASS will be available, it will be possible use GLONASS only timing. Beforetime distributed synchronization architecture is the most effective solution for service timing at the new generation mobile networks.
For implement distributed synchronization architecture needs to be approved of the existing national telecommunication standards. Central Science Research Telecommunication Institute is working about it and it will be ready very soon.
References
1. Melnik S.V., Petrova E.N., Smirnov N.I. GLONASS usage for more effective mobile networking. Reports of seminar «Synchronization systems for signal processing in telecommunication networks ». Nizhny Novgorod, June 2010. - M.: Insvyazizdat.
2. Melnik S.V., Petrova E.N. GLONASS usage for time and frequency synchronization in Telecommunication networks. Reports of « IQ Berlin», 2012. - Berlin, Germany.
3. Melnik S.V., Petrova £.N.. Smirnov N.I. GLONASS usage for LTE networks synchronization. Reports of seminar «Synchronization systems for signal processing in telecommunication networks ». Izhevsk, 2012. - M.: Insvyazizdat.
4. Melnik S.V., Petrov D.A. Overview of Single Frequency Multipoint Transmission Concepts for HSDPA and Performance Evaluation of Intra-site Multiflow Aggregation Scheme. Proceedings of 75th IEEE Vehicular Technology Conference (VTC). Yokogama 2012, Japan.
5. Melnik S.V., Petrov D.A. Computationally Efficient Modulation of Well-Localised Signals for OF DM. Proceedings of the International Conference on Wireless and Mobile Computing, Networking and Communications (WiMob 2012). Barcelona 2012. Spain.
6. Melnik S.V., Sm/rnov N.I. CAPEX reduce for mobile 4G network installation based on synchronization nodes usage. 2013. International Informatisation Forum (IIF) 2013. International congress «Communication networks and technologies»,Moscow, 201 3 - M.: Insvyazizdat.
7. Melnik S.V., Petrov D.A., Petrova E.N., Smirnov N.i. Service timing for 4G and 5G mobile networks. IEEE «Problems of Infocommunications. Science and Technology» (PICS&T-2014) . 14-17 October, 2014.
