(19)
(11)EP 2 434 673 B1

(12)EUROPEAN PATENT SPECIFICATION

(45)Mention of the grant of the patent:
21.09.2016 Bulletin 2016/38

(21)Application number: 09844982.0

(22)Date of filing:  13.11.2009
(51)Int. Cl.: 
H04J 3/06  (2006.01)
(86)International application number:
PCT/KR2009/006694
(87)International publication number:
WO 2010/134674 (25.11.2010 Gazette  2010/47)

(54)

NETWORK SYNCHRONIZATION METHOD AND APPARATUS FOR PERFORMING TIME SYNCHRONIZATION BETWEEN NODES

NETZWERKSYNCHRONISATIONSVERFAHREN UND VORRICHTUNG ZUM DURCHFÜHREN VON ZEITSYNCHRONISATION ZWISCHEN KNOTEN

PROCÉDÉ ET APPAREIL DE SYNCHRONISATION DE RÉSEAUX POUR EFFECTUER UNE SYNCHRONISATION TEMPORELLE ENTRE DES NOEUDS


(84)Designated Contracting States:
AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO SE SI SK SM TR

(30)Priority: 18.05.2009 US 179029 P
12.11.2009 KR 20090109009

(43)Date of publication of application:
28.03.2012 Bulletin 2012/13

(73)Proprietor: Samsung Electronics Co., Ltd.
Suwon-si, Gyeonggi-do, 443-742 (KR)

(72)Inventors:
  • GARNER, Geoffrey, M.
    Red Bank, New Jersey 07701 (US)
  • RYU, Hyunsurk
    Suwon-si Gyeonggi-do 443-706 (KR)
  • PARK, Keun Joo
    Seoul 135-942 (KR)
  • LEE, Jun Haeng
    Hwaseong-si Gyeonggi-do 445-734 (KR)

(74)Representative: Appleyard Lees IP LLP 
15 Clare Road
Halifax HX1 2HY
Halifax HX1 2HY (GB)


(56)References cited: : 
EP-A2- 1 130 801
US-A1- 2005 207 387
WO-A1-2005/051034
  
  • GEOFFREY M GARNER: "Improvement of Jitter, Wander, and Time Synchronization Performance in 802.1AS Wired Transport using Propagation Time Averaging 1 ; as-garner-prop-time-averaging-0107", IEEE DRAFT; AS-GARNER-PROP-TIME-AVERAGING-0107, IEEE-SA, PISCATAWAY, NJ USA, vol. 802.1, 22 January 2007 (2007-01-22), pages 1-13, XP068006917, [retrieved on 2007-01-22]
  • JUDAH LEVINE: 'Introduction to time and frequency metrology.' REVIEW OF SCIENTIFIC INSTRUMENTS vol. 70, no. 6, 01 June 1999, MELVILLE, NY., pages 2567 - 2596, XP012037477
  
Note: Within nine months from the publication of the mention of the grant of the European patent, any person may give notice to the European Patent Office of opposition to the European patent granted. Notice of opposition shall be filed in a written reasoned statement. It shall not be deemed to have been filed until the opposition fee has been paid. (Art. 99(1) European Patent Convention).


Description

Technical Field



[0001] One or more embodiments relate to a network synchronization method and apparatus for performing a time synchronization between nodes.

Background Art



[0002] GM Garner, "Improvement of Jitter, Wander and Time Synchronization Performance in 802.1AS Wired Transport using Propagation Time Averaging 1", IEEE Draft, vol. 802.1, 22 January 2007, pages 1-13, XP068006917, IEEE Piscataway, NJ USA, describes a method of propagation time measurement averaging using a difference equation and a corresponding transfer equation for synchronization of a network. The claims have been characterized in view of this document.

[0003] A variety of schemes to transmit timing between nodes of a network using time stamps exist.

[0004] In such schemes, each of nodes constituting the network attempts time synchronization with a single node among the nodes included in the network. The single node is known as a grandmaster. Specifically, in such schemes, timings is traceable to a grandmaster.

[0005] Each node, except for the grandmaster node, contains a single slave port and possibly one or more master ports. The grandmaster node contains one or more master ports, but no slave ports. Master ports transmit messages including time stamps to all nodes connected to the master ports. Here, the time stamps may be the time when the messages are transmitted. If the port at the other end of the link attached to a master port is a slave port, it receives the time stamp and adds the propagation time between the slave port and the master port to the time stamp. The slave port also time stamps the message on receipt, relative to its local clock. Then, the node that contains the slave port, i.e., the slave node, compares the received time stamp with propagation time added with the time stamp of the message on receipt. The slave node then computes the offset, i.e., the difference between the time stamp of the message on receipt and the received time stamp with propagation time added, and thereby perform a time synchronization. In this instance, the offset corresponds to the difference between master node and slave node time. Each node that sends messages through a master port to synchronize a slave node is known as a master node. Each master node, except for the grandmaster, is also a slave node to another master node.

Disclosure of Invention



[0006] According to the present invention there is provided a network synchronization method as set forth in claim 1. Also according to the present invention there is provided a network synchronization apparatus as set forth in claim 13. Other features of the invention will be apparent from the dependent claims, and the description which follows.

[0007] One or more embodiments may provide a network synchronization method and apparatus that may enhance jitter, wander, and a time synchronization performance by gradually increasing a window size for a propagation time measurement when a system starts up and then a time synchronization between nodes is initiated.

[0008] One or more embodiments may also provide a network synchronization method and apparatus that may enhance jitter, wander, and a time synchronization performance by applying an exponential weight to a computation of an average propagation time value when a full window of propagation time measurements is collected.

[0009] According to an aspect of one or more embodiments, there may be provided a synchronization method, including: measuring a propagation time from a slave node to a master node at a k-th time step where k denotes a natural number; computing an average propagation time value at a (k+1)-th time step according to a first computation formula, when k is less than a sliding window size, and computing the average propagation time value at the (k+1)-th time step according to a second computation formula, when k is greater than or equal to the sliding window size; and computing an offset based on a message transmission time at the master node, a message receipt time at the slave node, and the average propagation time value.

[0010] The first computation formula and the second computation formula may compute the average propagation time value at the (k+1)-th time step using an average propagation time value at the k-th time step and the propagation time measured at the k-th step time.

[0011] The first computation formula may be computed according to the following Equation:

where xk denotes the average propagation time value at the k-th time step and dk denotes the propagation time measured at the k-th time step.

[0012] The second computation formula may be computed according to the following Equation:

where xk denotes the average propagation time value at the k-th time step, dk denotes the propagation time measured at the k-th time step, and a denotes a filter coefficient of a first-order digital filter.

[0013] Here, a may be increased as the sliding window size increases.

[0014] Also, a may be computed according to the following Equation:

where M denotes the sliding window size.

[0015] Also, a may be computed according to the following Equation:

where M denotes the sliding window size and P denotes a constant.

[0016] In one aspect a network synchronization apparatus may be provided for performing a time synchronization between nodes, where the apparatus implements the features as discussed herein.

Advantageous Effect



[0017] According to embodiments, when a system starts up and a time synchronization between nodes is initiated, it is possible to enhance jitter, wander, and a time synchronization performance by gradually increasing a window size for a propagation time measurement.

[0018] Also, according to embodiments, it is possible to enhance jitter, wander, and a time synchronization performance by applying an exponential weight to a computation of an average propagation time value when a full window of propagation time measurements is collected.

Brief Description of Drawings



[0019] 

FIG. 1 illustrates an example of computing a propagation time according to a related art;

FIG. 2 illustrates another example of computing a propagation time according to the related art;

FIG. 3 illustrates a flowchart of a method of measuring a propagation time according to an embodiment;

FIG. 4 illustrates a detailed process of computing an offset of FIG. 3;

FIG. 5 illustrates a configuration of a network apparatus according to an embodiment;

FIG. 6 illustrates a method of enhancing an accuracy in a propagation time measurement according to an embodiment; and

FIGS. 7 through 9 illustrate a signaling for a parameter setting according to an embodiment.


Best Mode for Carrying Out the Invention



[0020] Reference will now be made in detail to example embodiments, examples of which are illustrated in the accompanying drawings, wherein like reference numerals refer to the like elements throughout. Example embodiments are described below in order to explain example embodiments by referring to the figures.

[0021] FIG. 1 illustrates an example of computing a propagation time according to a related art.

[0022] Referring to FIG. 1, a master node 101 may transmit a message to a slave node 102. A time stamp T1 may be stored in the message. The time stamp T1 is a time 103 when the master node 101 transmits the message. When receiving the message, the slave node 102 may confirm a receipt time T2 that is a time 104, relative to the slave node's local clock, when the message is received. When the slave node 102 is aware of a propagation time D1 from the master node 101, the slave node 102 may compute an offset using "offset = T2 - T1 - D1". The offset corresponds to a difference between the master node 101 time and the slave node 102 time. Accordingly, the slave node 102 may perform a time synchronization with the master node 101 using the offset.

[0023] However, to compute the offset as described above, the slave node 102 may need to be aware of the propagation time D1 from the master node 101. Specifically, the slave node 102 may need to compute the propagation time D1.

[0024] FIG. 2 illustrates another example of computing a propagation time according to the related art.

[0025] Referring to FIG. 2, a propagation time between clocks attached to two nodes, that is, a slave node 201 and a master node 202 may be measured by the clocks exchanging time stamps. A requester, which is one of the clocks, may store a transmission time T1 203 of a first message in the first message when the requestor transmits the first message to another clock. The other clock corresponding to a responder may store a receipt time T2 204 of the first message.

[0026] After storing the receipt time T2 204, the responder may transmit a second message corresponding to the first message to the requester. The responder may store a transmission time T3 205 of the second message in the second message when transmitting the second message. The requester may confirm a receipt time T4 206 of the second message.

[0027] The requester may measure the propagation time using the transmission time T1 203 of the first message, the receipt time T2 204 of the first message, the transmission time T3 205 of the second message, and the receipt time T4 206 of the second message. The propagation time may be computed according to the following Equation 1:



[0028] Here, it is assumed that the propagation time measurement scheme satisfies the following conditions:
  1. (1) The propagation time is symmetrical. That is, the propagation time is identical in both directions.
  2. (2) The propagation time is stable. That is, over the timescale to obtain the above-described four time stamps, for example, the time interval T4 - T1, the propagation time is constant.
  3. (3) A frequency difference between the slave node 201 and the master node 202 is small.
  4. (4) Time stamp values are accurate.


[0029] However, the first condition that the propagation time is symmetrical and the second condition that the propagation time is stable indicate that a source of a variable delay between the requester and the responder may not exist. Specifically, there may be a relay device such as a network switch or a network bridge between the slave node 201 and the master node 202. The existence of the above relay device may cause inaccuracy in jitter, wander, and a time synchronization.

[0030] When there exists a network switch or a network bridge between the slave node 201 and the master node 202, a propagation time variation and/or a propagation time asymmetry may result. To mitigate the propagation time variation and the propagation time asymmetry, the relay device such as the network switch or the network bridge may store a receipt time and a transmission time of each message in the message. Accordingly, the relay device includes a clock and a time stamping hardware, known collectively as a transparent clock. However, a transparent clock requires time stamping capability, and not all relay devices have this capability. In addition, the measured propagation time may not be symmetric due to time stamp error (i.e., the time stamps may have some inaccuracy, see the fourth condition above.

[0031] FIG. 3 illustrates a flowchart of a method of measuring a propagation time according to an embodiment.

[0032] According to an embodiment, the propagation time measurement method may average propagation time measurement values, and use the computed average value as a value to estimate an actual propagation time Dave,k. Accordingly, an offset may be computed using "offset = T2 - T1 - Dave,k".

[0033] The propagation time measurement method of FIG. 3 may be performed at a network apparatus that operates as a slave. Hereinafter, the network apparatus that operates as the slave is referred to as a slave node. A network apparatus that operates as a master is referred to as a master node.

[0034] In operation 310, the slave node may measure a propagation time from the slave node to the master node. The propagation time from the slave node to the master node may be a sum or an average of a propagation time T2-T1 from the slave node to the master node and a propagation time T4-T3 from the master node to the slave node. The propagation time measurement may be performed k times every predetermined time unit. Here, k denotes a natural number. To measure the propagation time at a k-th time unit may be used for the same meaning as to measure the propagation time at a k-th time step. For example, the slave node may measure the propagation time from the slave node to the master node at the k-th time step.

[0035] In operation 320, the slave node may compute an average propagation time value at a (k+1)-th time step. In operation 320, when k is less than a sliding window size, the slave node may compute the average propagation time value at the (k+1)-th time step according to a first computation formula. When k is greater than or equal to the sliding window size, the slave may compute the average propagation time value at the (k+1)-th time step according to a second computation formula. Here, "sliding window" may also be briefly expressed as "window". The sliding window size M may indicate a number of times that the measured propagation time is stored. M denotes a natural number. The sliding window size M may be associated with a size of a buffer included in the slave node.

[0036] Operation 320 may include operations 401, 403, and 405 of FIG. 4.

[0037] In operation 401, the slave node may determine whether k corresponding to a previous time step of the (k+1)-th time step is less than the sliding window size M. When k < M, operation 403 may be performed. Conversely, when k ≥ M, operation 405 may be performed.

[0038] In operation 403, the slave node may compute the average propagation time value at the (k + 1)-th time step according to the first computation formula.

[0039] In operation 405, the slave node may compute the average propagation time value at the (k + 1)-th time step according to the second computation formula. Here, the first computation formula and the second computation formula compute the average propagation time value at the (k+1)-th time step using an average propagation time value at the k-th time step and the propagation time measured at the k-th step time.

[0040] The first computation formula is associated with "gradually increasing a window size for a propagation time measurement when a system starts up and a time synchronization between nodes is initiated". The second computation formula is associated with "applying an exponential weight to a computation of an average propagation time value when a full window of propagation time measurements is collected".

[0041] The first computation formula may be defined by the following Equation 2:

where xk denotes the average propagation time value at the k-th time step and dk denotes the propagation time measured at the k-th time step. Here, for the computation of the above Equation 2, the slave node may include a unit to store the average propagation time value xk or kxk at the k-th time step.

[0042] The second computation formula may be defined by the following Equation 3:

where xk denotes the average propagation time value at the k-th time step, dk denotes the propagation time measured at the k-th time step, and a denotes a filter coefficient of a first-order digital filter. For the computation of the above Equation 3, the slave node may include a unit to store the average propagation time value xk at the k-th time step. The above Equation 3 may be computed using the first-order digital filter, for example, an infinite impulse response (IIR) filter.

[0043] In the above Equation 3, a may be set to increase as the sliding window size M increases. Accordingly, a may be computed according to the following Equation 4 or Equation 5:



where P denotes a constant. The above Equation 5 may be obtained by generalizing the above Equation 4.

[0044] The sliding window size M may be selected to be relatively large compared to a number of samples over which the propagation time dk measured at the k-th time step varies. Also, the sliding window size M may be selected to be relatively large compared to a time scale of a variation of an actual propagation time of a message. In any case, M may be set to 1000, however, may not be limited thereto.

[0045] In operation 330, the slave node may compute an offset based on a message transmission time at the master node, a message receipt time at the slave node, and the average propagation time value. The offset may be computed using "offset = T2 - T1 - Dave,k". In this instance, each of the slave node and the master node may compute and maintain the average propagation time value.

[0046] FIG. 5 illustrates a configuration of a network apparatus 500 according to an embodiment.

[0047] The network apparatus 500 may operate as a slave node.

[0048] The network apparatus 500 may include a propagation time measurement unit 510, an average propagation time value computing unit 520, and an offset computing unit 530.

[0049] The propagation time measurement unit 510 may measure a propagation time from a slave node to a master node at a k-th time step. Here, k denotes a natural number. The propagation time measurement unit 510 may measure the propagation time from the slave node to the master node using a time stamp. To measure the propagation time at a current time step, the propagation time measurement unit 510 may include, in a first message, a first time stamp that is a transmission time of the first message, and transmit the first message to the master node. Here, the master node may include, in a second message, the first time stamp, a second time stamp that is a receipt time of the first message, and a third time stamp that is a transmission time of the second message corresponding to the first message, and transmit the second message to the network apparatus 500. The propagation time measurement unit 510 may verify a fourth time stamp that is a receipt time of the second message. The propagation time measurement unit 510 may measure the propagation time at the current time stamp using the four time stamps.

[0050] The average propagation time value computing unit 520 may compute an average propagation time value at a (k+1)-th time step. The average propagation time value computing unit 520 may include a buffer to maintain M propagation time measurement values. When k is less than a sliding window size M, the average propagation time value computing unit 520 may compute the average propagation time value at the (k+1)-th time step according to a first computation formula. When k is greater than or equal to the sliding window size M, the average propagation time value computing unit 520 may compute the average propagation time value at the (k+1)-th time step according to a second computation formula.

[0051] The offset computing unit 530 may compute an offset based on a message transmission time at the master node, a message receipt time at the slave node, and the average propagation time value. The offset may be computed using "offset = T2 - T1 - Dave,k".

<A scheme of enhancing an accuracy of a propagation time measurement>



[0052] According to an embodiment, an accuracy of a propagation time measurement may be enhanced using an average propagation time value. Here, a time accuracy for a network synchronization may be determined by an application that needs a highest time accuracy among applications used in a corresponding node. For example, an application that needs a time synchronization between an image and a sound may need a relatively low time accuracy compared to an application that determines a frequency value for a mobile communication.

[0053] FIG. 6 illustrates a method of enhancing an accuracy in a propagation time measurement according to an embodiment.

[0054] Referring to FIG. 6, N-1 hops from a grandmaster 610 to a slave node 640 may exist. A total number of N nodes may exist. Different relay nodes may exist between a first relay node 620 and a second relay node 630. The first relay node 620 and the second relay node 630 may each correspond to a time-aware system.

[0055] In operation S601, the second relay node 630 may receive a first message from the slave node 640. The first message corresponds to a message that requests performance parameter information associated with a time accuracy. The first message may include a field n to indicate a number of hops. Here, a value of the field n may be "1".

[0056] Since the second relay node 630 is not a grandmaster node, the second relay node 630 may transmit a second message to the first relay node 620 in operation S602. The second message corresponds to a message that requests performance parameter information associated with a time accuracy. The second message may include a field n to indicate a number of hops. Here, a value of the field n may be "2".

[0057] Since the first relay node 620 is not the grandmaster node, the first relay node 620 may transmit a third message to the grandmaster 620 in operation S603. The third message may include a field n' to indicate a number of hops. Here, a value of the field n' may be "n' = n + 1".

[0058] The first message, the second message, and the third message may include performance parameter information of a corresponding node. The performance parameter information of the corresponding node may include at least one of a digital filter type, a digital filter size, a sliding window size, and a local clock performance. The second message and the third message may further include performance parameter information of a previous node. Specifically, the second message may include performance parameter information of the slave node 640 and performance parameter information of the second relay node 630. The third message may include performance parameter information of the slave node 640, performance parameter information of the second relay node 630, and performance parameter information of the first relay node 620.

[0059] Accordingly, when generally describing operations S601 through S603 as an operation performed at a random relay node, the description may be made as follows: When the random relay node is assumed as an (N-2)-th relay node, the (N-2)-th relay node may receive a performance parameter request message from an (N-1)-th relay node, and determine whether the (N-2)-th relay node is the grandmaster node. When the (N-2)-th relay node is not the grandmaster node, the (N-2)-th relay node may transfer, to an (N-3)-th relay node, the performance parameter request message received from the (N-1)-th relay node. Here, the performance parameter request message received from the (N-1)-th relay node may include performance parameter information of the (N-1)-th relay node. The performance parameter request message transferred to the (N-3)-th relay node may include performance parameter information of the (N-1)-th relay node and performance parameter information of the (N-2)-th relay node.

[0060] When the grandmaster 610 receives the third message, the grandmaster 610 may generate a performance table. The performance table may include performance parameter information of N-2 relay nodes and performance parameter information of the slave node 640. The grandmaster 610 may generate a performance information packet containing performance parameter information of all nodes, using the generated performance table, and may return the performance information packet. The performance information packet may be transferred to the slave node 640 through operations S604, S605, and S606.

[0061] The performance information packet may be updated in each node and then be transferred to a subsequent node. For example, the first relay node 620 may update information of the first relay node 620 that is included in the performance information packet and then transfer the updated performance information packet to the second relay node 630.

[0062] The slave node 640 or the grandmaster 610 may request each relay node to change a performance parameter by referring to the information performance packet or the performance table. For example, the slave node 640 or the grandmaster 610 may request each relay node to increase the sliding window size or to change a filter constant. When an application currently in use needs a high time accuracy, the slave node 640 may request each relay node to increase the sliding window size or to change the filter constant.

[0063] Operations S601 through S603 may be performed after operation S330 of FIG. 3. Each node may perform operations S601 through S606 prior to the propagation time measurement, or may periodically or aperiodically perform operations S601 through S606. Each node may periodically or aperiodically perform operations S601 through S606 to thereby enhance an accuracy of a propagation time measurement.

<Signaling for a parameter setting>



[0064] FIGS. 7 through 9 illustrate a signaling for a parameter setting according to an embodiment. The signaling of FIGS. 7 through 9 may be applicable when the slave node 640 or the grandmaster 610 requests each relay node to change a performance parameter. Also, the signaling may be applicable to set a parameter for computing an average propagation time value. The parameter for computing the average propagation value may be, for example, a window size. The signaling may be applicable to set a parameter necessary in a network. In FIGS. 7 through 9, time-aware bridges 2, 3, 4, 5, and 6 may perform a relay node function.

[0065] The signaling may be performed by transmitting a signaling message. The signaling message may carry information, requests, and/or commands between time-aware systems. The signaling message may transfer the information, the requests, and/or the commands using at least one type-length-value (TLV).

[0066] FIG. 7 illustrates an example where a time-aware end-station 1 requests a parameter setting. The time-aware end-station 1 may perform a slave node function. The time-aware end-station 1 may transfer a first signaling message with a request TLV to the time-aware bridge 5. The first signaling message may correspond to a message that requests to set a parameter for computation of the average propagation time value using values included in the request TLV.

[0067] The time-aware bridge 5 may receive the first signaling message via a master port, store information associated with the request TLV, and then transfer the request TLV using a second signaling message. Here, information associated with the request TLV may be, for example, a sequence number of the first signaling message and a TLV type. In this instance, information associated with the request TLV may be stored so that a request-denied TLV may be transferred only via an inverse path of a transfer path of the request TLV when the parameter setting request is denied.

[0068] As shown in FIG. 7, the request TLV may be forwarded to a grandmaster time-aware bridge 1 via the time-aware bridge 2. The grandmaster time-aware bridge 1 may perform the same function as the grandmaster 610 of FIG. 6. The grandmaster time-aware bridge 1 may also store information associated with the request TLV. The grandmaster time-aware bridge 1 may determine a grant or a denial with respect to the parameter setting request. Whether to grant or deny the parameter setting request may be determined based on performance parameter information.

[0069] FIG. 8 illustrates an example where a parameter setting request is denied at the grandmaster time-aware bridge 1. In FIG. 8, a request-denied TLV corresponds to a signaling message indicating that the parameter setting request is denied.

[0070] Referring to FIG. 8, when the time-aware bridge 2 receives the request-denied TLV, The time-aware bridge 2 may remove information associated with a pre-stored request TLV, and thereby transfer the request-denied TLV. As shown in FIG. 8, the request-denied TLV may be transferred to the time-aware end-station 1.

[0071] FIG. 9 illustrates an example where a parameter setting request is granted at the grandmaster time-aware bridge 1.

[0072] When the parameter setting request is granted, the grandmaster time-aware bridge 1 may set a parameter for computation of the average time propagation value using values included in the request TLV. The grandmaster time-aware bridge 1 may remove information associated with a pre-stored request TLV, and thereby transfer a request-granted TLV. The request-granted TLV indicates that the parameter setting request is granted. The request-granted TLV may be transferred via all master ports of the grandmaster time-aware bridge 1, which is different from the request-denied TLV. Accordingly, the time-aware bridges 2, 3, and 4 may receive the request-granted TLV.

[0073] The time-aware bridges 2, 3, and 4 may set the parameter for computation of the average propagation time value using values included in the request-granted TLV, and may transfer the request-granted TLV via corresponding master ports. In this instance, the time-aware bridge 2 may remove information associated with a pre-stored request TLV, and thereby transfer the request-granted TLV. When a time-aware end-station 2 receives the request-granted TLV, the time-aware end-station 2 may set the parameter for computation of the average propagation time value using the values included in the request-granted TLV. The time-aware end-station 2 may not transfer the request-granted TLV.

[0074] According to the signaling described with reference to FIGS. 7 through 9, a signaling message for a parameter setting may be transferred to a grandmaster. When a parameter setting request is granted, parameters included in the parameter setting request may be transferred to all of nodes included in a network. All the nodes included in the network may set the parameters included in the parameter setting request. Conversely, when the parameter setting request is denied, a parameter setting-denied message may be transferred via only a transfer path of the parameter setting request.

[0075] Examples include computer-readable media including program instructions to implement various operations embodied by a computer. The media may also include, alone or in combination with the program instructions, data files, data structures, tables, and the like. The media and program instructions may be those specially designed and constructed for the purposes of example embodiments, or they may be of the kind well known and available to those having skill in the computer software arts.

[0076] Although a few example embodiments have been shown and described, the present disclosure is not limited to the described example embodiments. Instead, it would be appreciated by those skilled in the art that changes may be made to these example embodiments.


Claims

1. A network synchronization method, comprising:

measuring a propagation time from a slave node (102, 202) to a master node (101, 201) at a k-th time step where k denotes a natural number;

computing an average propagation time value at a [k+1]-th time step according to a first computation formula, when k is less than a sliding window size, being a predetermined natural number M, and computing the average propagation time value at the [k+1]-th time step according to a second computation formula, when k is greater than or equal to the sliding window size M; and

computing an offset based on a message transmission time at the master node (103, 203), a message receipt time at the slave node (104, 204), and the average propagation time value at the [k+1]-th time step;

characterized by:

in the step of computing the average propagation time value, the first computation formula and the second computation formula compute the average propagation time value at the [k+1]-th time step using an average propagation time value at the k-th time step and the propagation time measured at the k-th step time.


 
2. The method of claim 1, where the first computation formula is computed according to the following Equation:

wherein xk denotes the average propagation time value at the k-th time step and

dk denotes the propagation time measured at the k-th time step.


 
3. The method of claim 1, wherein the second computation formula is computed according to the following Equation:

where xk denotes the average propagation time value at the k-th time step, dk denotes the propagation time measured at the k-th time step, and a denotes a filter coefficient of a first-order digital filter.
 
4. The method of claim 3, wherein a increases as the sliding window size M increases.
 
5. The method of claim 3, wherein a is computed according to the following Equation:

where M denotes the sliding window size.
 
6. The method of claim 3, wherein a is computed according to the following Equation:

where M denotes the sliding window size and P denotes a constant.
 
7. The method of any preceding claim, further comprising:

transmitting a performance parameter request message to a relay node, and receiving a performance information packet corresponding to the performance parameter request message.


 
8. The method of claim 7, further comprising:

requesting the relay node to change a performance parameter based on the performance information packet.


 
9. The method of claim 7, wherein the performance information packet includes at least one of a digital filter type, a digital filter size, the sliding window size M, and a local clock performance.
 
10. The method of any preceding claim, further comprising:

signaling a parameter setting request for computing of the average propagation time value.


 
11. The method of claim 10, further comprising:

receiving a grant or a denial with respect to the parameter setting request.


 
12. The method of claim 10, wherein, when the parameter setting request is granted, the parameter setting request is transferred to all nodes included in a network.
 
13. A network synchronization apparatus for performing a time synchronization between nodes, the apparatus comprising:

a propagation time measurement unit to measure a propagation time from a slave node (102, 202) to a master node (101, 201) at a k-th time step where k denotes a natural number;

an average propagation time value computing unit to compute an average propagation time value at a [k+1]-th time step according to a first computation formula, when k is less than a sliding window size, being a predetermined natural number M, and to compute the average propagation time value at the [k+1]-th time step according to a second computation formula, when k is greater than or equal to the sliding window size M; and

an offset computing unit to compute an offset based on a message transmission time at the master node (103, 203), a message receipt time at the slave node (104, 204), and the average propagation time value at the [k+1]-th time step;

characterized in that:

the average propagation time value computing unit uses the first computation formula and the second computation formula to compute the average propagation time value at the [k+1]-th time step using an average propagation time value at the k-th time step and the propagation time measured in the k-th step time.


 
14. The apparatus of claim 13, wherein the second computation formula is computed according to the following Equation:

where xk denotes the average propagation time value at the k-th time step, dk denotes the propagation time measured at the k-th time step, and a denotes a filter coefficient of a first-order digital filter.
 
15. The apparatus of claim 14 wherein a is computed according to the following Equation:

or wherein a is computed according to the following Equation:

where M denotes the sliding window size and P denotes a constant.
 


Ansprüche

1. Netzwerksynchronisationsverfahren, umfassend:

Messen einer Laufzeit von einem Slave-Knoten (102, 202) zu einem Master-Knoten (101, 201) an einem k-ten Zeitschritt, wobei k eine natürliche Zahl bezeichnet;

Berechnen eines durchschnittlichen Laufzeitwerts an einem [k+1]-ten Zeitschritt gemäß einer ersten Berechnungsformel, wenn k kleiner als eine Gleitfenstergröße ist, die eine vorgegebene natürliche Zahl M ist, und Berechnen des durchschnittlichen Laufzeitwerts an dem [k+1]-ten Zeitschritt gemäß einer zweiten Berechnungsformel, wenn k größer gleich der Gleitfenstergröße M ist; und

Berechnen eines Versatzes basierend auf einer Nachrichtensendezeit an dem Master-Knoten (103, 203), einer Nachrichtenempfangszeit an dem Slave-Knoten (104, 204) und dem durchschnittlichen Laufzeitwert an dem [k+1]-ten Zeitschritt;

dadurch gekennzeichnet, dass
bei dem Schritt des Berechnens des durchschnittlichen Laufzeitwerts die erste Berechnungsformel und die zweite Berechnungsformel den durchschnittlichen Laufzeitwert an dem [k+1]-ten Zeitschritt mittels eines durchschnittlichen Laufzeitwerts an dem k-ten Zeitschritt und der an der k-ten Schrittzeit gemessenen Laufzeit berechnen.
 
2. Verfahren nach Anspruch 1, wobei die erste Berechnungsformel gemäß der folgenden Gleichung berechnet wird:

wobei xk den durchschnittlichen Laufzeitwert an dem k-ten Zeitschritt bezeichnet und dk die an dem k-ten Zeitschritt gemessene Laufzeit bezeichnet.
 
3. Verfahren nach Anspruch 1, wobei die zweite Berechnungsformel gemäß der folgenden Gleichung berechnet wird:

wobei xk den durchschnittlichen Laufzeitwert an dem k-ten Zeitschritt bezeichnet, dk die an dem k-ten Zeitschritt gemessene Laufzeit bezeichnet und a einen Filterkoeffizienten eines digitalen Filters erster Ordnung bezeichnet.
 
4. Verfahren nach Anspruch 3, wobei a mit Zunehmen der Gleitfenstergröße M zunimmt.
 
5. Verfahren nach Anspruch 3, wobei a gemäß der folgenden Gleichung berechnet wird:

wobei M die Gleitfenstergröße bezeichnet.
 
6. Verfahren nach Anspruch 3, wobei a gemäß der folgenden Gleichung berechnet wird:

wobei M die Gleitfenstergröße bezeichnet und P eine Konstante bezeichnet.
 
7. Verfahren nach einem beliebigen vorhergehenden Anspruch, ferner umfassend:

Senden einer Leistungsparameteranforderungsnachricht an einen Weiterleitungsknoten und Empfangen eines der Leistungsparameteranforderungsnachricht entsprechenden Leistungsinformationspakets.


 
8. Verfahren nach Anspruch 7, ferner umfassend:

Anfordern, dass der Weiterleitungsknoten einen Leistungsparameter basierend auf dem Leistungsinformationspaket ändert.


 
9. Verfahren nach Anspruch 7, wobei das Leistungsinformationspaket mindestens eines aus der Gruppe umfassend einen Digitalfiltertyp, eine Digitalfiltergröße, die Gleitfenstergröße M und eine lokale Taktleistung aufweist.
 
10. Verfahren nach einem beliebigen vorhergehenden Anspruch, ferner umfassend:

Signalisieren einer Parametrieranforderung zum Berechnen des durchschnittlichen Laufzeitwerts.


 
11. Verfahren nach Anspruch 10, ferner umfassend:

Empfangen einer Freigabe oder Sperre in Bezug auf die Parametrieranforderung.


 
12. Verfahren nach Anspruch 10, wobei, wenn die Parametrieranforderung freigegeben ist, die Parametrieranforderung an alle in einem Netzwerk enthaltenen Knoten übermittelt wird.
 
13. Netzwerksynchronisationsvorrichtung zum Durchführen einer Zeitsynchronisation zwischen Knoten, wobei die Vorrichtung aufweist:

eine Laufzeitmesseinheit, um eine Laufzeit von einem Slave-Knoten (102, 202) zu einem Master-Knoten (101, 201) an einem k-ten Zeitschritt zu messen, wobei k eine natürliche Zahl bezeichnet;

eine Durchschnittslaufzeitwert-Berechnungseinheit, um einen durchschnittlichen Laufzeitwert an einem [k+1]-ten Zeitschritt gemäß einer ersten Berechnungsformel zu berechnen, wenn k kleiner als eine Gleitfenstergröße ist, die eine vorgegebene natürliche Zahl M ist, und um den durchschnittlichen Laufzeitwert an dem [k+1]-ten Zeitschritt gemäß einer zweiten Berechnungsformel zu berechnen, wenn k größer gleich der Gleitfenstergröße M ist; und

eine Versatzberechnungseinheit, um einen Versatz basierend auf einer Nachrichtensendezeit an dem Master-Knoten (103, 203), einer Nachrichtenempfangszeit an dem Slave-Knoten (104, 204) und dem durchschnittlichen Laufzeitwert an dem [k+1]-ten Zeitschritt zu berechnen;

dadurch gekennzeichnet, dass
die Durchschnittslaufzeitwert-Berechnungseinheit die erste Berechnungsformel und die zweite Berechnungsformel verwendet, um den durchschnittlichen Laufzeitwert an dem [k+1]-ten Zeitschritt mittels eines durchschnittlichen Laufzeitwerts an dem k-ten Zeitschritt und der in der k-ten Schrittzeit gemessenen Laufzeit zu berechnen.
 
14. Vorrichtung nach Anspruch 13, wobei die zweite Berechnungsformel gemäß der folgenden Gleichung berechnet wird:

wobei xk den durchschnittlichen Laufzeitwert an dem k-ten Zeitschritt bezeichnet, dk die an dem k-ten Zeitschritt gemessene Laufzeit bezeichnet und a einen Filterkoeffizienten eines digitalen Filters erster Ordnung bezeichnet.
 
15. Vorrichtung nach Anspruch 14, wobei a gemäß der folgenden Gleichung berechnet wird:

oder wobei a gemäß der folgenden Gleichung berechnet wird:

wobei M die Gleitfenstergröße bezeichnet und P eine Konstante bezeichnet.
 


Revendications

1. Procédé de synchronisation de réseau, consistant à :

mesurer un temps de propagation à partir d'un noeud esclave (102, 202) jusqu'à un noeud maître (101, 201) à un kième intervalle de temps où k représente un entier naturel ;

calculer une valeur de temps de propagation moyen à un [k + 1]ième intervalle de temps selon une première formule de calcul, lorsque k est inférieur à une taille de fenêtre coulissante, qui est un entier naturel prédéterminé M, et calculer la valeur de temps de propagation moyen au [k + 1]ième intervalle de temps selon une seconde formule de calcul, lorsque k est supérieur ou égal à la taille de fenêtre coulissante M ; et

calculer un décalage sur la base d'un temps de transmission de message au niveau du noeud maître (103, 203), d'un temps de réception de message au niveau du noeud esclave (104, 204) et de la valeur de temps de propagation moyen au [k + 1]ième intervalle de temps ;

caractérisé par le fait que:

au cours de l'étape de calcul de la valeur de temps de propagation moyen, la première formule de calcul et la seconde formule de calcul calculent la valeur de temps de propagation moyen au [k + 1]ième intervalle de temps à l'aide d'une valeur de temps de propagation moyen au kième intervalle de temps et du temps de propagation mesuré au kième temps d'intervalle.


 
2. Procédé selon la revendication 1, dans lequel la première formule de calcul est calculée selon l'équation suivante :

dans laquelle xk représente la valeur de temps de propagation moyen au kième intervalle de temps et

dk représente le temps de propagation mesuré au kième intervalle de temps.


 
3. Procédé selon la revendication 1, dans lequel la seconde formule de calcul est calculée selon l'équation suivante:

xk représente la valeur de temps de propagation moyen au kième intervalle de temps, dk représente le temps de propagation mesuré au kième intervalle de temps et a représente un coefficient de filtre d'un filtre numérique de premier ordre.
 
4. Procédé selon la revendication 3, dans lequel a augmente au fur et à mesure que la taille de fenêtre coulissante M augmente.
 
5. Procédé selon la revendication 3, dans lequel a est calculé selon l'équation suivante:

a = e-1/MM représente la taille de fenêtre coulissante.


 
6. Procédé selon la revendication 3, dans lequel a est calculé selon l'équation suivante:

où M représente la taille de fenêtre coulissante et P représente une constante.
 
7. Procédé selon l'une quelconque des revendications précédentes, consistant en outre à :

transmettre un message de demande de paramètre de performance à un noeud relais et recevoir un paquet d'informations de performance correspondant au message de demande de paramètre de performance.


 
8. Procédé selon la revendication 7, consistant en outre à :

demander au noeud relais de changer un paramètre de performance sur la base du paquet d'informations de performance.


 
9. Procédé selon la revendication 7, dans lequel le paquet d'informations de performance comprend un type de filtre numérique et/ou une taille de filtre numérique et/ou la taille de fenêtre coulissante M et/ou une performance d'horloge locale.
 
10. Procédé selon l'une quelconque des revendications précédentes, consistant en outre à :

signaler une demande de réglage de paramètre pour le calcul de la valeur de temps de propagation moyen.


 
11. Procédé selon la revendication 10, consistant en outre à :

recevoir un accord ou un refus par rapport à la demande de réglage de paramètre.


 
12. Procédé selon la revendication 10, dans lequel, lorsque la demande de réglage de paramètre est accordée, la demande de réglage de paramètre est transférée à tous les noeuds inclus dans un réseau.
 
13. Appareil de synchronisation de réseau pour effectuer une synchronisation temporelle entre des noeuds, l'appareil comprenant :

une unité de mesure de temps de propagation pour mesurer un temps de propagation à partir d'un noeud esclave (102, 202) jusqu'à un noeud maître (101, 201) à un kième intervalle de temps où k représente un entier naturel ;

une unité de calcul de valeur de temps de propagation moyen pour calculer une valeur de temps de propagation moyen à un [k + 1]ième intervalle de temps selon une première formule de calcul, lorsque k est inférieur à une taille de fenêtre coulissante, qui est un entier naturel prédéterminé M, et pour calculer la valeur de temps de propagation moyen au [k + 1]ième intervalle de temps selon une seconde formule de calcul, lorsque k est supérieur ou égal à la taille de fenêtre coulissante M ; et

une unité de calcul de décalage pour calculer un décalage sur la base d'un temps de transmission de message au niveau du noeud maître (103, 203), d'un temps de réception de message au niveau du noeud esclave (104, 204) et de la valeur de temps de propagation moyen au [k + 1]ième intervalle de temps ;

caractérisé par le fait que :

l'unité de calcul de valeur de temps de propagation moyen utilise la première formule de calcul et la seconde formule de calcul pour calculer la valeur de temps de propagation moyen au [k + 1]ième intervalle de temps à l'aide d'une valeur de temps de propagation moyen au kième intervalle de temps et du temps de propagation mesuré au kième temps d'intervalle.


 
14. Appareil selon la revendication 13, dans lequel la seconde formule de calcul est calculée selon l'équation suivante :

xk représente la valeur de temps de propagation moyen au kième intervalle de temps, dk représente le temps de propagation mesuré au kième intervalle de temps et a représente un coefficient de filtre d'un filtre numérique de premier ordre.
 
15. Appareil selon la revendication 14, dans lequel a est calculé selon l'équation suivante:

ou dans lequel a est calculé selon l'équation suivante:

où M représente la taille de fenêtre coulissante et P représente une constante.
 




Drawing






























REFERENCES CITED IN THE DESCRIPTION



This list of references cited by the applicant is for the reader's convenience only. It does not form part of the European patent document. Even though great care has been taken in compiling the references, errors or omissions cannot be excluded and the EPO disclaims all liability in this regard.

Non-patent literature cited in the description