JP2004336354A - Method to use uplink band in optical burst transmission/reception network - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a method to efficiently use an uplink transmission band, even if there is frequency deviation between an OLT (Optical Line Termination) and an ONU (Optical Network Unit) and variable-length frame data are transmitted. <P>SOLUTION: In an optical burst transmission/reception network, a plurality of ONUs share a transmission medium and a transmission band, an OLT 20 measures a transmission time according to the distance from it to each ONU and notifies each ONU of a time delay according to respective measurement value, and each ONU appends delay to a transmission signal to the OLT according to the value. The OLT 20 establishes a slot start time (SST) for a send-out time of a REPORT message for each ONU. According to the slot start time (SST), a time delay Tofst from a GATE message reception time at an ONU 10 to a REPORT message send-out time becomes a predetermined constant period for each ONU. The OLT 20 transmits a GATE message including the established slot start time (SST) to each ONU. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
According to the present invention, a plurality of child devices share a transmission medium and a transmission band, a parent device measures a transmission time according to a distance from each child device, and transmits to each child device an uplink signal corresponding to each measured value. Notifying the scheduled transmission time, each child device controls the transmission time of the transmission signal to the parent device according to the value, the parent device controls the allocation of the used band of each child device, and each child device The present invention relates to a method of using an upstream band in an optical burst transmission / reception network that transmits data to a parent device based on control.
[0002]
[Prior art]
With an optical burst transmission / reception network, a plurality of child devices share a transmission medium and a transmission band, a parent device measures a transmission time according to a distance from each child device, and each child device responds to each measured value. Notify the delay time, each child device adds a delay to the transmission signal to the parent device according to the value, the parent device controls the allocation of the used band of each child device, and each child device performs the control of the parent device. A network based on a method of transmitting data to a parent device based on this.
[0003]
In such an optical burst transmission / reception network, Non-Patent Document 1 discloses a method in which a plurality of child devices share a transmission medium and a transmission band, and each child device transmits data to the parent device by controlling the band of the parent device. Technology. FIG. 9 shows a network configuration shown in FIG. 5 of Non-Patent Document 1, which includes n ONUs (Optical Network Units) and one OLT (Optical Line Termination). Here, the OLT corresponds to the parent device, and the ONU corresponds to the child device. ODN (Optical Distribution Network) is a transmission medium including an optical fiber and an optical multiplexer / demultiplexer.
[0004]
FIG. 10 shows the format of data from the OLT to the ONU (hereinafter referred to as “down”) and the format of data from the ONU to the OLT (hereinafter referred to as “up”) (for example, FIG. 11 in Non-Patent Document 1). reference). In this configuration, transmission of fixed-length cells of 53 bytes in the downlink direction and 56 bytes in the uplink direction is assumed, data in the downlink direction is broadcast from the OLT to all ONUs, and data in the uplink direction is each time slot of each cell. Data transmission is performed in such a manner that transmission of one ONU is permitted in advance by the OLT. Further, it is assumed that all cell data in the upstream direction transmitted from each ONU are transmitted by a clock generated in synchronization with the frequency of the downstream data transmitted from the OLT.
[0005]
Since the allocation of time slots in an upstream frame is controlled by the OLT for all ONUs connected to the OLT, as long as each ONU obeys this control, each upstream time slot is only controlled by one ONU. Can be used. However, it is impossible to transmit data by actually sharing the upstream transmission path only by this band control function. The reason is that the distance between each ONU and the OLT varies, and the phase of the upstream frame recognized by each ONU does not match when viewed from the OLT. According to Non-Patent Document 1, the distance between each ONU and the OLT may vary within a range of 0 to 20 km, and the variation in the distance difference is caused by the actual geographical positions of the accommodation station and the subscriber's house. This is a condition that must be taken into account when considering the relationship.
[0006]
This distance difference appears as a difference in data transmission delay, so that the OLT sends the transmission delay to each of the ONUs so that the uplink data transmitted from each ONU does not collide at the time of arrival at the OLT without collision. It is necessary to measure each ONU, determine the amount of additional delay in the ONU based on the measurement result, and notify each ONU. At the same time, each ONU needs to add the notified delay amount to the uplink data and transmit it to the OLT. This additional delay amount is smaller for ONUs having a larger measured value, and is larger for ONUs having a smaller measured value, and the phase of an upstream frame recognized by each ONU with respect to a downstream frame transmitted from the OLT is higher than the OLT. Set to match. In Non-Patent Document 1, the delay measurement operation performed by the OLT is referred to as ranging.
[0007]
In the prior art shown in Non-Patent Document 1, as described above, since each ONU is subordinately synchronized with the OLT, as long as the delay time itself measured as described above does not change, when viewed with the OLT, Are actually matched and received. However, in Non-Patent Document 1, the measurement time is periodically updated even after the end of the delay measurement in consideration of the change in the transmission delay time due to the environmental conditions of the transmission path and the like. This periodic update corrects a slight phase shift from the OLT standby phase, and does not re-execute the delay measurement.
[0008]
The 56 bytes of data in the upstream direction have a 3-byte area called a PON overhead in addition to the 53-byte ATM cell as in the downstream direction. This includes an unused portion for data transmission so that even if the phase of the uplink data from each ONU is slightly shifted, it does not collide with the preceding and following data. This area is called a guard time, and this length is set by the OLT notifying each ONU of a common value.
[0009]
Since the delay measurement is accurately performed up to the bit unit of the transmission rate, the phase shift is several bits, and the guard time needs to be at most about 8 bits. Therefore, in this method, time-division multiplexing of uplink burst data is performed very efficiently.
[0010]
On the other hand, there is also a method of configuring a PON system that performs transmission in a format of a MAC frame (Ethernet (registered trademark) frame). In this method, transmission data is limited to a MAC frame, but since it can be easily and inexpensively connected to an IP network, in recent years when the Internet has spread, ITU-T Recommendation G. This is an important access network configuration method separately from the 983.1 method. In this method, the conventional MAC frame transmission method is used for transmission on the PON as it is. The fixed-length data transmission and the dependent synchronization of the upstream data to the downstream data, which are the prerequisites of the 983.1 system, are not used, and the independent clock operation is performed in the variable length frame. This method has been discussed for standardization mainly in IEEE802.3ah Ethernet (registered trademark) in the First Mile (EFM) Task Force or the like. A delay measurement method different from the conventional example according to the 983.1 standard is used. A method of measuring a transmission delay (RTT: Round Trip Time) between the OLT and the ONU according to this method is performed as shown in FIG. 11 (for example, see Non-Patent Document 2).
[0011]
In the method according to Non-Patent Document 2, as a precondition, the OLT and the ONU manage time using timers that operate on their own clocks. In FIG. 11, T1, T2, T3, and T5 are times by the OLT timer, and T4 is time by the ONU timer. The OLT timer is not reset or corrected during operation, but the ONU timer performs an operation of adjusting the time based on time information from the OLT.
[0012]
The OLT transmits a message signal GATE in the downstream direction. In this message, T1 is written as the time when this message was transmitted. This transmission time is called “time stamp; Time Stamp (TS)”. Upon receiving this message, the ONU sets the current time of its own timer to this time T1. In addition to the time stamp (TS), the GATE message includes “slot start time (SST)”, which is the time at which the ONU transmits data in the upstream direction, and “slot length; Slot,” which is the length of data. Length (SL) "is written.
[0013]
FIG. 12 shows a GATE message transmission / reception process described in Non-Patent Document 3. When the ONU's own timer reaches this SST, the ONU transmits data in the uplink direction, and transmits a message REPORT in the uplink. At this time, the transmission time is written in the message as in the case where the OLT has transmitted GATE. In FIG. 11, the time when the ONU receives the GATE message in which the time T1 is written as the time stamp (TS) is T2 as seen by the OLT timer. Also, the time when the ONU transmits the REPORT message is T3 as viewed by the OLT timer and T4 as viewed by the ONU timer. Also, the time when the OLT receives the REPORT message is T5 as seen by the OLT timer. At this time, the transmission delay (RTT) can be calculated as RTT = T5-T4 from the relationship of RTT = T2-T1 + T5-T3, T3-T2 = T4-T1.
[0014]
According to the calculation method, the transmission delay (RTT) is obtained only by the transmission / reception time of the REPORT message, and the relative phase of the REPORT message with respect to a specific signal (for example, a GATE message) output from the OLT becomes irrelevant. For this reason, OLT complies with ITU-T Recommendation G. There is no need to measure a delay value with respect to a reference phase as in the 983.1 method, and a measurement result can be obtained only by comparing a simple time stamp value. Further, the delay measurement at the time of starting the ONU in the initial stage and the correction of the delay value at the time of operation can be performed by the same calculation method.
[0015]
However, in the method of performing the delay measurement as described above, if there is a frequency deviation between the clocks of the OLT and the ONU, the value of the transmission delay (RTT) may not be accurately obtained.
[0016]
FIG. 13 shows how the calculated value of the RTT is obtained when the clock frequency of the OLT is 1/20, that is, 5% higher than that of the ONU. For the sake of simplicity, the time of each timer in the OLT and ONU is displayed as hours to minutes, such as 10:00. Here, it is assumed that the RTT value actually observed by the OLT clock is 20 minutes. In FIG. 13, the GATE message transmitted at 10:00 is received by the ONU at 10:10 after 10 minutes, and at the same time, the ONU timer is set to 10:00. Here, it is assumed that the time at which the ONU outputs the REPORT message is 10:20 after 20 minutes in accordance with the SST written in the GATE message. However, this time is on the ONU timer. Since the OLT timer is ahead of the ONU by 5%, it takes 21 minutes from the reception of the GATE message to the transmission of the REPORT message, and the one-way transmission from the OLT to the ONU. Since the time of 10 minutes has passed, the OLT timer is 10:31. Since the time transmitted from here to the OLT is 10 minutes by the OLT timer, the reception time of the REPORT message is 10:41 by the OLT timer. In this case, the value of RTT is 21 minutes in the calculation method shown in FIG.
[0017]
Similarly, FIG. 14 shows an example in which the time from reception of the GATE message to transmission of the REPORT message is further doubled to 40 minutes. The calculated value of RTT in this case is 22 minutes as a result of performing the same processing.
[0018]
The OLT assumes a value obtained by adding the RTT to the SST written in the GATE message for the reception time of the uplink method data from each ONU. That is, if the OLT correctly recognizes the actual RTT value, the OLT should prepare a slot for data reception from 10:40 in the example of FIG. 13 and from 11:00 in the example of FIG. However, as shown in the above example, if there is a clock frequency deviation between the OLT and the ONU, and there is a variation in the time difference between the reception of the GATE message and the transmission of the REPORT message, the larger the difference, the more the error in the RTT measurement. Becomes larger. Therefore, the error also increases in the reception time of the upstream data from the ONU as seen from the OLT, and it is necessary to take a large guard time as a margin period for multiplexing with the upstream data from another ONU.
[0019]
When the guard time is increased, the time required to transmit the same amount of data becomes longer, so that the use efficiency with respect to the actual transmission speed decreases.
[0020]
Further, in the MAC frame method, another factor for taking extra guard time is a variable data length. However, this factor still becomes a problem when there is a frequency deviation.
[0021]
FIG. 15 shows, as an example similar to FIG. 13, an operation in which an ONU transmits uplink data at 10:20 by a GATE message. The only difference from FIG. 13 is that the slot length (SL) in the uplink direction is equivalent to 20 minutes. Since the OLT has written 20 as SL in the GATE message and transmitted it to the ONU, the ONU transmits data from its own timer at 10:20 for 20 minutes as instructed. Therefore, the transmission of the data ends at 10:40 by the ONU timer. However, as in the example of FIG. 13, this time is 10:52 according to the OLT timer. The data reception end time of the OLT is 11:02. Since the data reception start and end times obtained from the correct RTT = 20 and Slot Length = 20 are 10:40 and 11:00, respectively, the error in the data transmission / reception end time is larger than the start time.
[0022]
Further, FIG. 16 shows a case where the SL is equivalent to 40 minutes. In this case, if the same calculation is performed, the data reception end time is 11:23, and the deviation from the logical end time 11:20 is 3 minutes. 14 and 15 are larger than those in the examples of FIGS.
[0023]
When such a shift in the uplink data reception end time occurs in the OLT, it becomes necessary to shift the reception start time of the ONU data allocated to the next slot backward, that is, the guard time is extended. The magnitude of the deviation also increases as the variable range of the frequency deviation and the slot length (SL) increases.
[0024]
[Non-patent document 1]
ITU-T Recommendation G. 983.1, "Broadband optical access systems based on Passive Optical Networks (PON)", 1998/10
[Non-patent document 2]
On Haran et al., "Presentation Materials MPCP: Timing Model", IEEE 802.3ah Ethernet (registered trademark) in the First Mile (EFM) Task Force March, 2002, p. 6
[Non-Patent Document 3]
Dolors Sara et al., "Presentation Materials MPCP: MPCP Baseline Proposal Architecture and Layering Model", IEEE 802.3ah Ethernet, (registered trademark) in the Internet. 7
[0025]
[Problems to be solved by the invention]
In summary, the ITU-T scheme shown in Non-Patent Document 1 uses the uplink transmission band efficiently, but the premise is that the fixed length cell is used and the OLT and the ONU are completely synchronized. Since transmission is not possible, Internet connection services cannot be easily provided.
[0026]
Further, the IEEE method described in Non-Patent Documents 2 and 3 can directly transmit a MAC frame. It is necessary to take a large amount, and the transmission band cannot be used efficiently.
[0027]
SUMMARY OF THE INVENTION The present invention has been made in view of the above, and has an optical burst transmission / reception network that has a frequency deviation between an OLT and an ONU and can efficiently use an uplink transmission band even when transmitting variable-length frame data. It is an object of the present invention to obtain a method of using an upstream band.
[0028]
[Means for Solving the Problems]
In order to achieve the above object, a method of using an upstream band in an optical burst transmission / reception network according to the present invention includes a master device and a plurality of devices each having a timer whose operation clock frequency is asynchronous and managing the time of its own device by the timer. A child device, wherein the parent device transmits a first specific signal including a transmission time of the first specific signal and a scheduled transmission time of the second specific signal in the child device to each child device, and each child device transmits When the first specific signal is received, the time of the timer of the own device is adjusted to the transmission time of the first specific signal included in the first specific signal, and included in the first specific signal from the time of the time adjustment. By transmitting a second specific signal including the transmission time of the second specific signal to the scheduled transmission time by transmitting a second specific signal to the parent device by measuring the time until the scheduled transmission time of the second specific signal by the timer, The device measures a transmission time according to a distance from each child device based on a transmission time included in the second specific signal from the child device and a reception time of the second specific signal at the parent device. In the method of using the upstream band in the optical burst transmission / reception network, the parent device may be configured such that a delay time from a reception time of the first specific signal to a transmission time of the second specific signal is a predetermined fixed time for each child device. The transmission schedule time of the second specific signal is transmitted to each child device by including the scheduled time in the first specific signal.
[0029]
According to the present invention, the parent device is configured to set the second specific signal such that the delay time from the reception time of the first specific signal to the transmission time of the second specific signal is a predetermined fixed time for each child device. Is included in the first specific signal and transmitted to each child device, so that the scheduled reception time of uplink data as seen from the parent device does not change, and the transmission delay conventionally required Guard time due to time fluctuation is not required. Therefore, it is possible to efficiently use the upstream transmission band.
[0030]
BEST MODE FOR CARRYING OUT THE INVENTION
Hereinafter, preferred embodiments of a method for using an upstream band in an optical burst transmission / reception network according to the present invention will be described in detail with reference to the accompanying drawings.
[0031]
Embodiment 1 FIG.
In the first embodiment, the interval Tofst from the GATE message reception time to the REPORT transmission time, which is an arbitrary interval in the related art, is fixed (always a predetermined fixed time) for each ONU.
[0032]
Hereinafter, the first embodiment will be described with reference to FIGS. FIG. 1 shows a configuration of an optical burst transmission / reception network (PON: Passive Optical Network) to which the present invention is applied. In this optical burst transmission / reception network, a plurality of child devices share a transmission medium and a transmission band, and each child device transmits data to the parent device by controlling the band of the parent device. ONUs (Optical Network Units) 10-1 to 10-n (hereinafter referred to as ONU 10 when representing one ONU), an OLT (Optical Line Termination) 20 as one parent device, an optical fiber and An ODN (Optical Distribution Network) 30 is provided as a transmission medium constituted by an optical multiplexer / demultiplexer or the like.
[0033]
The operating clock frequencies of the OLT 20 and the ONU 10 are asynchronous, and a variable-length frame such as a MAC frame is basically used as a frame format to be transmitted.
[0034]
The basic operation of the method of measuring the transmission delay (RTT: Round Trip Time) between the OLT 20 and the ONU 10, that is, the transmission time, according to the first embodiment will be described with reference to FIG. In the present measurement method, as a precondition, the OLT 20 and the ONUs 10-1 to 10-n each manage the time by a timer that operates with a clock independently provided, and the operation clock is asynchronous. In FIG. 2, T1, T2, T3, and T5 are times by the timer of the OLT 20, and T4 is time by the timer of the ONU 10. The timer of the OLT 20 is not reset or corrected during operation, but the timer of the ONU 10 performs an operation of adjusting the time based on the time information from the OLT 20.
[0035]
The OLT 20 transmits a message signal (first specific signal) GATE in the downstream direction. In this case, T1 is written in the Gate message as the time at which the Gate message was transmitted. This transmission time information is called “time stamp; Time Stamp (TS)”. In this case, the TS value included in the GATE message is “TSd”. When receiving the GATE message, the ONU 10 sets (sets the time to) the current time of the timer of its own device to the value (T1 in this case) specified by the transmission time information TSd. The GATE message includes, in addition to the transmission time information TSd, “slot start time; Slot Start Time (SST)” as scheduled transmission time information that is the time at which the ONU 10 transmits data in the upward direction, and the length of the slot (data). "Slot Length (SL)" is written as the slot length information.
[0036]
FIG. 3 shows the transmission / reception processing of the GATE message. When the time value of the ONU 10 reaches the time designated by the slot start time (SST), the ONU 10 transmits data in the upstream direction. When transmitting the transmission data, the ONU 10 transmits a message signal REPORT (second specific signal). Send As in the case where the OLT 20 has transmitted the GATE message, transmission time information (here, “TSu”) of the REPORT message indicating the time at which the REPORT message was transmitted is written in the REPORT message.
[0037]
In FIG. 2, the time when the ONU 10 receives the GATE message in which the time T1 is written as the transmission time information TSd is T2 (= T1 + RTT / 2) as seen by the OLT timer. The RTT is a pure transmission delay time (transmission time) without error, which is the sum of the transmission delay time (transmission time) from the OLT 20 to the ONU 10 and the transmission delay time (transmission time) from the ONU 10 to the OLT 20. is there. The time when the ONU 10 transmits the REPORT message is T3 (= T1 + RTT / 2 + Tofst (1 + Δf)) as viewed by the OLT timer, and T4 (= SST = TSu) as viewed by the ONU timer. Δf is a frequency deviation between the OLT 20 and the ONU 10. The time when the OLT 20 receives the REPORT message is T5 (T1 + RTT + Tofst (1 + Δf)) as seen by the OLT timer. At this time, the transmission delay time RTTm (RTT value) actually calculated by the OLT 20 is RTTm = T5-T4 from the relationship of RTTm = T2-T1 + T5-T3, T3-T2 = T4-T1, as described above. Can be calculated by
[0038]
Here, when the frequency deviation Δf exists between the OLT 20 and the ONU 10, the RTT value (RTTm) obtained by the above calculation method corresponds to the time (T1) at which the ONU 10 sets its own timer by receiving the GATE message. RTTm differs from RTT as a pure transmission delay time by Tofst × Δf, depending on the value of the time Tofst until the transmission time T4 at which the REPORT message is transmitted. That is,
RTTm = RTT + Tofst × Δf
It becomes.
[0039]
The reason for this is that during the period in which the ONU 10 counts with the timer of the ONU 10 for Tofst from the time when the ONU 10 receives the GATE message to the scheduled transmission time (SST) T4 of the upstream signal in the GATE message, the timer of the OLT 20 uses Tostst × ( This is because the time elapses by 1 + Δf). That is, the error between both timers during this period is Tofst × Δf.
[0040]
Therefore, in the method of using the upstream band in the PON according to the present invention, the phase relationship between the GATE message and the REPORT message of each ONU is controlled so as to have a fixed phase as shown in FIG. In other words, the OLT 20 sets the scheduled transmission time of the REPORT message transmission time such that the delay time Tofst from the reception time of the GATE message at the ONU 10 to the transmission time of the REPORT message becomes a predetermined fixed time for each ONU. (SST) is set for each ONU, and a GATE message including the set scheduled transmission time (SST) is transmitted to each ONU.
[0041]
The operation of the OLT 20, the ONU 10-1, and the ONU 10-2 will be described with reference to FIG. The OLT 20 transmits a GATE message to each of the ONUs 10-1 and 10-2, and sets an SST value such that a delay time Tofst from a reception time of the GATE message at the ONU 10 to a transmission time of the REPORT message is always constant. I do. In this case, the SST for ONU 10-1 is Tofst1 (fixed value), and the SST for ONU 10-2 is Tofst2 (fixed value).
[0042]
When receiving the GATE message for each, ONU 10-1 and ONU 10-2 transmit a REPORT message at the data transmission start time (scheduled transmission time) obtained based on the SST value. As a result, when focusing on one ONU, the time interval from the GATE reception time to the REPORT transmission time is always constant. For example, in the case of ONU 10-1, it is always constant at Tofst1, and in the case of ONU 10-2, it is always constant at Tofst2.
[0043]
The frequency deviation between the OLT 20 and the ONU 10-1 is Δf1, the frequency deviation between the OLT 20 and the ONU 10-2 is Δf2, and the original RTTs of the ONU 10-1 and the ONU 10-2 measured by the OLT clock are RTT1 and RTT2. Assuming that the results of the delay measurement processing are RTTm1 and RTTm2,
RTTm1 = RTT1 + Tofst1 × Δf1,
RTTm2 = RTT2 + Tofst2 × Δf2
Therefore, if Tofst1 and Tofst2 are fixed values, the delay measurement results RTTm1 and RTTm2 also become fixed values, respectively.
[0044]
Actually, there is also an element in which the actual delay time fluctuates due to a change with time of the transmission path delay or the like. Therefore, in the first embodiment, the OLT 20 transmits the GATE message to each ONU at a predetermined constant interval for each ONU, so that the measurement of the transmission delay time RTTm is repeatedly executed at a constant interval. I have to. Then, by appropriately selecting the measurement period, it is possible to follow the long-period fluctuation and suppress the short-period fluctuation, thereby optimizing the response to the delay fluctuation.
[0045]
If the delay measurement result for each ONU is constant, the element of the guard time is the error due to the relative time difference between the GATE message reception time and the REPORT transmission time, and the relative time difference between the data end time in the previous data slot and the GATE message transmission time. It is only an error due to the time difference, and the scheduled reception time of the uplink data viewed from the OLT 20 does not change. Therefore, according to the first embodiment, the guard time due to the fluctuation of the transmission delay time, which is conventionally required, becomes unnecessary.
[0046]
As described above, according to the first embodiment, since the guard time due to the fluctuation of the RTT measurement value of each of the ONUs 10-1 to 10-n is not required, the uplink transmission band is used efficiently. It becomes possible. In addition, since the RTT measurement is performed periodically, it is possible to cope with a change in an actual transmission delay.
[0047]
Embodiment 2 FIG.
In the second embodiment, in the delay measurement method according to the first embodiment, the delay time Tofst from the reception time of the GATE message at the ONU 10 to the transmission time of the REPORT message, that is, the value of the SST is set to the minimum value of the applicable range. Things.
[0048]
FIG. 5 shows the phase relationship between the GATE message and the REPORT message of each ONU according to the second embodiment. In the second embodiment, OLT 20 sets the value of SST for notifying a slot to be given to each ONU for RTT measurement as a minimum value within an applicable range. Other operations of the OLT 20 and the ONU 10 are the same as those of the first embodiment.
[0049]
As can be seen from the description of the first embodiment, the RTTm as a result of the actual delay measurement processing is expressed as RTTm = RTT + Tofst × Δf using the original RTT and the frequency deviation Δf, and the interval Tofst between the GATE message and the REPORT message. expressed. Therefore, when Tofst is the applicable minimum value Min, RTTm is a value closest to the original RTT value.
[0050]
FIG. 6 is a graph showing the relationship between the difference between RTTm and RTT and Tofst. When the range of the frequency deviation Δf is ± Lim and the maximum applicable value of Tofst is Max, the difference between RTTm and RTT when the difference between RTTm and RTT is the minimum value Min is ± Pmin, and when the maximum value is Max. Is ± Pmax.
[0051]
The effect of the difference between RTTm and RTT on band use efficiency will be described. As described in FIG. 13 and the like, the difference between RTTm and RTT is equal to the difference between the scheduled reception time of uplink data based on RTT viewed from the OLT 20 and the actual reception time. Therefore, the value of RTTm-RTT in FIG. 6 also represents a shift in the reception time of uplink data based on RTT. That is, when the expected reception time is calculated based on the RTT, if the frequency deviation is unknown, a deviation of ± Pmax may actually occur. When the expected reception time is calculated based on RTTm measured at Tofst = Max, for example, if the frequency deviation is −Lim, RTTm−RTT is the value of point B in the figure, and in fact, All data arrives only in the plus direction, but since the frequency deviation is unknown, the OLT 20 must also consider the shift on the minus side. The shift assumed in this case is ± 2 Pmax. Similarly, when the estimated reception time is calculated based on RTTm calculated with Tofst = Min, any one of which must be assumed is ± (Pmax + Pmin).
[0052]
As is clear from FIG. 6, when 2Pmax> Pmax + Pmin, and when using RTTm measured with Tofst = Min, the expected value of the difference between the expected reception time and the actual reception time of the data is minimized. Can be. Therefore, the guard time between the upstream and downstream data can be minimized, so that the transmission band can be used more efficiently.
[0053]
As described above, according to the second embodiment, the error between the measured delay value RTTm and the actual transmission path delay can be minimized, so that the error in the scheduled data reception time based on the error can be minimized, and the guard time can be reduced. Since it can be minimized, the transmission band can be used more efficiently.
[0054]
Further, in the OLT 20, if the ONU having a larger frequency deviation Δf from the OLT 20 has a smaller GST message and transmits an SST value such that the interval Tofst between the GATE message and the REPORT message becomes smaller, the ONT is transmitted to each ONU. The error between the delay measurement value RTTm and the actual transmission path delay between ONUs can be made constant, so that it is not necessary to take a large margin when setting the guard time, and as a result, the guard time can be reduced. Can be.
[0055]
Embodiment 3 FIG.
In the third embodiment, the delay measurement method in the second embodiment is further extended, and the interval Tofst from the GATE message to the REPORT message is set to the minimum value Min of the applicable range and the maximum value Max. RTTm is measured, and the actual RTT value and the frequency deviation Δf are estimated using each measurement result.
[0056]
FIG. 7 shows the phase relationship between the GATE message and the REPORT message of each ONU according to the third embodiment.
[0057]
In the third embodiment, the OLT 20 sends a GATE message for REPORT transmission to each of the ONUs 10-1 to 10-n with both Tofst = Min and Tofst = Max. Each of the ONUs 10-1 to 10-n returns a REPORT message according to each of these two GATE messages. At this time, uplink slot allocation may be performed simultaneously by one GATE message, or may be performed by separate GATE messages. The OLT 20 stores the delay measurement value RTTma when Tofst = Min and the delay measurement value RTTmb when Tofst = Max, respectively.
[0058]
RTTma and RTTmb are calculated using the original transmission delay time RTT and the frequency deviation Δf.
RTTma = RTT + MIN × Δf (1)
RTTmb = RTT + Max × Δf (2)
And using these relations,
Δf = (RTTmb−RTTma) / (Max−Min) (3)
Holds. Therefore, the frequency deviation Δf is obtained from the equation (3). Further, by using the obtained Δf, the value of RTT can be obtained by using the above equation (1) or equation (2).
[0059]
As described in the second embodiment, the difference between the scheduled data reception time and the actual data reception time is determined by the frequency deviation Δf and the time difference between the GATE message and the REPORT message, that is, Tofst. Is used, the difference between the scheduled data reception time and the actual data reception time can be predicted, and the scheduled data reception time can be matched with the actual reception time. That is, the Tofst value in the first embodiment, that is, the SST value is determined using the RTT and Δf obtained above.
[0060]
For example, as shown in FIG. 5, when the GATE message transmitted at time T1 is instructed to transmit uplink data at SST = Tofst, the reception time T5 is calculated using known T1, Tofst, RTT, and Δf.
T5 = T1 + RTT + Tofst × (1 + Δf)
[0061]
In this case, the cause of the error in the reception time based on the variable range of the RTT measurement and the Tofst can be set to zero, so that there is no need to arrange a guard time corresponding to the cause. Further, by using the SL value assigned to each ONU, the end time of the received data can be further predicted, and the error of the data end time based on the variable range of the data length can be reduced to zero.
[0062]
As described above, according to the third embodiment, the reception start time and the reception end time of uplink data can be predicted without error based on the frequency deviation Δf, the time difference Tofst between the GATE message and the REPORT message, and the variable range of the uplink data length SL. Therefore, the guard time can be made substantially zero, and the transmission band can be used efficiently.
[0063]
In the third embodiment, the maximum value Max and the minimum value Min are used as the two different Tofsts. However, the two different values are not limited to the maximum value Max and the minimum value Min. Any value between the value Max and the minimum value Min may be used.
[0064]
Embodiment 4 FIG.
In the fourth embodiment, the maximum value of the deviation amount of the data reception start time and the maximum value of the deviation amount of the data reception end time in the OLT 20 are predicted according to the magnitude of the Tofst value and the magnitude of the data length (slot length) SL. Then, a guard time according to the predicted value is set.
[0065]
FIG. 8 shows a phase relationship between the GATE message and the REPORT message of each ONU according to the fourth embodiment.
[0066]
In the fourth embodiment, OLT 20 stores information in the GATE message transmitted by itself, that is, transmission time information (TS), scheduled transmission time information (SST), and slot length information (SL) of the GATE message. From the difference between the stored SST value and the TS value, how much time difference is required for the uplink data from each ONU to be transmitted from the GATE message, that is, the Tofst value is obtained. The guard time is set such that the smaller the calculated Tofst value and data length value, the smaller the deviation from the scheduled uplink data reception time and the larger the deviation, the larger the deviation from the scheduled reception time.
[0067]
In FIG. 8, the start of the reception of DATAa is determined by TofstA, the end of the reception of DATAa is determined by TostA + SL, and the start of DATAb is determined by TofstB. When Tj2 to Tj4 are set in order, in the case of FIG. 8, Tj2 <Tj3 <Tj4. Similarly, if the shift amount of the end of DATAc is Tj1, the guard time GTa between DATAc and DATAa is (Tj1 + Tj2) / 2, and the guard time GTb between DATAa and DATAb is (Tj3 + Tj4) / 2.
[0068]
As described above, according to the fourth embodiment, the guard time that always assumes the maximum Tofst or the data length SL can be variably set instead of the guard time that assumes the maximum Tofst or the data length SL. The transmission band can be used efficiently.
[0069]
【The invention's effect】
As described above, according to the present invention, the parent device causes the delay time from the reception time of the first specific signal to the transmission time of the second specific signal to be a predetermined fixed time for each child device. Since the scheduled transmission time of the second specific signal is included in the first specific signal and transmitted to each child device, the scheduled reception time of the uplink data as seen from the parent device does not change. In this case, a guard time due to a necessary change in transmission delay time is not required. Therefore, it is possible to efficiently use the upstream transmission band.
[Brief description of the drawings]
FIG. 1 is a conceptual block diagram illustrating a configuration example of an optical burst transmission / reception network to which the present invention is applied.
FIG. 2 is a time chart for explaining a basic measurement operation of a transmission delay RTT between an OLT and an ONU.
FIG. 3 is a diagram showing a GATE message transmission / reception process between an OLT and an ONU.
FIG. 4 is a time chart for explaining a measurement operation of a transmission delay RTTm according to the first embodiment.
FIG. 5 is a time chart for explaining a measurement operation of a transmission delay RTTm according to a second embodiment.
FIG. 6 is a diagram illustrating a relationship between a difference between RTTm and RTT and Tofst.
FIG. 7 is a time chart for explaining a measurement operation of a transmission delay RTTm according to a third embodiment.
FIG. 8 is a time chart for explaining the operation of the fourth embodiment.
FIG. 9 is a diagram showing a conventional technique.
FIG. 10 is a diagram showing a data format according to the related art.
FIG. 11 is a time chart for explaining a basic measurement operation of a transmission delay RTT between an OLT and an ONU according to another conventional technique.
FIG. 12 is a diagram showing a GATE message transmission / reception process between the OLT and the ONU.
FIG. 13 is a time chart for explaining a problem of the related art.
FIG. 14 is a time chart for explaining a problem of the related art.
FIG. 15 is a time chart for explaining a problem of the related art.
FIG. 16 is a time chart for explaining a problem of the related art.
[Explanation of symbols]
10 (10-1 to 10-n) ONU (child device), 20 (OLT) parent device, 30 ODN.

Claims (6)

It has a master device and a plurality of slave devices each having a timer whose operation clock frequency is asynchronous and managing the time of the own device by the timer, and the master device transmits the first specific signal at the sending time and the second slave device at the slave device. The first specific signal including the scheduled transmission time of the second specific signal is transmitted to each child device, and when each child device receives the first specific signal, the time of the timer of its own device is included in the first specific signal. The time is adjusted to the transmission time of the first specific signal, and the time from the time adjustment time to the scheduled transmission time of the second specific signal included in the first specific signal is measured by a timer. A second specific signal including the transmission time of the second specific signal at the scheduled time is transmitted to the parent device, and the parent device transmits the transmission time included in the second specific signal from the child device and the second specific signal. On the parent device In upstream bandwidth usage in optical burst receiving network which is adapted to measure the transmission time corresponding to the distance between the child device based on the signal time,
The parent device is configured to transmit the second specific signal such that a delay time from the reception time of the first specific signal to the transmission time of the second specific signal becomes a predetermined fixed time for each child device. A method of using an upstream band in an optical burst transmission / reception network, wherein a time is included in the first specific signal and transmitted to each child device. 2. The transmission time measurement is executed at a constant interval by transmitting the first specific signal to each slave device at a predetermined constant interval for each slave device. A method of using an upstream band in an optical burst transmission / reception network. The delay time from the reception time of the first specific signal to the transmission time of the second specific signal is set to a minimum value of an applicable range. A method of using an upstream band in an optical burst transmission / reception network. The scheduled transmission time of the second specific signal such that the delay time from the reception time of the first specific signal to the transmission time of the second specific signal becomes smaller as the child device has a larger frequency deviation from the parent device. 3. The method according to claim 1, wherein the first specific signal is transmitted to each child device. The parent device sets two different values as the delay time from the reception time of the first specific signal to the transmission time of the second specific signal, and sets the two different values to the set delay time. Two first specific signals each including the scheduled transmission time of the second specific signal are transmitted to each child device, and a second specific signal transmitted from the child device is received according to the two first specific signals. 2. The method according to claim 1, further comprising: calculating a deviation of an operation clock frequency between the master device and the slave device based on two transmission time measurement results based on the measurement result; and determining the predetermined fixed time based on the calculation result. 3. The method of using an upstream band in the optical burst transmission / reception network according to 2. The first specific signal also includes a slot length assigned to each child device,
The parent device is configured based on a slot length included in the first specific signal and a time length from a transmission time of the first specific signal included in the first specific signal to a scheduled transmission time of the second specific signal. 3. The method according to claim 1, wherein an interval between each slot allocated to each child device is controlled.

Images