JP3444397B2 - Digital signal receiving apparatus, method, and storage medium storing program according to the method - Google Patents

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a transmission system for packetizing and transferring a digital audio signal, and more particularly to a digital signal receiving apparatus and method, and a storage medium storing a program according to the method.

[0002]

2. Description of the Related Art When transmitting a digital audio signal, a method of transferring N audio signals (samples) generated in each service interval into one packet is considered. The number N of samples in each packet is not constant unless the sampling frequency of the audio signal is constant and the service interval is an integral multiple of the sampling period unless both are completely synchronized.
Therefore, in the case of such packetization, for example, if the time of occurrence of the sample at the beginning of the packet is added to the packet as a time stamp, the time stamp will exist for each non-constant number of samples.

On the receiving side of the audio signal, for example, an analog PLL (Phase-Locked Loop)
Is used to generate the same sampling clock as on the transmitting side. Here, it is well known that when a sampling clock is generated from a series of time stamps, if the time stamp intervals are constant and a power of 2, the implementation becomes simple. This is because the receiving side can predict the time when the next time stamp will be sent, and the configuration may be adapted to it.

[0004]

However, when a sampling clock is generated from a series of time stamps that exist for each non-constant sample number as described above, packetization becomes complicated and transmission / reception delay becomes long. Problems such as occur.

The present invention is a system for transmitting a digital audio signal, wherein a time stamp is sent from the transmitting side for each non-constant number of samples, and a sampling clock is generated at the receiving side. It is an object of the present invention to provide a digital signal receiving apparatus, a method, and a storage medium that stores a program related to the method, which can generate the digital signal and can be easily mounted with a simple configuration.

[0006]

In order to achieve the above object, a digital signal receiving apparatus according to a first aspect of the present invention provides a packet data including a plurality of consecutive samples and a time stamp indicating a time position of a leading sample of the samples. And estimating the temporal position of the first sample based on the time stamp of the received packet data, and arranging samples other than the first sample at predetermined time intervals from the temporal position of the first sample. Means for estimating the temporal position of each sample, and means for correcting the value of the time interval according to the difference between the estimated temporal position and the time stamp of the next packet data, Means for generating a sampling clock from the temporal position of each sample.

According to a second aspect of the present invention, there is provided a digital signal receiving method, which comprises a step of receiving packet data including a plurality of consecutive samples and a time stamp indicating a temporal position of a leading sample of the samples, and a time stamp of the received packet data. Estimating the temporal position of the first sample based on the above, and estimating the temporal position of each sample by arranging samples other than the first sample at predetermined time intervals from the temporal position of the first sample. Modifying the value of the time interval according to the difference between the estimated temporal position and the time stamp of the next packet data, and generating a sampling clock from the estimated temporal position of each sample. It is characterized by including.

A storage medium storing a digital signal receiving program according to claim 3, wherein the program receives packet data including a plurality of consecutive samples and a time stamp indicating a temporal position of a leading sample of the sample. And estimating the temporal position of the first sample based on the time stamp of the received packet data, and arranging samples other than the first sample at predetermined time intervals from the temporal position of the first sample. Estimating the temporal position of
Modifying the value of the time interval according to the difference between the estimated temporal position and the time stamp of the next packet data; and generating a sampling clock from the estimated temporal position of each sample. It is characterized by including.

[0009]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings.

FIG. 4 shows a configuration example of a transmission system to which the present invention is applied. From the transmitting node 401 to the transmission line 404
A digital audio signal is transmitted to the receiving node 402 via the receiver at regular time intervals (called service intervals). The transmitted / received digital audio signal is packet data including a plurality of samples and a time stamp. Within the receiving node 402 is a digital PLL 403.
Is provided, and a sampling clock for the audio signal is generated from the received packet data. The transmitting node 401 and the receiving node 402 each operate on a reference clock (GLOBAL) operating at a frequency sufficiently higher than the frequency of the sampling clock of the audio signal to be transferred.
_TIME), and these reference clocks shall always be adjusted to point to the same time with a certain accuracy.

FIG. 5 is a time chart showing how packetization is performed on the transmitting side. 502 is a time axis, 501
Indicates each sample of the digital audio signal arranged on the time axis in a cycle according to a predetermined sampling clock. At 501, the positions numbered as samples 1, 2, 3, ... (Positions marked with *) are the temporal positions of the respective samples. These samples 1,
Packet data is created by combining 2, 3, ...

The packet data is constructed by setting a time stamp at the beginning of the data and the number of samples included in the packet, and subsequently setting the samples in order. 511 to 513 are samples 1 to 1 shown in 501.
4 is a packetized example. Packet data 511
Has a time stamp T1 set at the beginning, and is the number of samples included in this packet data next to "5".
Is set. After that, samples 1 to 5 are set in order. The time stamp is a value indicating the temporal position of the first sample of the samples included in the packet. For example, the time stamp T1 of the packet data 511 indicates the temporal position of sample 1, which is the first sample in this packet. Similarly, the packet data 512 and 513 also have time stamps T from the beginning.
2, T3, the number of samples, and each sample are set in this order.

The packet data 511 to 51 as described above
3 is transmitted from the transmission node 401 in FIG. 4 to the reception node 402 via the transmission line 404. In the receiving node 402, a sampling clock is generated by the digital PLL 403 based on the received packet data 511 to 513. At that time, it is desired to reproduce the sampling clock on the transmitting side described in FIG. 5 as accurately as possible. Therefore, in the system of this embodiment, the sampling clock is estimated on the receiving side by the digital PLL 403 as shown in FIG. Hereinafter, this estimation method will be described.

FIG. 6 is a time chart for explaining the estimation method of the sampling clock on the receiving side. Here, the temporal position of each received sample is estimated, and a sampling clock is generated based on the estimated position. Hereinafter, the temporal position of each sample is estimated as the estimated time Loca.
Call lTime. LocalTime is a variable that stores the value of the estimated time. A symbol that represents a variable such as LocalTime represents the variable itself and also a value stored in the variable (for example, LocalTime).
If so, it also represents the estimated time).

At the reception node 402 (FIG. 4), when the packet data 511 (FIG. 5) is received, the estimated time LocalTim of sample 1, which is the first sample of the packet data 511, is determined by the time stamp T1.
Let e be T1. On the other hand, at this point, it is not guaranteed whether or not the next packet data 512 is received. Therefore, for samples 2 to 5 included in the packet data 511, the estimated time of Sample 1, LocalTime = T1.
From the first time to the predetermined time interval Δt1. A line segment 603 in FIG.
5 shows a state in which 5 is arranged at each time interval Δt1. The estimated time LocalTime of each sample 2 to 5 is obtained by adding Δt1 to the value of the estimated time of the immediately preceding sample.

Further, the estimated time Loc of the next sample 6
Similarly, alTime is obtained. The position of the end point 6 ′ of the line segment 603 in FIG. 6 is the estimated time LocalTi of the sample 6.
It becomes the position of me. Below that, LocalTime = LocalTime + Δt1 * 5 is the estimated time LocalTime of sample 6.
Is the estimated time LocalTime (= T
1) shows that it is obtained by adding the product of Δt1 and 5 (there are five time intervals Δt1 from sample 1 to sample 6). Note that the Local on the left side of the above equation
Time indicates the estimated time of sample 6, and Loc on the right side
Let alTime indicate the estimated time for sample 1.

In the case of FIG. 6, the packet data 512 including the time stamp T2 (samples 6 to 6) at time T2.
9) is received. Here, the estimated time LocalTime of the sample 6 was estimated to be the position of the end point 6 ′ of the line segment 603 as described above, but since the packet data 512 is received and the time stamp T2 thereof is acquired, the accuracy of the sample 6 is accurate. It can be seen that the correct temporal position is T2.
Therefore, the estimated time LocalTime of the sample 6 is changed to T2 which is an accurate time position. The position of the end point 6 ′ estimated as the position of the sample 6 and the accurate position T
With 2, the estimated temporal position is ahead, so
The correction is performed so as to reduce the time interval for arranging the samples 7 to 9. Specifically, Δt2 (= Δt1−α) is obtained by subtracting a predetermined value α from the time interval Δt1 used when obtaining the estimated times of Samples 2 to 5, and this is obtained from Samples 7 to
9 time intervals. That is, the packet data 512
For samples 7 to 9 included in the above, the time interval Δt from the estimated time LocalTime = T2 of sample 6
Place every two. A line segment 604 in FIG. 6 shows a state in which samples 7 to 9 are arranged at time intervals Δt2 from the position of T2 of sample 6.

Further, the estimated time Lo of the next sample 10 is
calTime is calculated in the same manner. Line segment 604 in FIG.
Is the estimated time Loca of the sample 10.
It becomes the position of lTime. Below that, LocalTime = LocalTime + Δt2 * 4 is the estimated time LocalTime of the sample 10.
e is the estimated time of the sample 6, LocalTime (= T
2) shows that the product of Δt2 and 4 (there are four time intervals Δt2 from sample 6 to sample 10) is added. Note that the Loca on the left side of the above equation
lTime indicates the estimated time of sample 10, and L on the right side
ocalTime indicates the estimated time of sample 6.

In the case of FIG. 6, packet data 513 (samples 10 to 10) including the time stamp T3 at time T3.
14) is received. Here, although the estimated time LocalTime of the sample 10 was estimated to be the position of the end point 10 ′ of the line segment 604 as described above, the packet data 513
Is received and its time stamp T3 is obtained, it can be seen that the accurate temporal position of the sample 10 is T3. Therefore, the estimated time Local of the sample 10 is
Change Time to T3, which is the exact time position. Further, since the estimated temporal position is delayed between the position of the end point 10 ′ estimated as the position of the sample 10 and the accurate position T3, correction is made to increase the time interval in which the samples 11 to 14 are arranged. To do. Specifically, the time interval Δt2 used when obtaining the estimated times of Samples 7 to 9
Is added with a predetermined value β to obtain Δt3 (= Δt2 + β),
This is the time interval of samples 11-14. That is, the samples 11 to 1 included in the packet data 513
For 4, the estimated time LocalTi of sample 10
It is arranged at every time interval Δt3 from me = T3. A line segment 605 in FIG. 6 shows a state in which the samples 11 to 14 are arranged at time intervals Δt3 from the position of T3 of the sample 10.

Similarly, each time packet data is received, the first sample of the packet data is arranged based on an accurate time stamp, and the time interval Δt is calculated based on the amount of deviation between the time stamp and the estimated time. Is corrected and samples are arranged at each time interval Δt to obtain an estimated time of each sample. A sampling clock is generated based on the estimated time of each sample.

FIG. 1 and FIG. 2 are flowcharts showing a processing procedure for obtaining an estimated time LocalTime of each sample on the receiving side, and a digital PLL on the receiving side.
403 shows a specific procedure for executing the processing described in FIG. Figure 1 shows a sample with a timestamp (for example,
FIG. 2 shows the processing in the case of Samples 1, 6, and 10) described in FIGS. 5 and 6, and FIG. 2 is a sample having no time stamp (for example, Samples 2 to 2 described in FIGS. 5 and 6).
5, 7-10, 11-14).

In FIG. 1, in the case of a sample having a time stamp, first, in step 101, the time stamp is obtained from the received packet and set in timestamp (variable). Next, in step 102, the current estimated time L
ocalTime is time stamp timestamp
Is exceeded, that is, whether the estimated time is ahead of the time stamp.

When the estimated time is ahead in step 102, in step 103, the time is not reversed, and Work = LocalTime LocalTime = timestamp is set. Work is a work variable for saving the current estimated time. “Do not reverse the time” means, for example, in the example of FIG. 6, when the estimated time of the sample 5 is ahead of the time stamp T2, the estimated time of the sample 6 is T
When set to 2, the sample 5 is placed after the sample 6 and the time is reversed, so that such a situation should be prevented.

When a time reversal due to such advance occurs, the Loc of the process of step 103 is performed.
The process of alTime = timestamp is not executed (that is, the time stamp value is not used), and the current estimated time LocalTime is used as it is. Reference numerals 606 and 607 in FIG. 6 show an example in which the estimated time is used as it is without using the time stamp value. 606 is 6
The same as 03. 607 indicates samples 6 to 10 from the position of the estimated time 6 ′, not the position of the time stamp value T2.
Are arranged.

Next, in step 104, Work (estimated time before processing in step 103 is saved) and t
The difference from imestamp is a predetermined threshold value THRESHOL
It is determined whether D_GAIN or more. If so, in step 105, the predetermined value S is obtained from the time interval Δt.
LOWER is subtracted and the process returns. Step 104
And the difference between Work and timestamp is the threshold value THR.
When it is not equal to or more than EHOLD_GAIN, the process directly returns. It should be noted that Δt is a variable in which the time interval Δt1, Δt2, or Δt3 described in FIG. 6 is set. The predetermined value SLOWER corresponds to the predetermined value α described in FIG.

If in step 102 the estimated time is equal to or behind the time stamp, step 106.
Then, Work = LocalTime LocalTime = timestamp is set. Next, in step 107, Work (estimated time before processing in step 106 is saved) and t
The difference from imestamp is a predetermined threshold value THRESHOL
It is determined whether D_LOSE or more. If so, in step 108, the time interval Δt is set to the predetermined value FA.
Add STER and return. In step 107,
The difference between Work and timestamp is the threshold THRES.
If it is not more than HOLD_LOSE, the process directly returns. The predetermined value FASTER corresponds to the predetermined value β described in FIG.

A method of obtaining the estimated time in the case of a sample having no time stamp will be described with reference to FIG. In this case, in step 201, the estimated time is obtained from LocalTime = LocalTime + Δt, and the process returns.

FIG. 3 shows details of step 103 in FIG. Step 103 is replaced with steps 301 and 302 in FIG.
Should be replaced with In step 301, the estimated time Lo
Timestamp timestamp from calTime
It is determined whether or not the difference obtained by subtracting is greater than or equal to the time interval Δt. When the above difference is Δt or more, LocalT
If you set time = timestamp, the time will be reversed, so the current Local
In step 302, Work = is used because Time is used.
After setting only LocalTime, the process proceeds to step 104. If the above difference is not equal to or greater than Δt in step 301, no time reversal occurs, so in step 303 Work = LocalTime and Loc.
After setting alTime = timestamp, the process proceeds to step 104.

From the above, the receiving side digital PLL 40
In 3, it is possible to generate the same sampling clock as the sampling clock on the transmitting side.

The processing of steps 103 and 106 is as follows.
It may be modified in various ways, including not doing anything. Also,
In the process of FIG. 1, THRESHOLD_GAI
Since the operation changes depending on the values of N, THRESHOLD_LOSE, FASTER, and SLOWER, it is possible to adjust the manner of temporal arrangement of each sample by appropriately controlling these values. Also, FAS
The values of TER and SLOWER are Work and timestamp evaluated in step 104 or 107.
If it is determined according to the magnitude of the difference between and, Δt can be converged to an appropriate value earlier.

In the above embodiment, the digital PLL 4
Although the sampling clock is generated by 03, an analog PLL may be used after this digital PLL 403.

[0032]

As described above, according to the present invention, the receiving side estimates the temporal position of the leading sample based on the time stamp of the received packet data, and samples other than the leading sample. The temporal position of each sample is estimated by arranging from the temporal position of the first sample at predetermined time intervals, and the time interval is determined according to the difference between the estimated temporal position and the time stamp of the next packet data. Since the value of is corrected, the system can accurately generate the sampling clock of the transmitting side even in the system where the transmitting side sends a timestamp for each non-constant number of samples and the receiving side generates the sampling clock. In addition to being possible, there is an effect that it can be easily mounted with a simple configuration.

[Brief description of drawings]

FIG. 1 is a flowchart showing a processing procedure for obtaining an estimated time of a sample having a time stamp.

FIG. 2 is a flowchart showing a processing procedure for obtaining an estimated time of a sample having no time stamp.

FIG. 3 is a detailed diagram of step 103

FIG. 4 is a diagram showing a configuration example of a transmission system to which the present invention is applied.

FIG. 5 is a time chart diagram showing how packetization is performed on the transmission side.

FIG. 6 is a time chart diagram for explaining a sampling clock estimation method on the receiving side.

[Explanation of symbols]

401 ... Transmission node, 402 ... Reception node, 403 ... Digital PLL, 404 ... Transmission line, 511-513 ... Packet data.

─────────────────────────────────────────────────── ─── Continuation of the front page (56) Reference JP-A-4-362827 (JP, A) JP-A-9-64860 (JP, A) JP-A-9-149016 (JP, A) JP-A-5- 14399 (JP, A) JP-A-8-111685 (JP, A) JP-A-6-30043 (JP, A) JP-A-11-136224 (JP, A) (58) Fields investigated (Int.Cl. ⁷ , DB name) H04L 12/56 H04L 7/00 H04L 12/28

Claims (3)

(57) [Claims] 1. A means for receiving packet data including a plurality of consecutive samples and a time stamp indicating a temporal position of a head sample of the sample, wherein the time of the head sample is based on the time stamp of the received packet data. And a means for estimating the temporal position of each sample by arranging samples other than the first sample at predetermined time intervals from the temporal position of the first sample, and the estimated temporal position. And means for correcting the value of the time interval according to the difference between the time stamp of the next packet data and means for generating a sampling clock from the estimated temporal position of each sample. Signal receiving device. 2. A step of receiving packet data including a plurality of consecutive samples and a time stamp indicating a time position of a head sample of the sample, and a time step of the head sample based on the time stamp of the received packet data. Estimating the position and estimating the temporal position of each sample by arranging samples other than the first sample at predetermined time intervals from the temporal position of the first sample; and the estimated temporal position. A digital signal comprising the steps of modifying the value of the time interval according to the difference from the time stamp of the next packet data, and generating a sampling clock from the estimated temporal position of each sample. Receiving method. 3. A storage medium storing a digital signal reception program, the program receiving packet data including a plurality of consecutive samples and a time stamp indicating a temporal position of a leading sample of the samples. , Estimating the temporal position of the first sample based on the time stamp of the received packet data, and arranging samples other than the first sample at predetermined time intervals from the temporal position of the first sample Estimating the temporal position, modifying the value of the time interval according to the difference between the estimated temporal position and the time stamp of the next packet data, and estimating the temporal position of each sample Generating a sampling clock from the digital signal. A storage medium that stores a signal reception program.