JP2001111580A - Data transfer system - Google Patents

Abstract

PROBLEM TO BE SOLVED: To improve data transfer time by specifying a data transfer path between nodes and optimizing the transfer speed of the data. SOLUTION: This data transfer system having a plurality of electronic equipment is provided with a means, which acquires the connection information on the plurality of electronic equipment and transfer speed information about transferring the data of each of the electronic equipment, specifies a data transfer path on the basis of the connection information and also decides the speed, at which the data are transferred according to the transfer speed information.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a data transfer system provided with means for acquiring connection information of a plurality of electronic devices and transfer speed information for transferring data of each of the electronic devices.

[0002]

2. Description of the Related Art Conventionally, as an example of a digital interface used for electronic equipment (hereinafter, referred to as "node") such as digital AV equipment and computer peripheral equipment, there is an IEEE1394 interface. The IEEE 1394 interface is an interface standardized by IEEE (I Triple E) as a high-speed serial interface for next-generation multimedia.

[0003] A node can output data to another terminal after synchronizing data input from one terminal with a local clock output from a local clock circuit (not shown). This means that it has a function of relaying the input and output of signals, and as described later, data output from any node is transmitted to all other nodes in principle.

The transfer speed of each node is currently 10
0Mbps, 200Mbps, 400Mbps 3
The type is defined. Of these, 100 Mbps is guaranteed to be supported by all nodes. However, when data is transmitted / received at a speed of 200 Mbps or more, there is a data transfer source node, a data transfer destination node, and a data transfer between these nodes. It is necessary that all the data transfer speeds of the nodes to be executed be 200 Mbps or more. The transfer speed will be described in detail again with reference to the drawings.

[0005] The data transfer rate of each node can be determined by information contained in a self-identification packet output from each node at the time of bus reset. Further, by acquiring the self-identification packet, the number of nodes connected on the bus, the connection state of the bus, and the like can be known.

FIG. 14 is a diagram showing a topology (tree connection) based on IEEE1394. As shown in FIG. 14, each of the nodes 30 to 36 has a plurality of ports p0 to p
2 are provided, and the ports are connected by a cable of a maximum length of 4.5 m.

In the IEEE 1934 standard, the bus is reset when a new node is connected to the bus or when the power of the connected node is turned on / off. Each of the nodes 30 to 36 is assigned a physical layer identification number (hereinafter referred to as “phy_ID number”) 00 to 06 when the bus is initialized. phy_I
The one with the largest D number is called the root node. FIG.
Here, the personal computer 36 corresponds to this.

[0008] A terminal node at the end of the topology is called a leaf node. In FIG. 14, the CD-ROM disk 30, digital camera 31, and printer 33 correspond to this. Nodes other than the root node and leaf node are referred to as branch nodes. In FIG. 14, the stereo interface 32, the scanner 34 and the magnetic disk 35
Corresponds to this.

A port connected in the direction of the personal computer 36 as a root node is called a "parent port". For example, in the scanner 34, the port p
2 corresponds to this. On the other hand, a port connected to a direction farther from the personal computer 36 as opposed to the parent port is referred to as a “child port”. In the scanner 34, the port p1 corresponds to this. Note that a port that is not connected is “not connected” or “off”.

Next, data transfer and transfer speed will be described. For example, the stereo interface 32 of FIG.
Data output from port p via cable 2
Input from 2. The signal is output from the port p1 in synchronization with a local clock output from a local clock circuit (not shown).

The output packet data is input from the port p0 of the digital camera 31 via the cable 1. Thus, the stereo interface 32 transfers the data output from the personal computer 36 to the digital camera 31.

Here, for example, the data transfer rate of the personal computer 36 and the digital camera 31 is set to 4
Assuming that the speed of the stereo interface 32 is 200 Mbps and that of the stereo interface 32 is 200 Mbps,
6 must have a data rate of 200 Mbps or less.

Similarly, if the data transfer speed of the personal computer 36, the magnetic disk 35, the scanner 34, and the printer 33 is 400 Mbps, the speed of the data output from the personal computer 36 may be 400 Mbps or less.

FIG. 15 is a time chart showing processing after the bus is reset. As shown in FIG. 15, when a bus reset occurs, first, tree identification is performed, and then self-identification is performed. Then, after the self-identification has been performed, the bus manager having the restart of the cycle master, the assignment of the isochronous resource, and the management right becomes active.

Further, the bus manager becomes active 125 ms after the self-identification. Then, 625 ms elapses after the self-identification, and the isochronous resource manager becomes active. Also, a new isochronous resource is allocated 1000 ms after self-identification.

FIG. 16 is a diagram showing a state in which an isochronous transfer area and an asynchronous transfer area are mixed on a bus. In FIG. 16, an isochronous transfer area and an asynchronous transfer area coexist with one cycle of 125 μm.

In order to prevent a plurality of nodes from accessing one bus at the same time, an operation called arbitration is performed. Thus, when the isochronous packet and the asynchronous packet coexist on the bus, the isochronous packet has priority.

The asynchronous transfer area includes a period in which data is flowing on the bus, a period in which some short idle periods called sub-action gaps are combined, and a subsequent arbitration reset gap in the equal period. .

After each arbitration reset gap, arbitration for determining a node using the bus is performed next. In this section, each node is fairly given the right to access the bus. Thus, a plurality of nodes
This prevents simultaneous access to two buses.

[0020]

As described above, conventionally,
The data transfer rate at each node is 100 Mbps,
Although the three standards of 200 Mbps and 400 Mbps are defined in the standard, there is no technology for specifying the connection state of each node. For example, the data transfer speed of all nodes connected to the bus must be 200 Mbps or more. , The data was transferred at 100 Mbps uniformly.

The gap time is also recommended to be optimized according to the standard, and the optimum value is calculated. The gap count is "63", which is the default maximum value, and
The fixed value "42" was used. This is set to “63” when the bus is initialized according to the standard, and the maximum of 16 physical connections (hops) is possible as a standard. This is because the value becomes “42”.

Therefore, if the data transfer path between the nodes is specified and the data transfer speed is optimized, the data transfer speed can be improved.

Therefore, an object of the present invention is to improve the data transfer time by specifying the data transfer path between each node and optimizing the data transfer speed.

[0024]

In order to achieve the above object, the present invention provides a data transfer system having a plurality of electronic devices, which transfers connection information of the plurality of electronic devices and data of each of the electronic devices. Means for acquiring transfer speed information to specify a data transfer path based on the connection information, and determine a transfer speed of the data according to the transfer speed information.

[0025]

Embodiments of the present invention will be described below with reference to the drawings.

FIG. 1 is a configuration diagram of an apparatus for optimizing a data communication speed according to an embodiment of the present invention. As shown in FIG. 1, this device includes a storage device 1 that stores a connection state between each node and a data transfer speed of each node, and a data transfer path based on information stored in the storage device 1. And a data processing device 2 for performing identification and the like.

The storage device 1 stores a node information storage unit 11 which stores connection states of nodes output from each node onto the bus after a bus reset, and a data transfer path specified by the connection states of the nodes. A data transfer path storage unit 12 for storing data, a transfer speed storage unit 13 for storing an optimum transfer speed for each data transfer destination, and a gap count table 14 for storing gap counts calculated and presented according to the standard. I have.

The data processing device 2 includes a node information storage unit 1
1 to calculate the optimum transfer speed for each data transfer destination based on the information of the data transfer path specifying unit 21 for specifying the data transfer path and the like based on the information of 1 and the information of the node information storage unit 11 and the data transfer path storage unit 12. A transfer rate calculation unit 22, a maximum hop number calculation unit 23 that detects the number of hops between leaf nodes and calculates the maximum hop number, and a gap count change unit 24 that changes the gap count based on the maximum hop number and the gap count And

FIG. 2A is an internal configuration diagram of the node information storage unit 11. In the present embodiment, the node information storage unit 11
For example, a table memory is used. FIG.
FIG. 2B is a diagram showing a state where data indicating the port status is stored in the table memory of FIG. 2A, and stores data in the case of a topology as shown in FIG.

FIG. 2 (c) is a diagram for explaining the terms used in FIG. 2 (b).
“S400” indicates that the transfer bit rate is “98.30”.
4 Mbit / s "" 98.304 Mbit / s, 19
6.608 Mbit / s ”and“ 98.304 Mbit / s ”
s, 393.216 Mbit / s ± 100 ppm ”.

P, C, NC and NE are parent, child,
Each port is not present on the physical layer and the physical layer.
By the way, as a standard, since ports are supported up to p0-P26, it is assumed that the area for storing the port status can be extended up to the 2nd to 28th columns as needed.

As shown in FIGS. 2A and 2B, for example, each row of the table memory indicates a node identification number (NODE_ID) assigned to each node.
Each column indicates whether the node is a leaf node, the transfer speed, and the parent / child / unconnected status of each port. These data are stored in a self-identification packet output from each node onto the bus after the bus reset.

When the self-identification packet is input, the node information storage unit 11 shown in FIG. 1 extracts another data of parent / child / off assigned to each port from the self-identification packet.
(A) is stored in the second to fourth columns ([2] to [4]). Subsequently, the transfer rate is extracted and stored in the first column ([1]). Thereafter, by confirming whether or not a child port exists in the data stored in the second and subsequent columns, it is determined whether or not the node is a leaf node.
Store in the column ([0]).

It should be noted that the data stored in the 0th row only needs to indicate whether or not the node is a leaf node.
By checking whether or not a parent port exists in each data shown in the second to fourth columns, the 0th column specifies whether the node is a root node / leaf node / branch node. It may be stored.

FIG. 3A is an internal configuration diagram of the data transfer path storage unit 12. In this embodiment, a table memory is used as the data transfer path storage unit 12, for example. FIG. 3B is a diagram showing a state in which the phy_ID numbers assigned to the respective nodes are stored in the table memory of FIG. 3A, whereby the number of data transfer paths and the number of nodes interposed in the route are shown. Can be specified.

As shown in FIGS. 3 (a) and 3 (b), for example, each row of the table memory indicates the number of data transfer paths, and each column indicates the number of nodes interposed in the data transfer path and the allocation to the nodes. Phy_ID is shown.
That is, in FIG. 3, the phy_ID is “6-5-4-3”.
, A data transfer path “6-2-1”, and a data transfer path “6-2-0”. The data shown in FIG. 3 relates to the topology shown in FIG.

FIG. 4A is an internal configuration diagram of the transfer speed storage unit 13. FIG. 4B is a diagram illustrating a state in which the data transfer speed set for each node is stored in the storage unit illustrated in FIG. Numerals such as [0] shown in FIGS. 4A and 4B indicate phyIDs 0 to 6 shown in FIG. 14 allocated to each node, and data is transferred to the node. The transferable speed is stored in.

FIG. 5 is an internal configuration diagram of the gap count list 14. The value shown in FIG. 5 is calculated according to the standard, and stores a gap count related to the number of hops.

Next, the operation of this embodiment will be described. Here, as shown in FIG. 15, conventionally, tree identification and self-identification are performed after a bus reset, but in the present embodiment, for example, after the self-identification phase ends, the node information storage unit 11 and the data transfer path storage The predetermined data is stored in each of the unit 12 and the transfer speed storage unit 13.

Further, the bus manager becomes active 125 ms after the self-identification. Then, 625 ms elapses after the self-identification, and the isochronous resource manager becomes active. Then, for example, after this, if the own node is a bus manager or an isochronous resource manager when there is no bus manager, the maximum hop number is calculated and the gap count is changed. Also, a new isochronous resource is allocated 1000 ms after self-identification.

FIG. 6 is a flowchart showing the operation of this embodiment. First, a procedure for storing data related to ports of all nodes in the node information storage unit 11 will be described with reference to FIG. When a bus reset occurs, each node performs self-identification. Then, a self-identification packet is output from each node onto the bus. The node information storage unit 11 inputs them.

Then, as usual, there is no data inconsistency, which node is the root node, which node is the isochronous resource management,
Whether or not the link layer of each node is active is checked by checking means (not shown) (step A).
1).

Subsequently, the node information storage unit 11 extracts another parent port / child port / unused data allocated to each port from each input self-identification packet. Then, based on the extracted data, one of “P”, “C”, “NC”, and “NE” is stored in the second to fourth columns of FIG.

Next, the data transfer speed is set to 1 in FIG.
Store in the column. Further, it is determined whether or not the node is a leaf node, and the determination result is stored in the 0th column of FIG. 2A (step A2). Thus, as shown in FIG. 2B, data on ports of all nodes and the like are stored. Next, the process proceeds to step A3.

FIG. 7 is a flowchart showing a procedure for storing data relating to the data transfer path in the data transfer path storage unit 12 performed in step A3 of FIG. As shown in FIG. 7, first, ph allocated to the root node
The y_ID “06” is assigned to row 0, column 1 ([0,
1]: Hereinafter, row numbers and column numbers are similarly described. )
It is stored unconditionally (step A301).

Thereafter, from the sixth row in FIG. 2B where data on the root node is stored, each of the ports p0 to p0
Data regarding p2 is extracted in the order of p2, p1, and p0, and a child port is searched. In the present embodiment, for example, when each of the ports p2 to p0 is sequentially extracted, when a child port can be specified, the process shifts to searching for data of a node having a parent port to which the child port is connected. I do.

Therefore, in the sixth line of FIG. 2, although p2 and p0 are child ports, at the time when p2 is extracted, the data of the table for which search has been completed is rewritten with data indicating that search has been completed (step A302). Here, the rewriting to the data indicating that the search has been completed means, for example,
"C", "P", for example, "D" meaning "searched"
It means to be.

In this way, when the existence of the child port can be confirmed (step A303), the search on the sixth line is terminated, and the node phy_ID
The processing shifts to the search for the fifth row in which information on the 05th magnetic disk 35 is stored (step A304). On the other hand, when the existence of the child port cannot be confirmed in the root node (step A303), the operation shown in FIG. 7 is ended.

In the fifth row, first, the data stored in the second to fourth columns is extracted, and the presence or absence of the parent port is searched. Then, similarly to the above, the data in the table for which the search has been completed is rewritten to data indicating that the search has been completed (step A305). If there is a parent port (step A306), phy_ID05 of the magnetic disk 35 is stored in FIG. 3 [0, 2] (step A307).

Here, since p0 is the parent port, "5" is stored in [0, 2] in FIG. 3, and p0 is rewritten from "P" to "D" (step A305). Next, whether or not the magnetic disk 35 is a leaf node is confirmed from the data stored in [5, 0] in FIG. 2 (step A308).

Here, since it is not a leaf node, p
1 or p2 is a child port.
Therefore, a search is performed to determine which or both of these ports are child ports in the order of p2 and p1, and the data is rewritten with data indicating that the search has been completed (step A309).

As a result, if the existence of the child port can be confirmed (step A310), the process returns to step A304, and the procedure up to step A310 is repeated. on the other hand,
If the existence of the child port cannot be confirmed (step A310), the process shown in FIG. 7 ends. here,
Since p1 is a child port, the search on the fifth line ends,
It returns to step A304.

Subsequently, the child port p of the magnetic disk 35
The processing shifts to the search of the fourth row in which data related to the scanner 34, which is a node having a parent port connected to No. 1, is stored (step A304). In the fourth line, a search is performed in the same manner as in the fifth line.

That is, first, the presence or absence of the parent port is searched (step A305), and if it exists (step A306), the fact is stored in FIG. 3 (step A3).
07), search which port it is. Then, it is determined whether or not the node is a leaf node (step A308). If the node is not a leaf node, a child port is searched (step A309).

As a result of these searches, it is first confirmed that there is a parent port and that the parent port is p2, so that [4,] which is the phy_ID number of the scanner 34 is displayed in [0, 3] of FIG. Stored (Steps A305 to A3
07). Next, since it can be confirmed that p1 is a child port, the table is rewritten to "D" (step A3).
09), the search of the scanner 34 ends. And ph
The process proceeds to the search for the printer 33 with the y_ID 03 (steps A310 and A304).

In the third row, retrieval is performed in the same procedure as in the fifth row. As a result, it is confirmed that the parent port exists (steps A305 and A306), and [0,
Phy_ID number “3” is stored in [4] (step A307). Subsequently, whether or not the node is a leaf node is searched (step A308). Here, since the node is a leaf node, there is no child port. Therefore, the search on the third line is ended here.

In this manner, data relating to one data transfer path from the personal computer 36, which is the root node, to the printer 33, which is a leaf node, via the magnetic disk 35, which is a branch node, and the scanner 34, is transferred to the first line in FIG. To be stored.

Here, in the present embodiment, p
The child port is searched in the order of 2, p1, and p0, and when the child port is confirmed, the procedure shifts to the search for the next row. Therefore, in this procedure, some ports have not been searched. That is, for example, in the fourth line, it is confirmed that p2 is a parent port, and then it is confirmed that p1 is a child port. Has not been searched for p0.

If a port that has not been searched is a child port, there is a data transfer path that exists through the child port, so that it is necessary to search all the ports in the end. For this reason, in the present embodiment, an unsearched port is searched by a procedure that returns from the printer 33, which is a leaf node, to the personal computer 36, which is a root node.

Since "D" is stored in the table which has been searched in the order from the root node to the leaf node, when the search is performed in the order from the leaf node to the root node, a re-search is performed. It is possible to search efficiently without a table.

Next, a procedure for searching a table that has not been searched will be described. As described above, since the search of the third row in which the leaf node information is stored has been completed,
The tables that have not been searched earlier in the fourth row are p2, p1,
The search is performed in the order of p0 (steps A311 to A314).
Here, since the existence of a new child port is not confirmed,
Subsequently, a search for the fifth line is performed (step A311).

Since no new child port is confirmed by the same search in the fifth line, the search in the sixth line is performed (steps A311 and A312). Here, since p0 is a child port (step A313), a new child port can be confirmed by this search. Therefore, since there is a data transfer path including the child port, the data transfer path including the child port is specified and stored in FIG.

For this reason, first, a search is performed on the second line in which the data of the stereo interface 32 of phy_ID number 02 for which the parent port has not been searched is stored.
This search is performed in the same manner as the above-described search for the node from the root node to the leaf node. Therefore, in the second line, first, the presence / absence of the parent port is searched, and if the parent port exists, it is rewritten to the data indicating the search (step A305).

In this case, since p2 is the parent port, [2, 4] in FIG. 2 is rewritten to "D" (step A30).
5, A306). Then, phy_ID02 of the stereo interface 32 is stored in [1, 2] of FIG. 3 (step A307). Subsequently, it is confirmed whether or not the node is a leaf node from the data stored in [2, 0] in FIG. 2 (step A308).

Here, since it is not a leaf node, a search is made to determine which port the child port is. Furthermore, p
Since 1 is a child port, [2, 3] in FIG. 2 is rewritten from “C” to “D” (step A309), the search of the second row is completed, and the search moves to the search of the first row. (Step A
304).

The first line is searched according to the same procedure as the second line. Here, p0 is the parent port (step A
305, A306), and the phy_ID number 01 of the digital camera 31 is stored in [1, 3] of FIG. 3 (step A307). Further, since the node is a leaf node (step A308), there is no child port.
Therefore, the search on the first line is ended here.

The state of the bus connected from the personal computer 36 as the root node to the digital camera 31 as the leaf node via the stereo interface 32 as the branch node is shown in FIG.
Store in the line.

Subsequently, the printer 3 which is a leaf node
After confirming the bus state up to the leaf node, a search is again performed for the unsearched child port toward the root node, as in the procedure performed after the search up to 3 is completed. First, an unsearched port on the second line is checked (step A31).
1, A312). Here, since p0 is a child port (step A313), a data transfer path connected to a leaf node via p0 is stored in the third row of FIG.

Here, since this data transfer path is common from the personal computer 36 to the stereo interface 32, [2, 1] and [2, 2] in FIG. ], [1, 2] are stored (step A315).

Next, the CD-RO with the phy_ID number 00 on the 0th line in FIG. 2 for which the parent node has not been searched yet.
Retrieve data on M disk 30 (step A
304). This search is performed in the same procedure as the search from the root node to the leaf nodes.

Here, since p1 is the parent port (steps A305 and A306), this search results in [2, 3] of FIG.
The D number "0" is stored (step A307).
Then, “P” of [0, 3] in FIG. 2 is rewritten to “D”.

Further, since the node is a leaf node (step A308), a search is made toward the root node for the presence or absence of an unsearched child port (steps A311 to A311).
315). In this search, since there is no child port among the ports that have not been searched (step A314), there is no other data transfer path. In this way, three data transfer paths are specified as shown in FIG. 3B by the procedure of FIG.

Next, for example, how long the data output from the root node is input to the transfer destination node, in other words, the transmission speed of the data output from the root node is 100 Mbps, 200 Mbps, 400 Mbps.
Mbps will be described.

First, a method for optimizing the transmission rate of data output from the root node to each node will be described. First, the data transfer destination phy_I
D is, for example, “00”. Here, the data transfer path including the node with the phy_ID number 00 is the route in the second row in FIG. 3B, and only the node with the phy_ID number 02 exists between this node and the root node. .

Therefore, the data transfer rate of the node having the phy_ID number 02 is detected from FIG. 2B. Thus, the phy_ID numbers are “06”, “02”, “0”.
0 ”are 400 Mbps and 200 Mbps, respectively.
Mbps and 400 Mbps can be confirmed.

Also, from this result, phy_ID number 0
The optimal speed of the data output to node 0 is 200
Mbps. The optimum data transfer speed thus obtained is stored in FIG. Further, the data transfer rates of the phy_ID numbers “06” and “02” are 400M each.
bps and 200 Mbps, 200 Mbps is optimal for data to be output for phy_ID number 02. This data is stored in FIG.

Subsequently, the phy_ID from the root node
The transmission speed of the data to be output to the number 01 is determined by the same method. That is, here, phy_I
The data transfer speed of the node with the D number 01 is 100 Mb
Since it is ps, it must be 100 mbps regardless of the data transfer speed of the other nodes. Then, this data is stored in FIG.

Next, similarly, regarding the transmission speed of data to be output to the phy_ID number 03,
The data transfer rates of the numbers “06”, “05”, “04”, and “03” are extracted from the second column in FIG. Here, since each of these data transfer rates is 400 Mbps,
The optimal speed of the data output to phy_ID number 03 is 400 Mbps.

Further, phy_ID which is a leaf node
The optimum speed of data output to the node of number 03 is 40
0 Mbps means that the phy_ID number “0
The optimum speed of the data output to “6”, “05”, and “04” is also 400 Mbps, and is stored in FIG.

FIG. 8 is a flowchart showing a procedure for storing data in the transfer rate storage unit 13 by a method different from the above. Here, a procedure for calculating the transfer speed in the order of the phy_ID numbers and storing it in the transfer speed storage unit 13 will be described.

First, a data transfer path including a data transfer source is extracted from FIG. 2B (step A401). Here, for example, assuming that the root node is the data transmission source, all rows correspond. Next, the data transfer destination is, for example, phy_ID00 (step A40).
2). Then, it is determined whether the phy_ID number of the data transfer source node is different from the phy_ID number of the data transfer destination node (step A403).

Here, phy of the data transfer source node
_ID number is "00", phy of data transfer destination node
This is different because the _ID number is “06”. Therefore, the process proceeds to step A404. If they match, the process proceeds to step A405.

In step A404, a data transfer path including the data transfer destination node is searched from FIG. 2B. Here, it is the second line. Subsequently, a node on the data transfer path is searched from the second line in FIG. Here, only the phy_ID02 node exists. Next, the data transfer rate of the node on this data transfer path is extracted from the first column in FIG. 2B (step A405).

Here, phy_ID number “06”,
Since the data transfer speeds of the nodes “02” and “00” are 400 Mbps, 200 Mbps, and 400 Mbps, respectively, the optimum speed for data transfer to the node with the phy_ID number 00 is 200 Mbps. The optimum data transfer rate thus obtained is stored in FIG. 4 (step A405), and the process proceeds to step A405.

Subsequently, in step A405, the node for searching the data transfer destination is changed. Here, for example, a node whose phy_ID number is one larger is set as a search target (step A405). That is, the node that searches for the next data transfer destination is the node with the phy_ID number 01.

Next, in step A405, the phy_ID number of the data transfer source and the phy_ID of the data transfer destination
If the phy_ID number of the data transfer destination is smaller than the number, the process returns to step A404. Then, similarly, steps up to step A405 are repeated. On the other hand, if the phy_ID number of the data transfer destination is larger in step A405, the procedure shown in FIG. 8 ends. Thus, the data is stored in the transfer speed storage unit 13.

When data is stored in the transfer speed storage unit 13 by a method as shown in FIG. 8 or the like, as shown in FIG.
As usual, the isochronous resource acquired before the initialization of the bus is acquired again (step A5). In the procedure from step A5 to step A10, a process for isochronous resource management will be described as an example.

When 625 ms has elapsed from the time when the bus was initialized and there is no bus manager on the local bus, a cycle master determination process (step A9) and a Link-On packet transmission process (step A9) A10), and then confirm the maximum number of hops (step A11).

In step A11, the number of hops between leaf nodes is confirmed using the data stored in the table of FIG. 3 to determine the maximum number of hops. For example, as shown in FIG. 3, in this embodiment, since there are three data transfer paths, there are three leaf nodes. Therefore, first, for example, the number of hops from the leaf node on line 0 to the leaf node on line 1 is searched.

Here, the phy_ID numbers "3" and "4", "4" and "5", "5" and "6"
, And between "6" and "2" and between "2" and "1", the number of hops is five. Similarly, the number of hops from the leaf node in the first row to the leaf node in the second row is searched, and then the number of hops from the leaf node in the second row to the leaf node in the zeroth row is searched. Here, these hop numbers are 4 and 5, respectively. Therefore, the maximum hop number in this embodiment is 5
Becomes

FIG. 9 is a flowchart generalizing the procedure in step A11. In FIG. 9, the numbers of rows in FIG. 3 are m and n, the number of hops from the leaf node on the mth row to the leaf node on the nth row is Hop, and the maximum number of Hops is MaxHop.

First, n and MaxHop are set to 0, and m is set to n + 1 (steps A1101 to A1103). Next, the number of nodes existing between the leaf node on the nth row and the leaf node on the mth row is searched (step A1).
104). Since the number of hops can be represented by “the number of nodes−1”, the number of hops thus obtained is stored as Hop (step A1105).

Next, Hop is compared with MaxHop (step A1106). If MaxHop is smaller, it is updated to Hop (step A110).
7). Next, 1 is added to m (step A1108). Thereafter, m is compared with the number of data transfer paths (step A1109). If m is larger, step A
Returning to step 1104, the steps up to step A1109 are repeated. On the other hand, if m is smaller, step A11
Move to 10.

In step A1110, 1 is added to n. Then, in step A1111, n is compared with “the number of data transfer paths−1”, and if n is smaller,
Returning to step A1103, the steps up to step A1110 are repeated. On the other hand, when n is larger, FIG.
Are completed.

Next, the gap time is optimized (step A12). As described above, the gap count table 14 stores a gap count value that changes depending on the number of hops provided in the IEEE 1394 standard.

In step A11, the gap count table 1 is set using the maximum hop number obtained as an index.
4 (FIG. 5), the current gap count is accessed by detecting the optimum gap count value for the topology and accessing the PHY register of the root node;
If this differs from the gap count optimal for the topology, the setting of the gap count value is changed.

Then, after the bus is initialized, 1000
After waiting for msec (step A13), a new isochronous resource is allocated (step A14) after this time has elapsed, and the processing after self-identification ends.

FIG. 10 is a diagram showing each phase and gap time. Here, the cable delay refers to a delay generated in a cable connecting nodes. A subaction gap is one that appears before an unconnected asynchronous subaction during an equal interval. arb_d
Elay refers to the time that a node must wait for all nodes to guarantee the observation of a gap.
The arbitration reset gap is the one that appears before the unconnected asynchronous subaction at the start of the equal interval.

FIG. 11 is a diagram showing values defined in the standard for the constants in the calculation formula of FIG. Here, the term “cable assembly” refers to a configuration of a cable, and particularly to a length here. The cable propagation speed indicates the performance (delay per unit meter) of the cable. ARB_SPEED_SIGNAL_ST
ART refers to a delay time from when a transmitting port generates a data_prefix signal to when the same transmitting port generates a transmission speed signal.

Also, DATA_PREFIX_TIME
Is the remaining time from the end of speed sampling to the start of data synchronized with the clock. Also, it refers to the time between recognition and response data in a connected response. MAX_DA
TA_PREFIX_DELAY means that R is
This is the maximum delay from when X_DATA_PREFIX appears to when TX_DATA_PREFIX is transmitted at the transmission port. ACK_RESPONSE_TIME is
After reporting the end of the packet (DATA_END),
It means the time until a request is made to recognize the response of the link layer using the arbitration request.

Further, the recognition gap (ack_gap) refers to the one appearing between the end of the main asynchronous packet and the recognition packet. acknowledg g
“aptime” refers to the time from the end of a certain packet to the start of recognition. MAX_PHY_DARA_DE
LAY represents the maximum delay from when a node receives data until the node transmits the same data.

FIG. 12 is a diagram showing an equal section in the asynchronous area according to the present embodiment and the related art. FIG. 14 is a diagram showing calculation results of each phase and gap time shown in FIG.

Here, the equal section N is approximately represented by the following equation from FIG.

Equality section N = (arbitration phase × 3) + (recognition phase × 3) + total of data transfer phases + (subaction gap × 2) + arbitration reset gap FIG. It is a figure showing time. The equal section N requires 69.28 μs in the related art, but can be reduced to 25.37 μs in the present embodiment. This is because, by optimizing the data transfer speed, optimizing the sub-action gap and the arbitration reset gap, unnecessary time among them is reduced.

That is, according to the present embodiment, for example, the arbitration phase takes about 3.58 μsec to about 1.5 μs
ec, and the sub-action gap is about 10.
The time is reduced from 5 μsec to about 2.25 μsec, and the arbitration reset gap is reduced from about 21 μsec to about 4.45 μs.
ec. This can reduce the time required for the equal section N to about 63% of the conventional one.

[0106]

As described above, the present invention obtains connection information of a plurality of electronic devices and transfer speed information for transferring data of each of the electronic devices, and specifies a data transfer path based on the connection information. And the data transfer speed is determined according to the transfer speed information. Therefore, by optimizing the data transfer speed, the data transfer time can be improved.

Further, the present invention detects the number of connection media between the electronic device at the end of the data transfer path and the electronic device at the end of another data transfer path, and calculates the maximum value of the number of connection media. , The gap count is changed accordingly. Therefore, the subaction gap and the arbitration reset gap can be optimized, and the data transfer time can be improved.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of an apparatus for optimizing a data communication speed according to an embodiment of the present invention.

FIG. 2 is a diagram illustrating an internal configuration of a node information storage unit in FIG. 1 and a state in which data is stored therein;

3 is a diagram showing an internal configuration of a data transfer path storage unit in FIG. 1 and a state in which a phy_ID number assigned to each node is stored; FIG.

FIG. 4 is a diagram showing an internal configuration of a transfer speed storage unit in FIG. 1 and a state in which a data transfer speed is stored therein.

FIG. 5 is an internal configuration diagram of a gap count list of FIG. 1;

FIG. 6 is a flowchart showing the operation of the embodiment of the present invention.

FIG. 7 is a flowchart showing a procedure in step A3 of FIG. 6;

FIG. 8 is a flowchart illustrating a procedure for storing data in a transfer speed storage unit.

FIG. 9 is a flowchart generalizing the procedure in step A11.

FIG. 10 is a diagram showing each phase and gap time.

11 is a diagram showing values defined by standards for constants in the calculation formula in FIG. 10;

FIG. 12 is a diagram showing an equal section in an asynchronous area according to the present embodiment and the related art.

FIG. 13 is a diagram showing each time required for the present embodiment and the related art.

FIG. 14 is a diagram showing a topology (tree connection) conforming to IEEE1394.

FIG. 15 is a time chart showing processing after a bus reset occurs.

FIG. 16 is a diagram showing a state in which an isochronous transfer area and an asynchronous transfer area are mixed on a bus.

[Explanation of symbols]

 REFERENCE SIGNS LIST 1 storage device 2 data processing device 11 node information storage unit 12 data transfer path storage unit 13 transfer speed storage unit 14 gap count list 21 data transfer path identification unit 22 transfer speed calculation unit 23 maximum hop count calculation unit 24 gap count change unit Reference Signs List 30 CD-ROM disk 31 Digital camera 32 Stereo interface 33 Printer 34 Scanner 35 Magnetic disk 36 Personal computer

 ──────────────────────────────────────────────────続 き Continued on the front page F term (reference) 5K032 AA01 CC06 DA06 DB19 DB22 DB24 5K034 AA01 BB01 BB06 CC02 DD03 EE10 FF12 HH01 HH02 HH06 HH25 KK01 LL04 LL05 QQ02 QQ04 SS01

Claims (5)

[Claims] 1. A data transfer system having a plurality of electronic devices, comprising: means for acquiring connection information of the plurality of electronic devices and transfer speed information for transferring data of each of the electronic devices, based on the connection information. To specify the data transfer path, and
A data transfer system, wherein a data transfer speed is determined according to the transfer speed information. 2. Detecting the number of connection media between an electronic device at an end of the data transfer path and an electronic device at an end of another data transfer path, calculating a maximum value of the number of connection media, 2. The data transfer system according to claim 1, wherein the gap count is changed to a gap count according to the maximum value. 3. The data transfer system according to claim 1, wherein the connection medium is a serial bus conforming to IEEE 1394. 4. The data according to claim 1, wherein the electronic device having the largest physical layer identification number is provided with a means for acquiring the connection information and the transfer speed information. Transfer system. 5. The data transfer system according to claim 1, wherein each of the plurality of electronic devices includes a unit that acquires the connection information and the transfer speed information.