JPH11284649A - Network system, and network management device and method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To inform an operator that there is no effect and data transfer efficiency is improved when devices connected to a network are divided. SOLUTION: A network managing device recognizes a route of data transferred through a network in a step S5, collates it against the configuration of a network in a step S6 to determine whether or not the network can be divided, and also whether or not the data transfer efficiency of the network is improved by the division in steps S7 and S8 when the network is divided; when it is determined that the data transfer efficiency is improved, a message recommending the division is sent to proper nodes on the network in a step S10.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for connecting a plurality of electronic devices (hereinafter referred to as "devices") using a data communication bus capable of communicating control signals and data in a mixed manner.
The present invention relates to a network system for performing data communication between devices and a network management apparatus and method for managing the system.

[0002]

2. Description of the Related Art Conventionally, IEEE1394-1995 High
In a case where data transfer is performed simultaneously by a plurality of devices in a plurality of devices connected by a performance serial bus (hereinafter, a 1394 serial bus), transfer between a certain set of devices (for example, a PC and a printer) Between other devices (for example, another PC and printer, DVCR, HD, etc.)
, Even if they use a completely different transmission path than
Without occupying each other's bus, fair data transfer was performed while arbitrating.

[0003]

In the data transfer using the 1394 serial bus as described above, if a large amount of data is transferred and the bus is congested, the data is transferred between a pair of devices performing data communication. Is independent of the data transmission path between a pair of other devices (for example, nodes AB and CG), each of them can arbitrate without occupying the bus and obtain fair data while arbitrating. Since the transfer is performed, the data transmission efficiency is reduced.

[0004] The present invention has been made in view of the above-mentioned conventional example. In the case where the network is divided, the data transfer efficiency increases, and when there is no obstacle in data transfer between devices on the network, Provided is a network system, a network management device, and a method that enable more efficient data transfer by notifying a user of the fact.

[0005]

To achieve the above object, a network management apparatus according to the present invention has the following configuration. That is, connection means for connecting to a network,
Recognize the path of data transferred on the network via the connection means, determine whether the network can be divided in light of the configuration of the network, and if it is determined that division is possible, It is provided with a network management means for judging whether the data transfer efficiency is improved, and outputting a recommendation message for recommending the division when the data transfer efficiency is judged to be improved.

In order to achieve the above object, a network management method according to the present invention has the following configuration. That is, the path of data transferred on the network is recognized, and it is determined whether the network can be divided according to the configuration of the network. If it is determined that the network can be divided, the data transfer efficiency on the network is improved by the division. If it is determined that the division is to be improved, a recommendation message recommending division is output.

In order to achieve the above object, a network system according to the present invention has the following configuration. That is, the above network management device is connected as one of the nodes.

In order to achieve the above object, a computer readable storage medium of the present invention has the following configuration.
That is, the path of data transferred on the network is recognized, and it is determined whether the network can be divided according to the configuration of the network. If it is determined that the network can be divided, the data transfer efficiency on the network is improved by the division. It is characterized in that a program for causing a computer to function as network management means for outputting a recommendation message for recommending division is stored when it is determined to improve.

[0009]

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

FIG. 1 shows an example of a network configuration when implementing the present invention.

In the present invention, since an IEEE 1394 serial bus is used as a digital I / F for connecting each device, the IEEE 1394 serial bus will be described in advance. << Overview of IEEE 1394 Technology >> Digital VT for Home Use
With the advent of R and DVD, it is necessary to support real-time and high-information-volume data transfer of video data and audio data. In order to transfer such video and audio data in real time, and to transfer it to a personal computer (PC) or other digital devices, an interface capable of high-speed data transfer with the necessary transfer functions is required. The interface developed from such a viewpoint is IEEE 1394-1995 High Performance Ser
ial Bus (hereinafter 1394 serial bus).

FIG. 7 shows an example of a network system configured using a 1394 serial bus. This system is provided with devices A, B, C, D, E, F, G, and H. A-B, A-C, B-D, D-E, C-F
, CG, and CH are connected by a twisted pair cable of a 1394 serial bus.
The devices A to H are, for example, PC, digital VTR, DV
D, digital camera, hard disk, monitor, etc.

The connection method between the devices is such that the daisy chain method and the node branch method can be mixed.
A highly flexible connection is possible.

Each device has a unique ID, and recognizes each other to form a single network within a range connected by a 1394 serial bus. One 1394 connection between each digital device
Just by sequentially connecting with a serial bus cable, each device plays a role of relay, and constitutes one network as a whole. In addition, it has a function of automatically recognizing the device and recognizing the connection status when the cable is connected to the device by the Plug & Play function, which is a feature of the 1394 serial bus.

In the system shown in FIG. 7, when a certain device is deleted from the network or newly added, the bus reset is automatically performed, and the network configuration up to that time is reset. From
Rebuild a new network. With this function, the configuration of the network at that time can be constantly set and recognized.

The data transfer rate is 100/200 /
It has a transmission rate of 400 Mbps, and a device having a higher transfer rate supports a lower transfer rate and is compatible.

As a data transfer mode, asynchronous data such as a control signal (Asynchronous data: A
Asynchronous transfer mode for transferring sync data)
There is an isochronous transfer mode for transferring synchronous data (Isochronous data: hereinafter, iso data) such as real-time video data and audio data. This Asyn
In each cycle (usually 125 μS per cycle), the c data and the Iso data follow the transfer of a cycle start packet (CSP) indicating the start of the cycle, followed by the Is data.
o Data is transferred together in a cycle while giving priority to data transfer.

Next, FIG. 8 shows the components of the 1394 serial bus.

The 1394 serial bus has a layer (layer) structure as a whole. As shown in FIG.
The most hardware type is a 1394 serial bus cable, which has a connector port to which a connector of the cable is connected.
There are layers and link layers.

The hardware part is a substantial part of an interface chip, of which the physical layer performs coding and control related to connectors, and the link layer performs packet transfer and control of cycle time.

The transaction layer of the firmware section manages data to be transferred (transacted), and issues commands such as Read and Write.
The serial bus management is a part that manages the connection status and ID of each connected device and manages the configuration of the network.

The hardware and firmware are the actual configuration of the 1394 serial bus.

The software application
The layer differs depending on the software used, and is a part that defines how data is placed on the interface.
It is specified by a protocol such as the V protocol.

The above is the configuration of the 1394 serial bus.

Next, FIG. 9 shows a diagram of the address space in the 1394 serial bus.

Each device (node) connected to the 1394 serial bus always has a 64-bit address unique to each node. By storing this address in the ROM, it is possible to always recognize the node address of oneself and the other party, and perform communication specifying the other party.

The addressing of the 1394 serial bus is based on the IEEE 1212 standard, and the first 10 bits are used for specifying the bus number.
The next 6 bits are used for specifying the node ID number. The remaining 48 bits become the address width given to the device,
Each can be used as a unique address space. The last 28 bits store information such as identification of each device and designation of use conditions as an area of unique data.

The above is the outline of the technology of the 1394 serial bus.

Next, the technical portion which can be said to be a feature of the 1394 serial bus will be described in more detail. << Electrical Specifications of 1394 Serial Bus >> FIG.
4 shows a sectional view of a serial bus cable.

In the 1394 serial bus, a power supply line is provided in a connection cable in addition to two twisted pair signal lines. As a result, power can be supplied to a device having no power supply, a device whose voltage has dropped due to a failure, and the like.

The voltage of the power supply flowing in the power supply line is 8 to 40.
V and the current are specified as the maximum current DC 1.5A. << DS-Link encoding >> DS-Link of data transfer format adopted in 1394 serial bus
FIG. 11 is a diagram for explaining the encoding method.

In the 1394 serial bus, DS-Lin
The k (Data / Strobe Link) coding method is adopted. This DS-Link coding scheme is suitable for high-speed serial data communication, and its configuration requires two signal lines. The main data is sent to one of the twisted pairs, and a strobe signal is sent to the other twisted pair.

On the receiving side, the clock can be reproduced by taking the exclusive OR of this communicated data and the strobe.

Advantages of using the DS-Link coding method include higher transfer efficiency as compared with other serial data transfer methods, and the circuit scale of the controller LSI can be reduced because a PLL circuit is not required.
Since there is no need to send information indicating the idle state when there is no data to be transferred, the power consumption can be reduced by setting the transceiver circuit of each device to the sleep state. << Bus Reset Sequence >> In the 1394 serial bus, each connected device (node) is given a node ID and recognized as a network configuration.

When there is a change in the network configuration, for example, a change occurs due to an increase or decrease in the number of nodes due to insertion / removal of a node or power ON / OFF, etc., and it is necessary to recognize a new network configuration. Each detected node sends a bus reset signal on the path,
Enter the mode to recognize the new network configuration. The method of detecting the change at this time is performed by detecting a change in the bias voltage on the 1394 port board.

When a bus reset signal is transmitted from a certain node, the physical layer of each node transmits the bus reset signal to the link layer at the same time as receiving the bus reset signal, and transmits the bus reset signal to another node. . After all the nodes finally detect the bus reset signal, the bus reset is activated.

The bus reset is also activated by the above-described activation by hardware insertion / removal due to cable disconnection or network abnormality or the like, and also by directly issuing a command to the physical layer by host control from a protocol.

Further, when the bus reset is activated, the data transfer is suspended, the data transfer during this period is waited, and after the end, the data transfer is resumed under a new network configuration.

The above is the bus reset sequence.
<< Node ID Determination Sequence >> After the bus reset, each node starts an operation of giving an ID to each node in order to construct a new network configuration. The general sequence from the bus reset to the determination of the node ID at this time will be described with reference to the flowcharts of FIGS. 19, 20, and 21.

The flowchart of FIG. 19 shows a series of bus operations from the occurrence of a bus reset to the determination of the node ID and the start of data transfer.

First, as step S101, the occurrence of a bus reset in the network is constantly monitored, and if a bus reset occurs due to power ON / OFF of a node, the process proceeds to step S102.

In step S102, from the reset state of the network, a parent-child relationship is declared between the directly connected nodes in order to know the connection status of the new network. As step S103,
When the parent-child relationship is determined between all nodes, step S
One route is determined as 104. Until the parent-child relationship is determined between all nodes, the parent-child relationship is declared in step S102, and the route is not determined.

After the route is determined in step S104, the operation of setting a node ID for giving an ID to each node is performed in step S105. Node IDs are set in a predetermined node order, and I
When the setting operation is repeatedly performed until D is given, and finally the IDs are set in all the nodes in step S106, the new network configuration is recognized in all the nodes. Then, the data transfer is started.

In the state of step S107, a mode for monitoring the occurrence of a bus reset again is entered.
When the bus reset occurs, the setting operation from step S101 to step S106 is repeatedly performed.

The above is the description of the flowchart of FIG. 19. The flowchart from FIG. 19 shows the part from the bus reset to the route determination and the procedure from the route determination to the end of the ID setting in a more detailed flowchart. Are shown in FIGS. 20 and 21, respectively.

First, the flowchart of FIG. 20 will be described.

When a bus reset occurs in step S201, the network configuration is reset once. The occurrence of a bus reset is constantly monitored in step S201.

Next, as step S202, as a first step of re-recognizing the reset network connection status, a flag indicating a leaf (node) is set for each device. Further, in step S203, each device checks how many ports it has are connected to other nodes.

According to the result of the number of ports in step S204, the number of undefined (undetermined parent-child) ports is checked in order to start the declaration of the parent-child relationship.
Immediately after the bus reset, the number of ports = the number of undefined ports. However, as the parent-child relationship is determined, the number of undefined ports detected in step S204 changes.

First, immediately after the bus reset, only the leaf can declare the parent-child relationship first. A leaf can be known by checking the number of ports in step S203. In step S205, the leaf declares "I am a child and the other is a parent" to the node connected thereto, and ends the operation.

A node which has a plurality of ports in step S203 and is recognized as a branch has a number of undefined ports> 1 in step S204 immediately after the bus reset.
The process proceeds to step S206, where a flag of branch is set, and in step S207, the process waits for reception of “parent” in the parent-child relationship declaration from the leaf.

The leaf declares a parent-child relationship, and the branch that has received the declaration in step S207 appropriately returns to step S20.
Confirm the number of undefined ports of 4 and find that the number of undefined ports is 1
If it becomes, it becomes possible to declare “I am a child” in step S205 for the node connected to the remaining port. After the second time, step S204
Even if the number of undefined ports is checked in step S207, for a branch having two or more ports, the process waits again in step S207 to accept a "parent" from a leaf or another branch.

Finally, if any one of the branches or exceptionally leaves (because the child declaration could not be performed quickly enough) could become zero as a result of the number of undefined ports in step S204, In this case, the declaration of the parent-child relationship of the entire network has been completed, and the only node for which the number of undefined ports has become zero (all are determined as parent ports) is flagged as a root in step S208, and the root is set in step S209. The recognition is made.

Thus, the process from the bus reset shown in FIG. 20 to the declaration of the parent-child relationship between all the nodes in the network is completed.

Next, the flowchart of FIG. 21 will be described.

First, in the sequence up to FIG.
Since the information of the flag of each node such as branch and route is set, classification is performed in step S301 based on this.

As an operation of assigning an ID to each node, the ID can be set first from the leaf. The IDs are set in ascending order of leaf → branch → route (node number = 0).

In step S302, the number N (N is a natural number) of leaves existing in the network is set. Thereafter, in step S303, each leaf requests the root to give an ID. If there are a plurality of such requests, the route performs arbitration (operation of arbitration into one) in step S304, and performs step S3.
As 05, an ID number is given to one winning node, and a failure result is notified to the losing node. In step S306, the leaf whose ID acquisition has failed fails issues an ID request again, and repeats the same operation. In step S307, the ID information of the node is transferred to all the nodes by broadcasting from the leaf whose ID has been acquired. One node I
When the broadcasting of the D information is completed, step S30
As 8, the number of remaining leaves is reduced by one. here,
In step S309, when the number of the remaining leaves is one or more, the operation from the ID request in step S303 is repeatedly performed, and finally, when all the leaves broadcast the ID information, N = 0 in step S309, and the next Moves to the branch ID setting.

The setting of the branch ID is performed in the same manner as in the case of the leaf.

First, as step S310, the number M (M is a natural number) of branches existing in the network is set. Thereafter, in step S311, each branch requests the root to give an ID. On the other hand, for the route, arbitration is performed in step S312, and the branch is given in order from the winning branch to the next youngest number given to the leaf. In step S313, the root notifies the branch that issued the request of ID information or a failure result, and in step S314, the branch whose ID acquisition has failed fails issues an ID request again and repeats the same operation. In step S315, the ID information of the node is broadcast and transferred to all the nodes from the branch where the ID has been obtained. When the broadcast of the one node ID information ends, the number of remaining branches is reduced by one in step S316. Here, step S3
When the number of the remaining branches is 1 or more, the operation from the ID request in step S311 is repeated until all branches finally broadcast ID information. When all the branches have acquired the node IDs, M = 0 in step S317, and the branch ID acquisition mode ends.

At this point, since only the root node has not acquired the ID information at the end, step S
318 is set as the own ID number among the unassigned numbers, and the root I
Broadcast D information.

Thus, as shown in FIG. 21, the procedure from the determination of the parent-child relationship to the setting of the IDs of all the nodes is completed.

Next, the operation in the actual network shown in FIG. 12 will be described as an example with reference to FIG.

Referring to FIG. 12, (root) node B
Are directly connected to node A and node C,
Further, a node D is directly connected below the node C, and a node E and a node F are directly connected below the node D in a hierarchical structure. The procedure for determining the hierarchical structure, the root node, and the node ID will be described below.

After the bus reset, a parent-child relationship is declared between the directly connected ports of each node in order to recognize the connection status of each node. The parent and child can be said to be such that the parent is higher in the hierarchical structure and the child is lower.

In FIG. 12, the node A first declares the parent-child relationship after the bus reset. Basically, a node (called a leaf) having a connection to only one port of the node can declare a parent-child relationship. Since the user can first know that only one port is connected, it recognizes that this is the edge of the network, and the parent-child relationship is determined from the node that operates earlier in the network. In this manner, the port on the side that has declared the parent-child relationship (node A between AB) is set as a child, and the port on the other side (node B) is set as a parent. Thus, between node AB, child-parent, node E-
The child-parent is determined between D and the child-parent is determined between the nodes FD.

Further up in the hierarchy, among nodes having a plurality of connection ports (referred to as branches), the parent-child relationship is declared further higher in order from the node that received the declaration of the parent-child relationship from another node. To go. In FIG. 12, after the parent-child relationship between the node D and DE and between DF is determined,
The parent-child relationship is declared for node C, and as a result, child-parent is determined between nodes D and C.

The node C that has received the parent-child relationship declaration from the node D becomes the node B connected to another port.
Declares a parent-child relationship. As a result, a child-parent is determined between the nodes C and B.

In this way, a hierarchical structure as shown in FIG. 12 is formed, and the node B that has become the parent in all finally connected ports is determined as the root node. There is only one route in one network configuration.

In FIG. 12, node B is determined to be the root node. Node B, which has received a parent-child relationship declaration from node A, makes a parent-child relationship declaration to other nodes at an early timing. If so, the root node may have moved to another node. That is, any node may become a root node depending on the transmission timing, and the root node is not always constant even in the same network configuration.

When the root node is determined, the process enters a mode for determining each node ID. Here, all nodes notify their determined node IDs to all other nodes (broadcast function).

The self ID information includes its own node number, information on the connected position, the number of ports it has, the number of connected ports, information on the parent-child relationship of each port, and the like.

As a procedure for assigning node ID numbers, it is possible to start from a node (leaf) having connection to only one port, and to assign node numbers = 0, 1, 2, and so on in this order. .

The node having the node ID broadcasts information including the node number to each node. As a result, it is recognized that the ID number is “assigned”.

When all the leaves have acquired their own node IDs, the next step is to move to a branch and the node ID number following the leaf is assigned to each node. Similarly to the leaf, the node ID information is broadcast sequentially from the branch to which the node ID number is assigned, and finally, the root node broadcasts its own ID information. That is, the root always owns the maximum node ID number.

As described above, the assignment of the node IDs of the entire hierarchical structure is completed, the network configuration is reconstructed, and the bus initialization operation is completed. << Arbitration >> In the 1394 serial bus, arbitration (arbitration) of the right to use the bus is always performed prior to data transfer. Since the 1394 serial bus is a logical bus-type network in which each device connected individually relays the transferred signal to transmit the same signal to all devices in the network, packet collision occurs. Arbitration is necessary to prevent This allows only one node to transfer at a given time.

As a diagram for explaining arbitration, FIG. 13A shows a bus use request diagram, and FIG. 13B shows a bus use permission diagram.

When arbitration starts, one or more nodes issue a bus use request to the parent node. Nodes C and F in FIG. 13A are nodes that have issued a bus use right request. The parent node (node A in FIG. 13) which has received the request further issues (relays) a bus use request toward the parent node. This request is finally delivered to the arbitration route.

The root node having received the bus use request determines which node uses the bus. This arbitration work can be performed only by the root node, and the node that has won the arbitration is given permission to use the bus. FIG.
In (b), use permission is given to the node C, and use of the node F is rejected. A DP (data prefix) packet is sent to the node that lost the arbitration to notify that the node has been rejected. The rejected node use request waits until the next arbitration.

As described above, the node that has won the arbitration and has obtained the bus use permission can start transferring data thereafter.

Here, a series of arbitration flows will be described with reference to a flowchart shown in FIG.

In order for a node to be able to start data transfer,
The bus must be idle. In order to recognize that the data transfer that has been performed earlier is completed and that the bus is currently idle, a predetermined idle time gap length (eg, sub-action. Each node determines that its own transfer can be started by passing the gap.

In step S401, it is determined whether a predetermined gap length corresponding to each data to be transferred, such as Async data and Iso data, has been obtained. Unless the predetermined gap length is obtained, the request for the right to use the bus required to start the transfer cannot be made, so the process waits until the predetermined gap length is obtained.

If a predetermined gap length is obtained in step S401, it is determined in step S402 whether there is data to be transferred. If so, a request for a bus use right is issued in step S403 to secure a bus for transfer. Emit to the route. At this time, the transmission of the signal indicating the request for the right to use the bus is finally delivered to the route while relaying each device in the network, as shown in FIG. If there is no data to be transferred in step S402,
Wait as it is.

Next, at step S404, when the route receives one or more bus use requests at step S403, the route checks at step S405 the number of nodes that have issued use requests. If the selection value in step S405 is the number of nodes = 1 (the number of nodes that issued the use right request is one), the immediately subsequent bus use permission is given to that node. The selection value in step S405 is the number of nodes> 1
If (the number of nodes requesting the use is plural), the root performs an arbitration operation of deciding one node to which use permission is given in step S406. This arbitration work is fair, and the same node does not always obtain permission each time, and the right is equally given.

As step S407, step S40
At step 6, the node arbitrates the route from among the plurality of nodes that have issued the use request and selects one node whose use has been granted and another node that has lost. Here, for one node that has been arbitrated and has obtained use permission, or a node that has obtained use permission without arbitration with the number of use request nodes = 1 from the selection value in step S405, the route is set to that node as step S408. A permission signal is sent to it. The node that has received the permission signal starts transferring data (packets) to be transferred immediately after receiving the permission signal. Further, the nodes that have been defeated in the arbitration in step S406 and have not been permitted to use the bus are given step S406.
At step 409, a DP (date prefix) packet indicating an arbitration failure is sent from the root, and the node receiving the packet returns to step S401 to issue a bus use request for performing the transfer again, until the predetermined gap length is obtained. stand by.

The above is the description of the flowchart of FIG. 6 for explaining the flow of arbitration. << Asynchronous transfer >> The asynchronous transfer is an asynchronous transfer. FIG. 14 shows a temporal transition state in the asynchronous transfer. The first sub-action gap in FIG. 14 indicates the idle state of the bus. When the idle time reaches a certain value, the node desiring transfer determines that the bus can be used and executes arbitration for acquiring the bus.

When the bus use permission is obtained by arbitration, the data transfer is executed in the form of a packet.
After the data transfer, the receiving node sets ack (reception confirmation return code) of the reception result for the transferred data to a.
After a short gap of ck gap, the transfer is completed by returning and responding or sending a response packet. The ack is composed of 4-bit information and a 4-bit checksum, and includes information such as success, busy status, and pending status, and is immediately returned to the source node.

Next, FIG. 15 shows an example of the packet format of the asynchronous transfer.

The packet has a header in addition to the data part and the data CRC for error correction, and the header part has the destination node ID, the source node ID, the transfer data length and the like as shown in FIG. Various codes and the like are written and transferred.

Asynchronous transfer is one-to-one communication from a self-node to a partner node. The packet transferred from the transfer source node is distributed to each node in the network, but the address other than the address for itself is ignored, so that only one destination node reads the packet.

The above is the description of the asynchronous transfer. << Isochronous (Synchronous) Transfer >> Isochronous transfer is synchronous transfer. This isochronous transfer, which can be said to be the greatest feature of the 1394 serial bus, is
This transfer mode is suitable for transferring data that requires real-time transfer, such as multimedia data such as O video data and audio data.

Also, when the asynchronous transfer (asynchronous) is 1
Unlike the one-to-one transfer, the isochronous transfer is uniformly transferred from one transfer source node to all other nodes by the broadcast function.

FIG. 16 is a diagram showing a temporal transition state in the isochronous transfer.

The isochronous transfer is executed on the bus at regular intervals. This time interval is called an isochronous cycle. The isochronous cycle time is 125 μS. A cycle start packet indicates the start time of each cycle, and plays a role of adjusting the time of each node. A node called a cycle master transmits a cycle start packet, and after a transfer in a previous cycle is completed, a predetermined idle period (subaction gap) is passed, and then the start of this cycle is announced. Send a cycle start packet. The time interval at which this cycle start packet is transmitted is 125 μS.

FIG. 16 shows channels A, B,
As indicated by the channel C, a plurality of types of packets can be separately transferred by being given channel IDs in one cycle. This allows real-time transfer between a plurality of nodes at the same time, and the receiving node fetches only the data of the channel ID desired by itself. The channel ID does not represent the address of the transmission destination, but merely gives a logical number for the data. Therefore, the transmission of a certain packet is 1
From one source node to all other nodes,
It will be transferred by broadcast.

Prior to packet transmission in isochronous transfer, arbitration is performed as in asynchronous transfer. However, since the communication is not one-to-one communication as in the asynchronous transfer, there is no ack (reception confirmation reply code) in the isochronous transfer.

The iso gap (isochronous gap) shown in FIG. 16 indicates an idle period necessary for recognizing that the bus is empty before performing the isochronous transfer. After the predetermined idle period has elapsed, a node that wishes to perform isochronous transfer determines that the bus is free, and can perform arbitration before transfer.

Next, an example of a packet format for isochronous transfer will be described with reference to FIG.

Each packet divided into each channel has a header in addition to a data part and data CRC for error correction, and the header part has a transfer data length and a channel as shown in FIG. NO and other various codes and a header CRC for error correction are written and transferred.

The above is the description of the isochronous transfer. << Bus Cycle >> In actual transfer on the 1394 serial bus, isochronous transfer and asynchronous transfer can coexist. FIG. 18 shows a temporal transition of the transfer state on the bus in which the isochronous transfer and the asynchronous transfer are mixed at that time.

The isochronous transfer is executed prior to the asynchronous transfer. The reason is that after the cycle start packet, the isochronous transfer can be started with a gap length (isochronous gap) shorter than the gap length (subaction gap) of the idle period required to start the asynchronous transfer. . Therefore, the isochronous transfer is executed with priority over the asynchronous transfer.

In the general bus cycle shown in FIG. 18, a cycle start packet is transferred from the cycle master to each node at the start of cycle #m. As a result, each node adjusts the time, and after waiting for a predetermined idle period (isochronous gap), the node that should perform isochronous transfer performs arbitration and starts packet transfer. In FIG. 18, the channel e, the channel s, and the channel k are sequentially and isochronously transferred.

After the operations from the arbitration to the packet transfer are repeatedly performed for the given channel, when all the isochronous transfers in the cycle #m have been completed, the asynchronous transfer can be performed.

When the idle time reaches the sub-action gap in which asynchronous transfer is possible, the node that wishes to perform asynchronous transfer determines that it can shift to execution of arbitration.

However, during the period in which the asynchronous transfer can be performed, a subaction gap for starting the asynchronous transfer is obtained between the end of the isochronous transfer and the time (cycle synch) at which the next cycle start packet should be transferred. Only when they are given.

In cycle #m in FIG. 18, isochronous transfer for three channels and asynchronous transfer (including two packets (pack 1 and packet 2))
Has been transferred. After this asynchronous packet 2, the time to start cycle m + 1 (cycle sync
h), the transfer in cycle #m ends here.

However, the time (cyc) for transmitting the next cycle start packet during asynchronous or synchronous transfer operation
If le synch) is reached, the cycle start packet of the next cycle is transmitted after waiting for an idle period after the transfer is completed, without forcibly interrupting the transfer. That is, when one cycle continues for 125 μS or more, it is assumed that the next cycle is correspondingly shorter than the reference 125 μS. Thus, the isochronous cycle is 125 μ
S can be exceeded or shortened based on S.

However, the isochronous transfer is always executed if necessary every cycle in order to maintain the real-time transfer, and the asynchronous transfer may be transferred to the next and subsequent cycles due to the shortened cycle time.

The cycle information including such delay information is
Controlled by the master.

The above is the description of the IEEE 1394 serial bus.

<Data Transmission Efficiency in Network>
FIG. 1 is a diagram showing a network of devices according to an embodiment of the present invention. Each device is connected using a 1394 serial bus. PC shown in FIG. 1 is a personal computer, DVCR is a digital video camera recorder, HD
Indicates a hard disk. In FIG. 1, node B (PC)
To the node A (printer). Further, data transfer (printing) from the node C (PC) or the node E (PC) to the node D (printer) is performed, and an image captured from the node F (DVCR) is transferred to the node C (PC) or the node E (PC). ), Stored in the node G (HD), and printed by the node D (printer).

In a network connected by a 1394 serial bus, a plurality of nodes do not transmit data at the same time, and a node that has won arbitration transmits data. Arbitration is performed so that transmission opportunities are imparted to each node fairly, and each node performs fair data transfer. However, for this,
When the amount of data handled by each node increases, a plurality of nodes attempt to transfer data at the same time, so that the bus becomes very congested, and it becomes difficult to transfer data efficiently.

For example, the node B (PC) in FIG. 1 transfers data (printing) effective only to the node A (printer).
And does not transfer valid data (such as print data) to the other nodes C to G. Conversely, valid data (such as performing printing or the like) is transmitted from nodes C to G to nodes A and B. ,
Video display, etc.) are not being transferred. In this case, for example, when the node A (PC) transmits data to the node B (printer), the bus is used by arbitration even though other nodes do not transmit data to the nodes A and B. You have to gain the right and then send the data. In addition,
Valid data means data that is valid for the receiving side. On the other hand, data that reaches a node flowing on the bus but is not the destination of the node is called invalid data for the node and is distinguished. This means that in IEEE1394, data transmitted from a certain node is sent to all other nodes, and in the asynchronous mode, only the destination node described in the data packet receives the data packet,
Also, in the isochronous mode, only the destination node receives data of the channel previously notified from the transmitting node. That is, in the asynchronous mode, the transmission data is valid only for the destination node,
In the isochronous mode, it can be said that this mode is effective only for a receiving node that is determined to receive data of a certain channel.

If the network of FIG. 1 whose efficiency has deteriorated as described above is divided into a network composed of the nodes A and B and a network composed of the nodes C to G, the data between the nodes A and B can be obtained. Since the exchange and the data exchange between the nodes C to G do not compete with each other and can be performed in parallel, the data transfer efficiency of each network can be improved.
In addition, since there is no data exchange between the networks, there is no problem due to the division.

FIG. 2 is a diagram showing a network configuration in which the network shown in FIG. 1 is divided into two parts.
4 are connected using a serial bus. Node B from FIG.
FIG. 5 shows a state in which the connection between node A and node C is disconnected, and the network is divided into two networks, a network including nodes A and B and a network including nodes C to G. As a result, the nodes A and B perform data transfer between the nodes A and B, and need not wait for data transfer due to invalid data for the nodes A and B, such as data broadcast from the nodes C to G. Disappears. Similarly, nodes C to G are nodes C
To G, data to be broadcasted from the nodes A and B, and the like, there is no need to wait for data transfer due to invalid transfer data for the nodes C to G. For this reason, more efficient data transfer is possible in both networks.

<Management procedure of network configuration> In order to improve the efficiency of the network in this way, the root node on the network is used as a network management device, the usage status of the network is monitored, and it is determined that the network can be divided. If so, notify the operator to that effect,
Encourage action. In the management procedure to be described later, it is necessary to know the bus contention state. Therefore, the root node that performs bus arbitration is used as a network management device, but any node on the network that can know the bus contention state is used. It does not matter.

FIG. 5 is a block diagram of a node operating as a network management device. The program stored in the main memory 232 is executed by the CPU 231 to implement the procedures of the flowcharts of FIGS. 3 and 4 described later. The CPU 231 controls the entire node functioning as a network management device. File memory 233
Is a device that stores programs and data as files, such as a hard disk, a magneto-optical disk, and a flexible disk. A program for implementing the flowcharts of FIGS. 3 and 4 is also supplied from the file memory 233. The display 234 and the keyboard 235 are input / output devices for use by an operator. This network management device is connected to the network by an IEEE 1394 interface 236 as one node of the network. Here, two ports are provided, but any number may be used as long as it is within the standard. The configuration in FIG.
It is a configuration assuming. A printer further includes a printer engine, and a digital video camera further includes an image photographing unit and a reproducing unit. Similarly, other specific devices will also have a unique configuration for each device.

FIGS. 3 and 4 show flowcharts of the management procedure in the network management device. A bus reset is performed because a device is newly connected to a network connected by the 1394 serial bus or a cable is disconnected. FIG. 3 starts when a bus reset occurs.

When a bus reset occurs, as step S2, the parent-child relationship of the connected devices is determined by the node determination sequence, and the node ID of each node is determined. When the node determination sequence is completed, step S3
The information such as the parent-child relationship and the node ID determined during the node ID determination sequence is detected to recognize the address / parent-child relationship of each device. These recognized relationships are stored, for example, as a table in which nodes having a parent-child relationship are associated.

In step S4, the configuration of the entire network is recognized from the information. That is, for example, the individual parent-child relationships are integrated to construct and store the table. In the example of FIG. 1, the node B is recognized as the parent of the node A, the node B is recognized as the parent of the node C, and the link between the nodes is recognized. Further, by integrating these, it can be recognized that the node B is the parent of the nodes A and C.
Thus, the tree structure composed of the parent-child relationship of the nodes in the network is recognized and stored as network configuration information. Thereafter, data transmission using the network is performed.

After the network configuration is determined as shown in FIG. 3, when data transmission occurs from any node, the procedure shown in FIG. 4 is performed.

In step S5, the source and destination nodes of data are detected from the source ID, destination ID, channel number, and the like included in the data packet, and are detected against network configuration information. Recognize meaningful "paths". A meaningful path is a path of data transferred from the transmitting node to the receiving node. In the case of asynchronous transmission, the path of the data can be easily known from the source ID and the destination ID. In the case of isochronous transmission, the receiving node and the channel number can be known by notifying the receiving node of the channel number performed by asynchronous transmission prior to the isochronous transmission. The transmitting node can be informed by an arbitration request performed prior to the transmission. In this manner, a meaningful path can be specified whether in the asynchronous mode or the isochronous mode. The arbitration request and response data to the arbitration request are excluded from the data transfer whose path is recognized in step S5.

In order to recognize the data transfer path, if a meaningful path is specified in this way, for example, a use / non-use flag or the like is prepared in the network configuration information, and the recognized path is identified. Is marked for each link between a pair of nodes having a parent-child relationship. For example, when data is transmitted from the node B to the node D, the links BC and CD constituting the path are marked.

In step S6, it is determined whether or not there is a link through which the data packet does not pass in the network configuration, that is, a link between unused nodes. If there is an unused link that is not used, the data transfer is continued as it is in step S9. For example, in FIG. 1, in addition to the node B (PC), the node E (P
If printing from C) to node A (printer),
Since the links of nodes BC are used, they cannot be divided as shown in FIG.

If there is an unused link in the network, the data transfer efficiency is determined in step S7. The data transfer efficiency can be determined by monitoring the number of times (percentage) that data transfer is normally completed with respect to the number of data transfer requests, such as winning or losing in arbitration. Note that it is dangerous to regard a certain link as an unused link unless data transmission has been performed a certain number of times.Therefore, a predetermined number of data transfers have been performed after a bus reset or a predetermined time has elapsed. The data transfer efficiency can also be determined on the condition.

In step S8, based on the determination in step S7, it is determined whether the path is congested and the transfer efficiency is poor, and whether dividing the network improves the data transfer efficiency. If it is more efficient to divide the network, decide to divide the network, and do not divide the network if the bus is not crowded or if the transfer efficiency is not expected to improve even if the network is divided Is determined. Since the IEEE 1394 network has a tree structure, if there are unused links, the network can be divided there. However, even if it is divided, if one of the divided networks is only a leaf or the link is not used for one entire network, the data transmission is congested by nodes in the other network Yes, improvement in data transfer efficiency cannot be expected. For example, in such a case, it is determined that the network is not divided. That is, when it is determined that the data transfer efficiency is equal to or less than the predetermined value and the data transfer occurs in both of the two sub-networks divided at the dividable positions, the data transfer efficiency is determined by the division. If it is determined not to divide the network which is determined to be improved, the data transfer is continued as it is in step S9. If it is determined to divide the network, the process proceeds to step S10. Step S
In 10, a message notifying that the transfer efficiency is improved by disconnecting the cable between the nodes and increasing the connection efficiency, and information on the link between the nodes or the port in the node which may be disconnected may be removed. No link to send to nodes at both ends. The node receiving the message outputs the message by displaying it on a display provided.

When the user of the node receiving the message disconnects the cable from the port notified that it is OK to disconnect the cable and disconnects the connection to the subsequent node, a bus reset occurs again and the removed cable is disconnected. Is reconstructed, and data transfer is performed in each network.

As described above, when the network is congested, it is determined whether the network can be divided by checking the data path defined by the source and destination of each data transmission with the network configuration. I do. If it can be divided, it is determined whether or not the congestion can be resolved by the division. If the congestion can be resolved, the operator is notified that the congestion can be resolved by dividing the network and how the division can be performed. By prompting the operator to divide the network in this way, more efficient data transfer becomes possible.

In the node notified from the root node that the link may be removed, if the operator considers it necessary in the future or does not divide the network for other reasons, there is no need to do anything. good. In this case, in order to prevent the same message from being issued again, the root node serving as the network management apparatus may be notified that the link notified that the link may be removed is excluded from the removal candidates. Good. Upon receiving the notification, the root node does not notify the link related to the notification that the transfer efficiency will be improved by disconnecting the cable between the nodes and thereby increasing the transfer efficiency.

Further, the root node may be set in advance as a root node in order to exclude desired links and nodes from nodes and links to be notified that disconnection is better. In this case, it is not notified that a preset link should be disconnected even if the link is unused.

[0132]

[Other Embodiments] Even if the present invention is applied to a system including a plurality of devices (for example, a host computer, an interface device, a reader, a printer, etc.), an apparatus (for example, a copying machine) Machine, facsimile machine, etc.).

Further, an object of the present invention is to provide a storage medium storing a program code of software for realizing the functions of the above-described embodiments to a system or an apparatus, and to provide a computer (or CPU) of the system or apparatus.
Or MPU) reads and executes the program code stored in the storage medium.

In this case, the program code itself read from the storage medium implements the functions of the above-described embodiment, and the storage medium storing the program code constitutes the present invention.

As a storage medium for supplying the program code, for example, a floppy disk, hard disk, optical disk, magneto-optical disk, CD-ROM, CD
-R, a magnetic tape, a nonvolatile memory card, a ROM, or the like can be used.

When the computer executes the readout program code, not only the functions of the above-described embodiment are realized, but also the OS (Operating System) running on the computer based on the instruction of the program code. ) Performs part or all of the actual processing, and the processing realizes the functions of the above-described embodiments.

Further, after the program code read from the storage medium is written into a memory provided in a function expansion board inserted into the computer or a function expansion unit connected to the computer, based on the instruction of the program code, The case where the CPU of the function expansion board or the function expansion unit performs part or all of the actual processing, and the function of the above-described embodiment is realized by the processing.

[0138]

As described above, according to the present invention, in a case where data transfer between devices is congested in a device connected by a network, even if the network is divided, other devices are affected. To determine whether there is a higher transfer efficiency by dividing the network, and to notify the operator that the division has no effect and that the data transfer efficiency will increase, thereby improving the data transfer efficiency of the network. Attention can be given to the operator for improvement.

[0139]

[Brief description of the drawings]

FIG. 1 is a diagram showing an example of a network according to the present invention.

FIG. 2 is a diagram showing an example of a network according to the present invention.

FIG. 3 is a flowchart showing an operation at the time of a bus reset in the network management device according to the present invention.

FIG. 4 is a flowchart showing an operation at the time of data transfer between nodes in the network management device according to the present invention.

FIG. 5 is a block diagram of a node operating as a network management device.

FIG. 6 is a flowchart for explaining arbitration.

FIG. 7 is a diagram illustrating an example of a network configuration connected using a 1394 serial bus.

FIG. 8 is a diagram illustrating components of a 1394 serial bus.

FIG. 9 is a diagram showing an address map of a 1394 serial bus.

FIG. 10 is a sectional view of a 1394 serial bus cable.

FIG. 11 is a diagram for explaining a DS-Link coding scheme.

FIG. 12 is a diagram illustrating an example of determining an ID of each node on a 1394 serial bus.

FIG. 13 is a diagram for explaining arbitration on a 1394 serial bus.

FIG. 14 is a basic configuration diagram illustrating temporal state transition of asynchronous transfer.

FIG. 15 is a diagram showing an example of a format of an asynchronous transfer packet.

FIG. 16 is a basic configuration diagram illustrating a temporal state transition of isochronous transfer.

FIG. 17 is a diagram showing an example of a format of a packet for isochronous transfer.

FIG. 18 is an example of a bus cycle showing a state of a packet transferred on an actual bus in a 1394 serial bus.

FIG. 19 is a flowchart from a bus reset to a determination of a node ID.

FIG. 20 is a flowchart of parent-child relationship determination in bus reset.

FIG. 21 is a diagram showing a state after a parent-child relationship is determined in a bus reset
It is a flowchart figure until a node ID is determined.

Claims (16)

[Claims] 1. A connection means for connecting to a network, a path of data transferred on the network via the connection means is recognized, and it is determined whether the network can be divided in light of a configuration of the network. If it is determined that division is possible, it is determined whether or not the data transfer efficiency on the network is improved by the division.If it is determined that the data transfer efficiency is improved, the network management unit outputs a recommendation message recommending the division. Network management device. 2. The network management device according to claim 1, wherein the network management device further recognizes a network configuration at the time of initialization. 3. The network management device according to claim 1, wherein the recommendation message includes information on a division position. 4. The network management device according to claim 1, wherein the recommendation message is notified to a node corresponding to a division position. 5. The network management unit according to claim 1, wherein if there is an unused link through which transfer data does not pass between nodes on the network, it is determined that the link can be divided at that portion. Network management equipment. 6. The network management means, when it is determined that division is possible, if data transmission has occurred in both of the two sub-networks divided at the division position,
6. The network management device according to claim 1, wherein it is determined that the division improves data transfer efficiency. 7. The network according to claim 1, wherein said network is IEEE 1394.
The network management device according to claim 1, wherein the network management device is configured by a serial bus. 8. Recognizing a route of data transferred on a network, determining whether the network can be divided in light of the configuration of the network, and, if it is determined that the network can be divided, transferring data on the network by division. A network management method comprising: determining whether efficiency is improved; and, if determined to be improved, outputting a recommendation message recommending division. 9. The network management method according to claim 8, wherein a network configuration is recognized at initialization before recognizing a route of data transferred on the network. 10. The network management method according to claim 8, wherein the recommendation message includes information on a division position. 11. The network management method according to claim 8, wherein the recommendation message is notified to a node corresponding to a division position. 12. A determination as to whether the network can be divided is made such that if there is an unused link through which transfer data does not pass between nodes on the network, division is determined at that portion. 9. The method according to claim 8, wherein
The network management method described in 1. 13. The method for determining whether data transfer efficiency on the network is improved includes determining whether data transmission has occurred in both of two sub-networks divided at a position where the network is determined to be divisible. 13. The network management method according to claim 8, wherein it is determined that the data transfer efficiency is improved by the above. 14. The network according to claim 1, wherein said network is IEEE 139.
9. The system according to claim 8, wherein said serial bus comprises four serial buses.
The network management method described in 1. 15. A network system, wherein the network management device according to claim 1 is connected as one of nodes. 16. Recognizing a path of data transferred on a network, determining whether the network can be divided in light of the configuration of the network, and, when determining that the network can be divided, transferring the data on the network by division. A computer-readable storage medium storing a program for causing a computer to function as network management means for outputting a recommendation message for recommending division when it is determined whether the efficiency is improved.