JPH0669921A - Communication network - Google Patents

Abstract

PURPOSE:To improve the utilizing efficiency of a communication resource by allocating a band flexibly in the communication network supporting the communication not requiring real time performance. CONSTITUTION:A communication node decides a time allocation of a communication band based on an application item, the time allocation is stored in a band management table 32 in the communication node, the communication node informs the start of communication to the host according to the band management table 32 to set a logic path for the communication. Furthermore, when the band different from the applied band of the host is allocated, after the communication node stores once the information sent by the host to a file system, stored information is transferred to the destination host by using the allocated band.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a communication network that simultaneously supports a plurality of communications that do not require real-time communication. In particular, by scheduling a communication band based on a notification from a user, communication resources can be efficiently used. Related to a communication network to be used in a specific manner.

[0002]

2. Description of the Related Art As a technique for scheduling a communication band based on a report from a user, there is a technique similar to the NTT technical reference material "Satellite Digital Communication Service" (September, 1986). In this conventional technique, the user terminal specifies a communication start date / time, a communication end date / time, a destination terminal address, and a communication attribute (bandwidth, broadcast, symmetry) with respect to the network, and makes a reservation request for communication. The network determines whether or not it is possible to accept the reservation application based on the bandwidth allocation schedule and the designated items of the reservation application when the reservation application occurs. When the reservation application can be accepted, the reservation number is notified to the user terminal (reservation registration).

On the other hand, if even one of the designated items is not satisfied, the user terminal is notified that reservation registration cannot be performed. For example, when a user terminal operates on July 11, 1992,
1.5 Mbps from 9am to 9:30 am on the same day
When I try to reserve a band, the available time slot is 1Mbp
If it is s, the reservation application is rejected. The host for which the reservation application is rejected will make another reservation application to secure the communication band. At this time, it is possible to inquire the current schedule to the network (inquiry of free time), whereby the host can make a reservation application after confirming the time zone in which the band is free.

[0004]

The above-mentioned prior art has the following problems.

First, in the above-mentioned conventional technique, the reservation application is immediately rejected if the band is not available between the communication start time and the communication end time requested by the host. For example,
If a band is not available at 9:30 when a reservation is made at 9:30, the reservation application is rejected even if the band is available from 9:31. When the host whose reservation has been rejected reserves the band from 9:31,
The procedure of free time inquiry / re-application must be included. Reservation application / registration (rejection) is frequently performed when the network becomes large in scale, so it is fully conceivable that free bandwidth will be allocated between free time inquiry and re-application. That is, with the method in which the host specifies the start time / end time as in the prior art, it becomes difficult to match the start time / end time with the free time of the band. Therefore, free time inquiry / re-application is repeated for bandwidth allocation, and there is a problem that the throughput of the network decreases.

Secondly, in the prior art, the reservation acceptance judgment is performed on the assumption that the bandwidth requested by the host is guaranteed. Therefore, the communication which can be reserved and registered is limited to the one which has applied for the bandwidth smaller than the free bandwidth. Was there. However, in the case of communication that does not require real-time processing, information from the host can be stored in the network, and the stored information can be transferred using a bandwidth that the network can accommodate. Therefore, even if the requested band is wider than the free band, information can be transferred, and more efficient communication band scheduling can be performed. Since the conventional technology does not consider the function of accumulating information from the host and the function of transferring information using the bandwidth that the network can accommodate, it is efficient when assuming communication that does not require real-time processing. There was a problem that scheduling could not be performed.

[0007]

[Means for Solving the Problems] As means for solving the above-mentioned first problem, a host is provided with means for declaring a used band and a used time to a network, and a time allocation of the band based on the declared value is assigned to the network. Means for determining
The information transfer route is set according to the time allocation, and further,
A means for notifying the host of the start of communication is provided.

As means for solving the second problem, means for accumulating information from the host in the network and means for allocating a band different from the used band declared by the host in response to the traffic load of the own network When,
There is provided means for temporarily accumulating information corresponding to the difference between the declared bandwidth and the allocated bandwidth in the accumulating means and transferring the accumulated information using the allocated bandwidth. Further, the network is provided with means for terminating the communication protocol defined between the hosts.

[0009]

In the communication network using the means for solving the first problem, the time zone declaration from the host is not the start time / end time but the use time. Therefore, the network specifies a specific time when allocating the communication band. The reservation can be assigned to an arbitrary time zone in which a band commensurate with the declared band is free, without being restricted by the band.

In the communication network using the means for solving the second problem, even when the bandwidth declared by the calling host and the bandwidth allocated by the network are different, information corresponding to the difference in bandwidth is stored in the information storage means. be able to. The information stored in the storage means is transferred in the network using the allocated bandwidth. With the above technology, if communication that does not require real-time processing is premised, inter-host communication can be realized even if the bandwidth declared by the host cannot be allocated. In this network, flexible bandwidth allocation becomes possible compared to conventional bandwidth allocation that has been performed with the bandwidth declared by the user (host) unchanged.
Efficient network operation can be performed. or,
The means for terminating the communication protocol enables the network to return the control information on behalf of the receiving host when the host needs the control information (for example, delivery confirmation) along with the information transfer. As a result, information can be stored in the network without disconnecting the connection between the hosts.

[0011]

FIG. 2 shows an example of a communication network to which the present invention is applied. In FIG. 2, the communication nodes 1, 2, and 3 are
Equipment that implements the communication protocol (hereinafter referred to as the host)
7, 8, 9, 10, 11 and the information storage device (hereinafter referred to as a file system) 4, 5, 6 are connected to the digital dedicated line 1
Connect via 2, 13, and 14. Hosts 7, 8, 9,
A logical path is set between the communication nodes 10 and 11 by the communication nodes 1, 2 and 3. If the logical path supports a communication service that does not require real-time processing, the information transmitted from the originating host is temporarily stored in the file system 4,
It is accumulated in 5 and 6. In this case, there are at least two types of logical paths managed by the network: source host-file system and file system-receiver host.
Incidentally, when N file systems are interposed between the originating host and the receiving host, there are (N + 1) types.

The file systems 4, 5 and 6 not only store user information between hosts but also terminate the communication protocol implemented in the host. Specifically, in a communication protocol defined between hosts for realizing a service, when exchanging control information (for example, delivery confirmation) with information transfer, the file systems 4, 5, 6 are protocols. Terminate. For example, in the case of supporting a service such as a packet communication in which the receiving host sends back a delivery confirmation every time the transmitting host sends information (packet), the file systems 4, 5, 6 are It recognizes the packet sent by the host, and returns a delivery confirmation of the packet on behalf of the receiving host. In such a case, the file system 4,
5 and 6 implement and terminate the protocol implemented in the host in order to realize the service.

When the exchange of control information associated with the information transfer is specified in the application layer, the file systems 4, 5 and 6 terminate up to the application layer. On the other hand, when there is no exchange of control information in the communication protocol defined between the hosts (for example, when the calling host supports a service that continues transmitting information until the communication is completed), the file systems 4, 5, 6 Has only the function of storing the information transferred from the network. That is, the protocols implemented by the file systems 4, 5 and 6 are only the information transfer protocol in the network, not the communication protocol implemented in the host.

In response to a reservation request for a communication band from the host, the band management system 15 performs band allocation in which the band used for each communication and the start / end time are clearly defined. The bandwidth management system of the present embodiment is not implemented in independent hardware, and it is the communication node that accommodates the host that actually allocates the bandwidth to the host. In this embodiment, one band management system is constructed as the entire network by notifying the other communication nodes of the band allocated to the host accommodated by the communication node.

FIG. 3 is a block diagram of modules of the communication node 1 in this embodiment. In FIG. 3, the communication node 1 is a subscriber interface (hereinafter LIF) 16, 1
7, transmission interface (hereafter TRK) 18, 19, file system interface (hereafter FIF) 20, central control unit (hereafter CPU) 21, exchange unit (hereafter SW) 22
Composed of.

The LIFs 16 and 17 convert the information from the hosts 7 and 8 into ATM (Asynchronous Transfer Mode) cells and transmit the cells to the SW 22. Furthermore, SW2
User information is generated from the cell received from 2 and transferred to the hosts 7 and 8. The CPU 21 LIFs the virtual channel identifier (VCI) and virtual path identifier (VPI) required for routing in the cell network and the route information required for cell routing in the SW22.
Notify 16,17. The TRKs 17 and 18 convert the cells transferred from the SW 22 into a predetermined transmission format and transmit the cells to the digital leased lines 12 and 13. Further, ATM cells are cut out from the digital dedicated lines 12 and 13, and SW
22 to 22. The FIF 20 is an interface between the SW 22 and the file system (hereinafter FS) 4. FIF
20 disassembles the ATM cell transferred from SW22, AT
The payload of M cells is sequentially written into FS4. That is, the logical path from the source host is once terminated by the FIFO 20, and the logical path to the destination host is FS by the FIFO 20.
-It is newly set between the receiving hosts. The ATM header indicating the logical path between the FS 4 and the receiving host is added to the payload by the FIF 20. The route information in the ATM header is notified by the CPU 21 like the LIF. CP
In addition to the above-described cell routing control, U21 performs band allocation control based on a communication band reservation application from the host.
The function of the CPU 21, particularly the time allocation control of the band will be described in detail later.

FIG. 4 is a functional block diagram of the CPU 21. The cell receiving unit 23 uses the VCI / V in the ATM cell header.
Based on the PI, the received cell is transferred to the band management unit 24 or the processor 25. Information transfer to the band management unit 24 and the processor 25 is not performed every time a cell is received, but a command / data to the band management unit 24 or the processor 25 is assembled by the cell reception unit 23 and then transferred. In addition, ATM
Conversion from cell to command / data is VCI, VPI,
And ATM adaptation header.

The band management unit 24 controls the band allocation of the communication network based on the command / data from the cell receiving unit 23. The bandwidth allocation control of the present invention will be described later. The band management unit 24 transfers the result of the band allocation control to the processor 25 together with the originating / receiving side host address and the transfer route information. The processor 25 sets the VCI / VPI / route number for communication based on the transfer route information,
Write to the route information management table 26. In the route information management table 26, the VCI / VPI / route number is managed for each logical path, and the information regarding the allocated bandwidth of each logical path is not managed. On the other hand, the bandwidth management unit 24
The bandwidth management table in the table manages the communication bandwidth allocated to the inter-host communication, and does not manage the route information of VCI / VPI / route number. The processor 25 also manages the bandwidth of the entire network by transmitting and receiving bandwidth management table information of each communication node to and from the processor of another communication node.

The billing management unit 27 manages billing for the user based on the amount of transfer information, the used bandwidth, the used time, etc. for each logical path. In this embodiment, VAN (Value Added N
Etwork: Value-added communication network) is assumed to accommodate a plurality of small-scale users under a single network, and charging control for users is performed. By applying the present invention, it becomes possible to perform a service which has not existed in the past and a charge control associated therewith. The charging control will be referred to after the description of the bandwidth allocation control.

The cell transmission unit 28 converts the call setting control information transmitted from the band management unit 24 and the logical path setting control information transmitted from the processor 25 into control cells and transfers them to the SW. The call setting control information is control information that prompts the host to transfer information from the network based on the schedule defined in the bandwidth management table. On the other hand, the logical path setting control information is a control signal for setting a logical path for inter-host communication according to the schedule defined in the bandwidth management table and the route defined in the route information management table.

Next, the band management unit 24 will be described with reference to FIG. When the reservation application acceptance circuit 29 receives a command (reservation application command) from the cell reception unit 23, the reservation parameters of the command (originating host address, recipient host address, used bandwidth-usage time, or transfer data amount) And the reservation acceptance time are written in the reservation management table 30. The scheduler 31 reads the reservation parameters from the reservation management table 30 at any time, and determines the time allocation of the used bandwidth according to the reservation parameters. Furthermore, scheduler 3
1 reads the bandwidth allocation information from the bandwidth management table 32 in synchronization with the reading of the reservation parameter. As a result, when the scheduler 31 newly allocates a band, the latest band management information can always be referred to. The newly determined time allocation of the band is written in the band management table 32 as band management table update data. Further, the scheduler 31 transmits the call setting control information to the control cell generation circuit 33 based on the bandwidth management table, and also notifies the processor to set the communication logical path.

Next, the control sequence from the reservation application of the calling host to the end of communication in this embodiment will be described with reference to FIGS. 5 and 6.

FIG. 5 shows a control sequence from the reservation application of the calling host to the completion of bandwidth allocation (reservation registration). First, the originating host makes a reservation application to the network. At the time of reservation application, the reservation parameters (source host address, recipient host address, bandwidth-usage time, or transfer data amount) are notified to the network. The network writes the reservation parameter and the time when the reservation application is accepted in the reservation management table, and transmits the reservation application registration to the originating host. In this embodiment, in principle, reservation application registration for all reservation applications, that is,
Although the reservation management table is written, if the reservation management table cannot be written because the reservation management table is full, the new reservation application is rejected. Next, the network performs time allocation of the bandwidth used for host-to-host communication based on the reservation parameters registered in the reservation management table. Upon completion of bandwidth allocation, the network sends a reservation registration to the originating / receiving host. The reservation registration process is completed by the host transmitting ACK to the network when the reservation registration is received. On the other hand, when the network cannot accept the reservation application, the registration refusal is notified to the calling host. At this time, the control information is not notified to the receiving host. The procedure for determining the time allocation from the reservation parameters will be described later with reference to FIGS. 7, 8 and 9.

FIG. 6 shows a control sequence from the completion of reservation registration to the end of communication. First, the network sends a call setting (communication start) command to the originating / receiving host at a predetermined communication start time after the reservation registration is completed. The communication start time is a time determined when the time allocation of the band is determined, and is recorded in the band management table. The host returns ACK to the call setup command, and the network
The communication start process is completed by returning an ACK response to K. Further, the network sends a call disconnection request command to the calling / receiving host at a predetermined communication end time.
In response to the call disconnection request command, the host returns a release command to the network, and the network notifies the host of the release command after the release processing, whereby the communication ends.

Next, with reference to FIGS. 7, 8 and 9, a band time allocation procedure in this embodiment will be described. Figure 7
Shows the structure of the reservation management table. A reservation number is assigned to each reservation parameter, and is stored in the management table in the order of reservation acceptance time, originating host address, recipient host address, bandwidth used, and usage time (or transfer data amount). . Reservation numbers 1, 2, and 3 are types for declaring used bandwidth and use time, and reservation number 4 is a type for declaring transfer data amount.

FIG. 8 shows the structure of the bandwidth management table. In this embodiment, first, the bands are allocated in the order in which reservation applications are received. From FIG. 7, reservation number 1 (used bandwidth 10
Mbps, usage time 1 hour and reservation number 2 (use bandwidth 100)
Mbps, usage time 30 minutes) from 0:00. Next, the reservation number 3 (used bandwidth 80 Mbps, used time 25 minutes) is allocated from 0:00, but since the entire bandwidth is 150 Mbps, the free bandwidth from 0:00 is only 40 Mbps. Therefore, the bandwidth of the reservation number 3 is assigned 40 Mbps from 0:00 to 0:30 and 80 Mbps from 0:30. As a matter of course, (40 Mbps x 30 minutes) + (8
0 Mbps × 10 minutes) can transfer the same amount of data as (80 Mbps × 25 minutes). Since the band notified by the reservation number 3 is 80 Mbps, information transfer must be performed in a band narrower than the band desired by the host until 0:30. Therefore, in this embodiment, the difference between the transfer band between the host and the network and the transfer band within the network is absorbed by the information storage device (file system: FS) within the network. Specifically, the calling host with reservation number 3 transfers information to the FS at 80 Mbps, the network temporarily stores the information in the FS, and then transfers to the receiving host at 40 Mbps until 0:30.
80 Mbps transfer is performed until 0:40. Another method of absorbing the band difference is to specify a transfer band from the network to the host and the host transfers information in the specified band. However, in this case, it is necessary for the host to have a function of transferring information in a band designated by the network.

Finally, the reservation number 4 is assigned. Since the reservation number 4 does not specify the band to be used, the network sets a band to be used based on the free band and notifies the host of the band. In this embodiment, a 60 Mbps band is allocated from 0:30 to 0:40, and a 140 Mbps band is allocated from 0:40 to 0:50.
The host with reservation number 4 transfers information using the notified band.

FIG. 9 shows the result of time allocation of the above bands. Of these, reservation numbers 1 and 2 are assigned the band designated by the host as they are. The reservation number 3 absorbs the difference in the used band by accumulating the transfer information in the FS, and efficiently uses the band of the communication network.

Next, the charge management will be described. In this embodiment, the bandwidth is set for the reservation numbers 3 and 4 for the convenience of the network. From the network manager's point of view, setting the communication bands for reservation numbers 3 and 4 as described above enabled more efficient use of resources, so it is cheaper for these communications. Set the charge. In other words, by applying the present invention, a high fee is set when information is transferred in compliance with the band or time specified by the host (user), and when the transfer band or transfer time is arbitrary. It will be possible to provide a service that sets a low price.

Next, a second embodiment will be described. In this embodiment, the allowable transfer time is declared when the host makes a communication band reservation for the network. The permissible transfer time indicates the permissible time from the reservation application time of the communication band to the completion of communication by the host. This parameter enables more flexible bandwidth allocation than the first embodiment in which bandwidth allocation is performed in the order of reservation application registration time. Hereinafter, the time allocation control of the band in consideration of the allowable transfer time will be described with reference to FIGS. 10, 11, and 12.

FIG. 10 shows the structure of the reservation management table in the second embodiment. A reservation number is assigned to each reservation parameter and stored in the management table in the order of reservation acceptance time, originating host address, recipient host address, bandwidth used, usage time (or transfer data amount), and allowable transfer time. Has been done. The allowable transfer time for the reservation numbers 1 and 4 is 2 hours, the allowable transfer time for the reservation number 2 is 6 hours, and the allowable transfer time for the reservation number 3 is 1 hour. From the relationship between the reservation reception time, the use time, and the allowable transfer time, it can be seen that the reservation number 2 can allocate bandwidth with a sufficient time, and the others have a relatively small time allowance.

FIG. 11 shows the structure of the bandwidth management table in the second embodiment. In the present embodiment, first, bands are allocated in order of reservation acceptance from those having a small time margin for band allocation. Reservation number 1 from Figure 10
(Reservation acceptance time 23:30, bandwidth used 10Mbps,
The use time of 1 hour and the allowable transfer time of 2 hours and the reservation number 3 (reservation reception time of 23:40, bandwidth of 80 Mbps, use time of 25 minutes, allowable transfer time of 1 hour) are allocated from 0:00. 0 for reservation number 4 (reservation acceptance time 23:45, transfer data amount 12 Gb, allowable transfer time 2 hours)
Bandwidth is allocated from 0:00, but the total bandwidth is 1
Since it is 50 Mbps, the free bandwidth from 0:00 is 6
There is only 0 Mbps. Therefore, the bandwidth of reservation number 4 is 60 Mbps allocated from 0:00 to 0:25, and
140 Mbps is allocated from the hour 25 to the hour 0:30 (the exact communication end time is 0:28:57). Finally, the reservation number 2 is allocated a band of 100 Mbps from 0:30 to 1:00. still,
If another reservation application occurs before the bandwidth allocation of the reservation number 2, the allocation of the reservation number 2 may be delayed. The result of band allocation in this embodiment is shown in FIG.

As another allocation control, a 140 Mbps band can be allocated to reservation number 2 from 0:30. The bandwidth used for reservation number 2 is 100 Mb
Since it is ps, it is necessary to store the information transferred by the calling host having the reservation number 2 in advance in a storage device (FS) in the network. For example, 0:00 to 0:30
Information is transferred from the originating host to the FS by the minute. The band of transfer between the originating host and the FS is arbitrary. Next, from 0:30 when the network bandwidth becomes available, the available bandwidth (140
Transfer to the receiving host using all of M). In this case, it takes about 22 minutes to occupy the network band (140 Mbps), and 10 minutes from 0:30 to 10 minutes.
The communication ends at a time earlier than when the 0 Mbps band is assigned. Furthermore, there is no vacant band from 0:30 to 0:52, which enables efficient network operation. FIG. 13 shows the allocation result when this allocation method is used.

Next, the charging management in this embodiment will be described. In the present embodiment, in particular, the bandwidth allocation of the reservation number 2 is postponed compared to the other areas, so that it becomes possible to preferentially allocate the bandwidth allocation for other communication with less time margin. Specifically, since the reservation number 2 is postponed, it becomes possible to allocate the bandwidth of the reservation number 4 within the allowable communication time, and as a result, the utilization rate of communication resources is improved. Therefore, from the viewpoint of contributing to the effective use of communication resources, the reservation number 2 which is postponed is set to a lower charge than other charges. That is, a low charge is applied to communications that are applied with sufficient time for bandwidth allocation,
A high fee service is available for other communications.

[0035]

According to the present invention, flexible bandwidth allocation can be realized and communication resource utilization efficiency can be improved in a communication network that supports communication that does not require real-time processing.

[Brief description of drawings]

FIG. 1 is a block diagram of a bandwidth management unit according to an embodiment of the present invention.

FIG. 2 is a block diagram of a communication network to which the present invention is applied.

FIG. 3 is a module block diagram of a communication node according to an embodiment of the present invention.

FIG. 4 is a block diagram of a central control unit according to an embodiment of the present invention.

FIG. 5 is an explanatory diagram of a control sequence from reservation application to reservation registration.

FIG. 6 is an explanatory diagram of a control sequence from reservation registration to completion of communication.

FIG. 7 is an explanatory diagram of a reservation management table according to an embodiment of the present invention.

FIG. 8 is an explanatory diagram of a bandwidth management table according to an embodiment of the present invention.

FIG. 9 is an explanatory diagram of a first example of band allocation to which the present invention is applied.

FIG. 10 is an explanatory diagram of a reservation management table according to the second embodiment of this invention.

FIG. 11 is an explanatory diagram of a bandwidth management table according to the second embodiment of this invention.

FIG. 12 is an explanatory diagram of a second example of band allocation to which the present invention is applied.

FIG. 13 is an explanatory diagram of a third example of band allocation to which the present invention is applied.

[Explanation of symbols]

24 ... Band management unit, 29 ... Reservation application receiving circuit, 30 ... Reservation management table, 31 ... Scheduler, 32 ... Band management table, 33 ... Control cell generation circuit.

Claims (5)

[Claims] 1. In a communication network that simultaneously supports a plurality of communication services that do not require real-time processing, a host has means for reporting at least a used band and a used time to the network, and the network has the declared value. A communication network comprising means for determining time allocation of bandwidth based on the time allocation, and means for setting an information transfer path according to the time allocation and notifying the host of the start of communication. 2. The network according to claim 1, wherein the network is
A means for accumulating information from the host, and
Means for allocating a band different from the band used by the host in response to the traffic load of its own network,
A communication network having means for temporarily accumulating information corresponding to the difference between the declared bandwidth and the allocated bandwidth in the accumulating means, and transferring the accumulated information using the allocated bandwidth. 3. In a communication network that simultaneously supports a plurality of communication services that do not require real-time processing, the host has means for reporting at least the transfer information amount to the network, and the network is based on the transfer information amount. Means for notifying the host of the used band and the used time of communication, and means for deciding the time allocation of the band based on the used band and the used time, and the information transfer route according to the time allocation. Set
A communication network comprising means for notifying the host of the start of communication. 4. The host according to claim 1, 2 or 3, wherein the host has means for reporting an allowable time until the communication is ended or an allowable time when the communication is ended, and the network is the allowable time or the allowable time. A communication network having means for determining time allocation of bandwidth based on time of day. 5. The communication network according to claim 1, 2, 3 or 4, wherein the network terminates a communication protocol defined between hosts for realizing a communication service.