JP2003101590A - Equipment, mechanism, method, and medium recording them, for dynamically exchanging bidirectional among devices having private address and devices having global address going over network routers - Google Patents

Abstract

PROBLEM TO BE SOLVED: To solve the problems of a G-side (global network) device being unable to specify the address of a P-side (private network) device and the G-side device being unable to directly transfer data to the P-side device, because the P-side device is connected to a G-side device by static address conversion of NAT of a router, and the number of transfer route (number of hops) increasing, if data transfer is made from the G-side device to the P-side device via a server installed in the G side. SOLUTION: The P-side device and the G-side device are connected to a session management server (S) installed in the G-side, and the S supervises data area which is assigned uniquely. Data are transferred directly between devices not passing through the S, according to the obtained session information, such as data transfer request, regarding own device. Direct data transfer from the G-side to the P-side becomes possible by replacing it to direct data acquisition from the P-side to the G-side. This method can be applied to the communication between P-side devices separated by the G-side network.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a mechanism, an apparatus and a method used for exchanging data using a computer and a network.

[0002]

2. Description of the Related Art A device having a globally assigned global address and a local area
a Network), but in order to exchange data between devices (hereinafter, “P-side device”) having private addresses that may be duplicated in the world,
There is a network router (hereinafter, "router") having a global address. This router is installed on the private network side of the perimeter where the private network connects to the global network,
P-side device has a global address (hereinafter,
In order to transfer data to and from the "G side device"), global address and private address conversion is performed.

In the conversion of the global address and the private address, NAT (Network) is used.
I using Address Translater)
Performs static address translation called P Forwarding. That is, it is set in advance to which P-side device in the private network the data transferred from the G-side device outside the private network to the global address assigned to the router is transferred. By doing so, data transfer from the G-side device to the P-side device was possible within the range planned by the administrator.

Next, description will be made with reference to FIG. NAT
When data is transferred from the G-side device to the P-side device without using, the G-side device cannot specify the P-side device having the private address. Further, since the address conversion is not performed, it is impossible for the G side device to transfer the data to the P side device. For this reason, G who wants to communicate
The device on the P side and the device on the P side are connected to a session management server (hereinafter abbreviated as “S”) having a global address, and data is stored in a data storage server (hereinafter abbreviated as “T”) installed on the global network. Data is exchanged while temporarily storing. Normally, S and T are integrated to form one server. That is, the communication control and the data transfer are usually performed by intermediating a server that integrates S and T or S and T.

Further, when P-side devices connected to independent private networks, that is, when private networks a and b exist, data transfer is performed between Pa and Pb connected to the respective networks. In this case, the other party's private address cannot be specified mutually. Therefore, first, P
a and Pb are connected to a session management server having a global address (hereinafter abbreviated as “S”) and temporarily store data in a data storage server (hereinafter abbreviated as “T”) installed on the global network. Exchange data while accumulating in. Normally, S and T are integrated to form one server. That is, the communication control and the data transfer are usually performed by intermediating a server that integrates S and T or S and T.

[0006]

In the case of interposing a router, it is possible to uniquely specify the G side device from the P side device, but it is impossible to specify the P side device from the G side device. Therefore, the data transfer from the P-side device to the G-side device was possible, but the data transfer from the G-side device to the P-side device was impossible. If NAT is used, it is possible for the router to access the specific P-side device from the G-side device outside the router by statically allocating the specific communication port to the specific P-side device. It was However, the present invention generally functions effectively even in a situation in which data is dynamically (variably) transferred between a P-side device and a G-side device.

When a server that integrates S and T or S and T is installed in a global network and communication is performed between devices (hereinafter referred to as “client device”) that communicate with each other by connecting to the server through the server, the client device is used. The amount of transfer or the number of transfer routes (the number of hops) is unnecessarily increased because they read from each other while writing to the server
Moreover, the server that stores the data was also burdened. Further, since the content of the communication between the client devices is stored in the server T managed by a third party, there is a problem in terms of security for maintaining the confidentiality of the communication.

[0008]

[Means for Solving the Problem] An S having a global address is installed on the global network (data is not stored in the session management server). Even if S and T are operated by the same computer, the following description will not be affected. The devices (both P side and G side) trying to communicate with each other first connect to S. At this time, S is unique to the client device, that is, a unique identifier in S (hereinafter, “session ID”).
Assign

S then attempts to connect from S to the client device in the conventional manner. If the connection is successful,
The connected device is the G-side device connected to the global network, and if the connection fails, it is the P-side device connected to the private network. S
Holds information on whether the connected device is a P-side device or a G-side device. Alternatively, when the G-side device and the P-side device can be easily discriminated from the IP address itself, it is possible to omit the confirmation by the above connection trial.

S prepares a data area (hereinafter abbreviated as "D") in which the client device can continuously connect and monitor the status. The client device maintains the connection to S constantly or periodically while continuing the communication according to the present invention.
Are watching. This is because S has a global address and can be specified by any device. For example, the specific P-side device Pa monitors the allocated data area Da, and the specific G-side device Gb monitors the allocated data area Db.

S holds the following data for each client device. (1) Session ID (2) IP address accessible to client device, or IP address and port number (3) IP address of communication control server that temporarily stores data in communication between private addresses, or IP address and port Number (4) Information that can specify the data to be transferred in the global network (hereinafter, "data ID") (5) Whether the client device exists in the global network or in the private network (6 ) From which device the access request is issued to the client device (7) Client device status (7) "Client device status" means that the client device is in communication or is on standby. State of communication such as noka It is.

Since the G side device can be specified from the P side device, in the P → G direction data transfer, the P side device directly transfers the data to the G side device in a normal manner without passing through T. In this data transfer, Pa which is a specific P-side device
Notifies S that communication has started with respect to a specific G-side device Ga. Pa monitors Da.

When data is to be transferred from the G-side device to the P-side device, the G-side device uses the private side of the P-side device.
Since the address cannot be specified, data cannot be directly transferred to the P-side device. Therefore, the specific G-side device Gb informs S that it wants to transfer data to the P-side device Pa. At this time, Gb cannot know the IP address of Pa existing on the private network, so Pa is specified on S. That is,
This data transfer from the G side to the P side is possible only when all are connected to S.

Upon receiving a data transfer request from Gb to Pa, S receives a data area Da monitored by Pa.
And (1) Gb requests data transfer to Pa, and (2) Gb global address,
(3) Data ID of the data that Gb is about to transfer
Write. The data of (1), (2) and (3) are collectively referred to as "transfer start data".

Since Pa monitors Da, when transfer start data is written in Da, the information is acquired. Subsequently, Pa uses the acquired global address of Gb to directly connect to Gb, and then uses the data ID to acquire the data that Gb tried to transfer to Pa from Gb to Pa. That is, in the data transfer in the G → P direction, it was impossible to specify the address of the P-side device, but in the present invention, the address of the G-side device including the server can be specified from the P-side device. By utilizing the fact that the data transmission in the G → P direction performed by G is replaced by the data acquisition in the P ← G direction performed by P, the data in the G → P direction is effectively It enables transmission.

In addition, in the present invention, S only mediates the communication as described above, and the data communication itself is carried out directly between the client devices that communicate with each other. Therefore, the efficiency of data transfer does not decrease as compared with the conventional method. In addition, as described below, the transfer method according to the present invention can improve the data transfer efficiency.

The improvement in transfer efficiency that can be achieved by using the present invention will be described. FIG. 1 is a schematic diagram of a case where client devices exchange data with each other without using the present invention, that is, a conventional data transfer. In FIG. 1, the devices connected to the network are Pa, Pb, Ga, and Gb, respectively. P
a and Pb are devices having private addresses inside the router, and Ga and Gb are devices having global addresses. Further, T indicates a data storage server that temporarily stores data in data transfer between client devices. Reference symbol S denotes a server that manages connection, that is, session management in data transfer.

FIG. 2 shows a list of the number of communication paths (the number of hops) required for the conventional data transfer of FIG. The numbers of hops required for Pa, Pb, Ga, and Gb to transfer data to each other are shown. The total number of hops required in FIG. 1 is calculated to be 100.

FIG. 3 is a schematic diagram in the case where client devices exchange data with each other using the present invention. The meanings of the symbols used in FIG. 3 are the same as those in FIG. Similar to FIG. 2, FIG. 4 shows a list of hop numbers required for data transfer when the present invention is used. In the present invention, the session management server S controls so that the data transfer is performed directly between the client devices, so that the number of hops is reduced in the communication between the G-side device and the P-side device. However, in the communication between P-side devices that do not mutually know their private addresses, no decrease is seen because it is necessary to transfer and acquire data from T as in the conventional case. The number of hops required in FIG. 3 is as listed in FIG. 4, and is calculated as 60 as a whole.

What is shown in FIG.
It is a figure in the case of implementing the invention described in. Install a communication control server with a global address on the boundary between the private network and the global network. In FIG. 5, these servers are shown as Ta and Tb. In addition, P side devices existing in the same private network are respectively Pa, Pa '. ． ． As shown.

In the situation shown in FIG. 4, it was not possible to reduce the number of hops in the communication between the P side devices. Therefore, in communication between devices having private addresses respectively installed in two different private networks that exist independently of each other, that is, in the communication between Pa ← → Pb, the communication control servers Ta and Tb mutually communicate with each other. Try to transfer data directly. Communication control between Ta and Tb is performed by the session management server S installed on the global network. Communication control when Ta and Tb are installed is performed as follows.

Private with Ta and Tb installed
In communication within the network, Pa and Pa 'can mutually know private addresses, and thus directly perform data transfer with each other. In this case, the data exchanged inside the private network will not be sent across the router to the global network. This also increases the security of the transferred data.

In the case of communication between the devices Pa and Pb, which are respectively connected to the independent private networks in which Ta and Tb are installed, it is not possible to mutually know the private address by a normal communication method. It is only possible to know the global addresses of the routers and communication control servers Ta and Tb that are in contact with the global network side.

By using the present invention, the number of hops required for communication between Pa and Pb can be greatly reduced. Pa, Pb
Connects to the session management server S on the global network and monitors the data areas Da and Db, respectively. For example, when transferring data in the Pa → Pb direction,
Pa sends data to Pb to Ta installed in its own private network (hereinafter referred to as “X”).
Abbreviated) is temporarily accumulated.

Subsequently, a data ID which is an identifier capable of specifying X is acquired. Pa issues a data transfer request to Pb to S. The data I of X is included in this data transfer request.
D is also included. S writes the transfer start data in the data area Db. This transfer start data includes Pa
It contains the information necessary for Pb to connect to Ta and acquire X.

When Pb acquires the transfer start data of Db, Pb recognizes that X originates from Pa and acquires X from Ta. By doing so, X is the conventional transfer method Pa → router → T shown in FIG.
→ Router → Pb, not Pa → Ta →
It is performed in Pb. The data transfer from Pb to Pa is also performed by the same means.

The number of hops reduced by these transfer methods is as listed in FIG. 6, and is 4 in total.
Calculated as 0. Thus, the present invention streamlines the transfer.

[0028] Further, conventionally, data is transferred to T via a global network, and T is managed by a third party.
After the data was temporarily stored in the network, the data was transferred again via the global network, and there was a high risk that the data being transferred could be intercepted. However, in the present invention, Ta is a server that is managed in the private network to which the data transmission source Pa belongs, data accumulation in Ta is relatively safe, and transfer to Tb via the global network is performed. Is also reduced to one route. These also improve the security of data transfer.

[0029]

The present invention enables dynamic data transfer from a device having a global address to a device having a private address, that is, data transfer in the G → P direction, which has been impossible in the past.

The present invention improves data transfer efficiency by reducing the number of routes required for data transfer.

The present invention shortens the data transfer path,
Improves data transfer security, especially by minimizing transfers over global networks.

The present invention enables private address-to-address communication, that is, Pa ← → Pb communication, without going through the storage server T. Omission of the storage server T also improves data transfer efficiency.

[0033]

BEST MODE FOR CARRYING OUT THE INVENTION The present invention can be generally applied to data exchange via a computer network such as the Internet.

[Brief description of drawings]

FIG. 1 is a schematic diagram of data transfer using a conventional method.

FIG. 2 is a table of the number of hops required for data transfer using the conventional method.

FIG. 3 is a schematic diagram of data transfer performed when at least one client device has a global address, which is performed using the present invention.

FIG. 4 is a table of the number of hops required for data transfer when at least one client device has a global address performed by using the present invention.

FIG. 5 is a schematic diagram of data transfer performed when both client devices have private addresses, which is performed using the present invention.

FIG. 6 is a table of the number of hops required for data transfer when both client devices have private addresses, which are performed using the present invention.

Claims (4)

[Claims] 1. A device, mechanism, which enables direct data transfer from a device having a global address to a device having a private address by installing a communication control server having a global address on a network. Methods and media recording these. 2. A device, a mechanism, a method, and a method for enabling data transfer between a device having a global address and a device having a private address by installing a communication control server having a global address on a network. A medium that records these. 3. A communication control server having a global address is installed on the network to directly transfer data from a device having a global address to a device having a private address, thereby improving transfer efficiency. A device, a mechanism, a method for improving the same, and a medium recording these. 4. A private network and a global network. By installing a communication control server having a global address on the boundary between the networks, the private network and the global network are installed in two different private networks that exist independently of each other. An apparatus, a mechanism, a method, and a recording medium for improving transfer efficiency by directly transferring data between respective communication control servers in communication between devices having private addresses.