JP2003258878A - Method for deciding address for network system, and network terminal - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To enable new nodes to automatically and selectively cope with both of a server system and a distribution system and to obtain a network address by either of the systems to participate in a network. <P>SOLUTION: In this network system consisting of a plurality of nodes Ni and Nx, each node simultaneously has the address obtaining functions of a server system for accepting the network address to be imparted from an address server 1 and a distribution system for allowing the network-participating nodes Ni and Nx to autonomously decide the network address. When the new node Nx participates in the network, an address request message is multi-addressed, and a subsequent address obtaining system is decided alternatively and autonomously from the server system or the distribution system depending on the content of a response and the presence/absence of the response within a prescribed time to obtain address decision or address decision information from the content of the response and the presence/absence of the response. <P>COPYRIGHT: (C)2003,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a server system in which, when a network terminal newly joins a network as a node, the network address necessary for identifying each node is acquired, and the network address is distributed from an address server. The present invention relates to a network system address determination method and a network terminal that automatically selects either a distributed method in which each network terminal autonomously determines a network address according to the state of the network, and determines the network address.

[0002]

2. Description of the Related Art Conventionally, an automatic selection method of a server method in which a network address is distributed from an address server and a distributed method in which each node autonomously determines a network address has been realized at the time of IP address distribution in an IP network. There is.

[0003]

However, in the conventional method for determining the address of the network system, the network goes down when switching between the server system and the distributed system in which network addresses having different address systems are used. There was a problem that caused it. In addition, since the IP network has a wide address space with respect to the number of nodes originally used, an address selection method that takes almost no attention to address duplication is adopted, and there is a problem in that it is wasteful when applied to a small-scale network. there were.

In FIG. 1, the address server 1 is a new node Nx.
Shows a server method for distributing network addresses. In this server system, if the communication between each node Ni and the address server 1 is guaranteed, the new node Nx
Network addresses can be distributed without conflict with. FIG. 2 shows a distributed system. In this distributed method, it is necessary to ensure communication between each existing node Ni and the newly joining node Nx.

Therefore, in the server system, the time required for one new node Nx to acquire the network address is short, and in the distributed system it is long. However, in the latter case, there is an advantage that the address server 1 is unnecessary. Due to the characteristics of each system, it is desirable to use the server system for large-scale networks such as buildings and the distributed system for small-scale homes.

According to the present invention, the new node automatically selects and supports both the server system and the distributed system, and is equipped with the function of acquiring the network address in any of the systems, and the usage can be easily selected. It is intended to provide an address determination technology for a network system.

Further, according to the present invention, the same network address system is used in the server system and the distributed system, and when switching between the server system and the distributed system, the network address is not changed and the address distribution system can be smoothly switched. Another object of the present invention is to provide an address determination technique for a network system in which the number of nodes used is equal to the address space, and the address duplication can be reduced even under the condition that address competition frequently occurs in the conventional case.

[0008]

An address determining method for a network system according to a first aspect of the present invention is a network system comprising a plurality of nodes, which accepts a network address given from an address server, and network participation. Each node is equipped with an address acquisition function, which is a distributed method in which each node autonomously determines a network address. Each node broadcasts an address request message when participating in the network, and responds within a predetermined time and whether there is a response or not. According to the method, the subsequent address acquisition method is autonomously determined in the alternative of the server method and the distributed method, and the address determination or the address determination information is obtained from the response content and the response presence / absence.

According to a second aspect of the present invention, in the address determination method of the network system according to the first aspect, in the address request message broadcasted by the new node when joining the network, the media address and the holding network address which are the unique information of each node. And data to which is added.

According to a third aspect of the present invention, in the address determination method of the network system according to the second aspect, the address server which receives the address request message writes the media address written in a table held by the address server using the media address as a key. The distributed network address stored in the pair is searched, and if the received network address is not used other than the corresponding media address and is unused, this is taken as the data, and if the holding value already exists at the applicable media address, the held value is taken as the data. , If it is not, select an empty address, use this as data, return to the sender as an address request response message containing one of the above as data, and receive the address request message and operate in a distributed system Each node has its own media address and its own And wherein the return to the sender a lifting network address as the address request response message.

According to a fourth aspect of the present invention, in the address determination method of the network system according to the third aspect, the self-maintained network address is included in a higher address of the self and returned to the transmission source.

According to a fifth aspect of the present invention, in the address determination method of the network system according to the third or fourth aspect, a node newly joining the network sets an upper limit of the waiting time from the transmission of the address request message to the reception of the address request response message. The maximum response time is when the maximum number of guaranteed nodes in the subnet is all participating, and when there is no response after this time, the system behaves as a distributed system thereafter.

According to a sixth aspect of the present invention, in the address determination method for a network system according to the fifth aspect, a node newly joining the network receives an address request response message from the address server within the waiting time as an address request response. In some cases, thereafter, it behaves as a server method, and when an address request response message is received from both the address server and the existing node within the waiting time, the error stop is performed.

According to a seventh aspect of the present invention, in the address determination method for the network system according to the sixth aspect, when the new node determines to the distributed system, the new node holds the address request response message received within the upper limit of the waiting time. It is characterized in that a self-holding network address not included in the network address is determined as a temporary address as the first candidate.

According to an eighth aspect of the present invention, in the address determination method for a network system according to the second to seventh aspects, when the address server participates in a network operating in a distributed system, it sends a server notification broadcast message. The existing node that has received the server notification broadcast message sends an address request message having its own media address and network address as data to the server and behaves as a server system.

According to a ninth aspect of the present invention, in the address determination method for a network system according to the first to eighth aspects, each node is a forced server capable of only transmitting an address request message and receiving an address request response message during plug and play. It is characterized by operating in a system.

According to a tenth aspect of the present invention, in the address determination method for a network system according to the ninth aspect, a node newly joining the network sets an upper limit of a waiting time from transmission of an address request message to reception of an address request response message as a subnet. It is the maximum response time when the maximum guaranteed number of nodes is all participating, and if there is no response after this time, when each node is operating with the forced server method, it will be stopped after that as an error And

A network terminal according to the invention of claim 11 is
Both the server method that accepts the network address given by the address server and the distributed method that decides its own network address that can be distinguished from other nodes autonomously participating in the same network when the address server does not exist. It is equipped with an address acquisition function, and broadcasts an address request message at the time of network participation, and the subsequent address acquisition method is autonomous between the server method and the distributed method depending on the response content within a predetermined time and the presence or absence of a response. The address determination information or the address determination information based on the response content and the presence or absence of the response.

According to the twelfth aspect of the invention, in the network terminal of the eleventh aspect, at the plug-and-play time, there is provided a setting means which operates in a forced server system capable of only transmitting the address request message and receiving the address request response message. It is characterized by that.

[0020]

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

3A and 3B are OSI layer diagrams of each node. The model shown in FIG. 9A includes an application layer L1, an upper middleware layer L2, and a TCP / IP layer L.
3, the lower media layer L4. In addition, the model shown in (b) of FIG.
/ IP layer L3 is omitted.

Regardless of how the layers are divided, it is necessary to have an individual address for each layer, but the media address of the lower media layer L4 is previously stored in the network communication device incorporated in the network terminal. It is common to embed it. Also, TCP / IP layer L3
The IP address of is defined in RFC (Request For Comments) as a DHCP server that dynamically assigns an IP address to each client or a method of autonomously acquiring the IP address, and these are adopted by the system. Then, the present invention relates to a technique of acquiring a higher-order address used by the application layer L1.

The upper address is not necessary when handling media defined by a single address system, but in order to handle each node in a unified manner in an application that uses media of various different address systems at the same time. Is mandatory. In addition, when an address space of only a few bytes can be prepared instead of a large address space such as an IP address, the concept of the upper address is useful because the IP address can be apparently used by mapping to the upper address. .

FIG. 4 shows the structure of the upper address, which is Net.
It consists of ID and NodeID. Each of these is a data string of about 1 byte, NetID is a subnet identification code,
NodeID is a holding network address defined as a code for identifying an individual node in the subnet. In the figure, NetID is something like "0x01", and NodeID is something like "0x9E".

Since this NetID is distributed by the routers existing in each subnet after the NodeID is acquired, it will be omitted in this description. MAC_A is No
It is an address (not a so-called MAC address) that is converted into a deID, and MAC_A is a server at the time of startup and autonomously acquired by some method and converted into a NodeID by a predetermined linear conversion. Normally No
The same value is taken for deID and MAC_A.

FIG. 5 shows a format of a packet generated in each layer of FIG. 3, and a processing header is added to each layer. TCP / I from upper middleware layer L2
When passed to the P layer L3, the UDP is used in the TCP / IP layer L3.
A header and an IP header are added and passed to the lower media layer L4. In the lower media layer L4, a lower header is added to this and sent out to the network transmission line.

FIG. 6 shows a network terminal 2 serving as each node.
2 shows the hardware configuration of the CPU 21, the ROM
22, a RAM 23, and a communication media processing block 24. The communication media processing block 24 is a part that exchanges signals with a physical transmission path (including wireless) 25.

The ROM 22 is, for example, a flash ROM,
The RAM 23 is a DRAM, which is a non-volatile memory such as an EEPROM or a hard disk. These ROM2
2. The above NetID and NodeI in the RAM 23
Address data such as D, MAC_A, IP address, and media address are stored. FIG. 7 shows an example of the address table.

Details of the network terminal 2 are as follows. The CPU 21 performs the application layer processing, the upper middleware layer processing, the TCP / IP layer processing, and the lower media layer processing, which have already been described, and acquires the upper self address / DB table and the upper other node address / DB acquired as a result of the processing.
The respective values of the table, the lower self address table, and the lower other node address table are written in the RAM 23 and read as necessary. Alive in RAM23
Address buffer is also set. Further, the value written in the RAM 23 is written in the ROM 22 and read out as necessary.

Lower media layer hardware processing block 24
Receives a packet generated by lower media layer processing in the CPU 21, modulates the packet, and wirelessly, power line, leased line (Ethernet (registered trademark), etc.) at appropriate timing.
And the like on the transmission path 25. Further, the lower media layer hardware processing block 24 appropriately demodulates packets of the wireless, power line, and leased line (such as Ethernet) sent from another node by the transmission path 25, and receives the packets in the lower media layer processing in the CPU 21. hand over.

The address table stored in the ROM 22 and RAM 23 will be described.

<Upper Self Address / DB Table> A set of upper self address and self media address is stored.

<Higher Other Node Address / DB Table>
It stores various data of the device to be communicated, not its own node. A pair of a higher address and a media address of the other party is stored for each communication target device.

<Lower self address table> The self lower MAC_A, IP address and media address required for communication are stored.

<Lower Other Node Address Table> MAC_A, IP of the other party required when the lower media layer communicates
Address and media address are stored.

<Alive Address Buffer> This is a buffer for temporarily pooling the high-order address, media address, and IP address returned in the address request response message by the new startup node or the node that has transmitted the duplicate NodeID. Cleared when processing is complete.

<Higher Address> NetID, N mentioned above
The node ID is a unique value within the subnet, and the NetID is a unique value assigned to each subnet. The upper middleware layer L2 specifies a higher address to identify and operate a node on the network. This value is associated with the media address, the IP address, the NodeID, and the upper address only during a certain indefinite period. That is, the set of the IP address, the NodeID, and the upper address with respect to the media address is a variable value.

<IP address> A so-called internet address. 32 bits for IPv4, 12 for IPv6
It is 8 bits.

<Media Address> This is a unique value of the communication node. In some cases, it refers to a value in combination with a hardware ID indicating the type of media (wireless, Ethernet, power line, etc.). At least, it is a globally unique value for each media type.

Next, the operation of the address data determination method implemented by the network system having the above configuration will be described with reference to the sequence diagram of FIG. The sequence diagram of FIG. 8 shows the server 1 as the default network system.
And the default nodes N1, N2, ..., Nm whose addresses have been acquired
Exists, and a new node Nx exists in this network system.
Shows a procedure for acquiring a network address of a newly-entry node Nx in the case of joining. There are cases where the server 1 exists (1) and cases where it does not exist (2), but in this sequence diagram, both cases are shown at once. Further, the vertical axis of FIG. 8 shows the time transition.

Each of the server 1, the network terminals Ni (i = 1 to m) and Nx serving as nodes holds the address table shown in FIG. 7 as a database.
This address table is held in the server 1 for MAC_A distribution management, and in the general node Ni for node destination table and self address management.

Further, in the network system of this embodiment, the transition to the server / server type operation node / distributed type operation node and the attribution message shown in FIG. 9 are adopted. According to this, the network terminal can perform the operation transition to either the server method or the distributed method by the operation determination trigger, and can automatically select whether or not the server is installed. However, even if the server exists, M
The server system is fixed so as not to operate in a distributed system without receiving an AC_A response. That is, the MAC_A response is set as the response obligation reception message, and only the address request is set as the transmission obligation message. Further, when a server is additionally joined in the network operating in the server system, the new server behaves as a general server Ni in the normal server system if it receives a MAC_A response when sending an address request. . When a server is added to a distributed network, the server is broadcast a server notification so that the existing node Ni operating in the distributed system can receive the notification.
Each returns its own media address and retained MAC_A to the server in the MAC_A request, and the MAC in the MAC_A response
It receives _A and switches to the server system to operate.

<1. Server method> When a new node Nx joins, if there is a media address and MAC_A given when joining the network before, if this is the data in the message shown in FIG. 5, this is code 01 in the table of FIG.
The address request message is broadcast to the subnet (step S1). This address request message is sent to the server 1
To the MAC_A request of code 02 in the table,
If the server 1 participates in the network, the server 1 determines the MAC_A when it receives this message, and the requesting new node Nx transmits the MA of code 03 in the table.
Pass as payload data of C_A response (step S
2).

Here, the server 1 has a database of address lists as shown in FIG. 7, and manages MAC_A using the media address as a key. Then, the server 1 always gives the same MAC_A when receiving the address request message including the same media address. Further, if the value corresponding to the media address in the server 1 is blank among the media address and the MAC_A value in the address request message, a value equal to MAC_A in the address request message is given to the new node Nx as a decision value. .

The new node Nx in the server system acquires MAC_A by this operation, and consequently the NodeID is also determined (step S3). The new entry node Nx is
It behaves as a server system, and does not respond even if it receives an address request and a server notification broadcast described later.

<2. Distributed method> In step S1, even if the new node Nx broadcasts the address request message of FIG. 10 to the subnet, the server 1 does not exist and the MAC_A response has a predetermined time Tmax (timeout time in the sequence diagram of FIG. 8. If it cannot be received during the period (see below), it is considered that the default node Ni is operating in the distributed mode. In this case, the default node Ni that has already started up
Is operating as a distributed system, the address request message is received as the communication rule of the distributed system node, and the address request response of code 04 in FIG. 10 is returned to the request source together with the upper address, media address, and IP address. (Step S4).

Immediately after the subnet broadcast of the address request message, the new node Nx receives the message (1) is an address request response from each predetermined node Ni, or
It is divided into three cases: (2) MAC_A response sent from the server 1 or (3) no received data.
It should be one of them and saves the data returned from the server 1 or the default node Ni in a buffer.
If the MAC_A response is not returned by the time Tmax, the time-up event is used as a trigger to determine the distributed method.

When there is no address request response, the new node Nx, if there is a MAC_A self-holding value,
If there is no holding value, the lower 1 byte of the self-media address is used, and if there is an address request response, the self-holding MAC_A is stored in the vacant MAC_A from all MAC_A obtained from the inverse conversion value of the received NodeID.
If there is a value, select that value. If there is no hold value, select this if the lower 1 byte of the self media address is included in the empty address. If not, select it randomly from the empty addresses. , Temporary MAC_A is determined (step S5). Then, the new node Nx notifies the data including the temporary MAC_A as the node confirmation of the code 05 in the table of FIG.
6).

Received default node N in operation in the distributed system
Each i has its own MAC_A and the received temporary MAC_A
Are compared, and if they match, the node confirmation response of code 06 in the table is returned to the new node Nx (step S
7). The new node Nx that received the node confirmation response
This is excluded, the vacancy is checked, the temporary MAC_A is determined again, and the notification is sent by the broadcast in the subnet. This is repeated a predetermined number of times, it is confirmed that the node confirmation response is not returned, the provisional value is set to the official MAC_A, and it is converted to the NodeID (step S8).

If it is determined that there are no free addresses, error stop is determined. If a node confirmation message is received from another node between the address request transmission in step S1 and the node confirmation transmission in step S6, the received temporary MAC_A is regarded as used and is excluded from its own selection candidates. Further, the time from the address request broadcast in step S1 until the waiting of the address request response is stopped (in the case of the distributed system, immediately before the provisional address is determined) is the above-mentioned maximum response time Tmax, and the maximum number of nodes is connected to the subnet. Set to the maximum response time of all nodes in the event of a breach. And if this time is exceeded, the temporary MAC is used as a distributed method.
_A is decided and a node confirmation message is transmitted.

When the server 1 is added later to each of the default nodes Ni operating in the distributed mode and the server notification broadcast of the code 07 issued by the server 1 is received in step S10,
A MAC_A request including the media address and MAC_A is sent to the server 1 (step S11), the MAC_A is distributed from the server 1 in the MAC_A response, and then the server system is switched to operate (step S12).

The new node Nx is MAC_A
Is acquired by the server method or the distributed method, the NetID is requested to the router, the NetID is acquired from the router, and the upper address is determined.

As described above, in the address determination method of the network system according to the present embodiment, the holding self MAC_A of the node Ni operating in the distributed system included in the MAC_A request in step S11 is not used in the server 1 (the new server Nx is Of course, it is unreserved and all unused.)
If so, this will be added, so even if the system switches from the distributed system to the server system, MAC_A, that is, No.
Since the deID, that is, the upper address does not change, the network is not interrupted.

Further, by using the same network address system for the server system and the distributed system, when switching between the server system and the distributed system, it is possible to change the network address and smoothly switch the address distribution system. Further, the number of used nodes is equal to the address space, and in the conventional case, address duplication can be reduced even under the condition that address conflict frequently occurs.

Next, an address determining method for the network system according to the second embodiment of this invention will be described with reference to the sequence diagram of FIG. The second embodiment is
When the network terminal Nx serving as a node newly joins the network, an address request is broadcast (step S
1) If the server 1 returns a server response (step S1)
A), the node Nx sends a MAC_A request to the server 1 (step S1B), and if it receives a MAC_A response, sends it to M
It is characterized in that it is determined to be AC_A (step S2).

In response to the address request broadcast from the new node Nx, if there is no response from the server within the maximum standby time Tmax because the server 1 does not exist, it is considered to be operating in the distributed system, and the Although the address determination is performed, the address determination procedure in this distributed system is the same as that of the first embodiment shown in the sequence diagram of FIG.

In the case of the second embodiment, the message type is as shown in FIG. 12, and the network terminal Ni
The address request message that each broadcasts first also serves as the server search message. Therefore, as shown in FIG. 12, the server response message from the server 1 is added to the message type of FIG. 10 of the first embodiment. According to the second embodiment, there is an advantage that the server participation in the network can be determined more surely and the address distribution can be made more robust.

In the above embodiment, the specified node transmits the entire upper address in response to the address request message from the new node. However, in this upper address, the own network address No.
The configuration may be such that only the deID is transmitted. in this case,
Since the NodeID has a one-to-one correspondence between the NodeID and the MAC_A because the MAC_A is a linearly converted code, the MAC_A that can be regarded as its own holding network address may be transmitted instead of the NodeID.

[0059]

As described above, according to the present invention, the new entrant node automatically selects either the server method or the distributed method as the address determination method according to the presence or absence of the server, and whichever method is selected. You can also get the network address with this method.

In the present invention, the network address system is the same between the server system and the distributed system, and when switching between the server system and the distributed system, the network address is not changed and the address distribution system can be smoothly switched.

[Brief description of drawings]

FIG. 1 is a block diagram of a general server-type network system.

FIG. 2 is a block diagram of a general distributed network system.

FIG. 3 is a diagram of the network terminal O according to the embodiment of the present invention.
Explanatory drawing of SI layer model.

FIG. 4 is an explanatory diagram showing a configuration of upper address data used in the above embodiment.

FIG. 5 is an explanatory diagram showing a data structure of each layer of a message transmitted and received in the above embodiment.

FIG. 6 is a block diagram showing a hardware configuration of the network terminal according to the above embodiment.

FIG. 7 is an explanatory diagram of an example of an address table used in the above embodiment.

FIG. 8 is a sequence diagram of an address determination procedure according to the above embodiment.

FIG. 9 is a diagram illustrating a server system operation of each of the server and the node according to the above-described embodiment, determination of a distributed system operation and determination of transition
A table showing the response obligation reception message and the transmission obligation message at initialization.

FIG. 10 is a table showing the types of messages used in the above embodiments and the address data contained therein.

FIG. 11 is a sequence diagram of an address determination procedure according to the second embodiment of the present invention.

FIG. 12 is a table of message types used in the above embodiment and address data included therein.

[Explanation of symbols]

1 server 2 network terminals 21 CPU 22 ROM 23 RAM 24 Lower Media Layer Processing Block 25 transmission lines

Claims (12)

[Claims] 1. A network system composed of a plurality of nodes, comprising an address acquisition function of a server system that accepts a network address assigned from an address server and a distributed system in which each network participating node autonomously determines a network address. Each node is installed at the same time, each node broadcasts an address request message when participating in the network, and the subsequent address acquisition method is selected from the server method and the distributed method depending on the response content within a predetermined time and the presence or absence of response. First, an address determining method for a network system, which autonomously determines and obtains an address or address determining information from the response content and the presence or absence of the response. 2. The address request message broadcasted by a new node when participating in a network includes a media address, which is unique information of each node, and data to which a holding network address is added. Method of network system address determination. 3. The address server receiving the address request message searches for the distributed network address held in pair with the media address written in the table held by itself, using the media address as a key,
If the received network address is other than the corresponding media address and is not used, this is used as the data, if the holding value already exists at the corresponding media address, the holding value is used as the data, and if not, an empty address is selected and used as the data. , Each of the existing nodes operating in the distributed system that returns the address request response message including one of the above as data to the transmission source and receives the address request message has its own media address and its own holding network address. Is returned to the transmission source as an address request response message.
The method for determining the address of the network system described in. 4. The address determination method for a network system according to claim 3, wherein the own-held network address is included in an upper address of the own and returned to the transmission source. 5. A node newly joining the network sets the upper limit of the waiting time from the transmission of the address request message to the reception of the address request response message as the maximum response time when all the maximum guaranteed nodes in the subnet are participating. 5. The address determination method for a network system according to claim 3, wherein when there is no response after this time, the system behaves as a distributed system. 6. When a node newly joining the network receives an address request response message from the address server within the waiting time as an address request response, it behaves as a server system thereafter and becomes an address server within the waiting time. 6. The address determination method for a network system according to claim 5, wherein when the address request response message is received from both the existing node, the error stop is performed. 7. The new node tentatively determines, as a first candidate, a self-holding network address that is not included in the holding network address included in the address request response message received within the upper limit of the waiting time when the distribution method is determined. The address determination method for a network system according to claim 6, wherein the address is determined. 8. The address server sends a server notification broadcast message when participating in a network operating in a distributed mode, and an existing node receiving the server notification broadcast message retains its own node. 8. An address determining method for a network system according to claim 2, wherein an address request message having the media address and the network address as the data is sent to the server to act as a server system. 9. The node according to claim 1, wherein each node operates in a forced server system capable of only transmitting an address request message and receiving an address request response message during plug and play. A method for determining the address of the described network system. 10. A node newly joining the network sets the upper limit of the waiting time from the transmission of the address request message to the reception of the address request response message as the maximum response time when all the maximum guaranteed nodes in the subnet are participating. 10. The method for determining an address of a network system according to claim 9, wherein when there is no response after this time, when each node is operating by the forced server method, an error stop is performed thereafter. 11. A server method for accepting a network address given from an address server, and a distribution for determining its own network address that can be distinguished from other nodes autonomously participating in the same network when the address server does not exist. It is equipped with an address acquisition function of both the system and the system, broadcasts an address request message at the time of network participation, and depending on the response content within a predetermined time and the presence or absence of a response, the subsequent address acquisition methods are the server method and the distributed method. A network terminal having a function of autonomously making an alternative decision and obtaining an address decision or address decision information from the response content and the presence or absence of a response. 12. The network terminal according to claim 11, further comprising setting means that operates in a forced server system capable of only transmitting an address request message and receiving an address request response message during plug and play. .