JP2009044371A - Overlay network forming method, overlay node, overlay network, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To solve the problem that an overlay network is configured by full-mesh topology in conventional technique and it is difficult to configure an overlay network when the total number of overlay nodes is large. <P>SOLUTION: Some overlay nodes meeting predetermined conditions among overlay nodes constituting an overlay network are logically connected as supernodes to form a supernode layer, and other general nodes are logically connected to the nearest supernodes to be subordinate thereto to form a layer wherein the general node group is present as a general node layer, thus forming a two-storied layer. Further, the respective nodes are given two-dimensional coordinate positions, and a two-dimensional space including each supernode in the center is divided into six sectors each having an angle range θ of π/3; and a supernode closest to the supernode among other supernodes positioned in a space of a sector j (j=1 to 6) is logically connected as an adjacent supernode to constitute the supernode layer. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a technique for improving end-to-end communication quality in an IP (Internet Protocol) network, and more particularly to a communication path using an overlay network logically formed on an IP network. The present invention relates to a technique suitable for efficiently improving end communication quality.

  The Internet, which represents IP networks, has been developed and spread to accommodate various applications. Recently, it is sensitive to VoIP (Voice over IP) and QoS (Quality of Service) typified by streaming. The containment of time applications etc. is also developing rapidly.

  Along with this, it is an important issue to realize technology ("end-to-end QoS management technology") on the Internet to avoid end-to-end congestion and improve quality. . However, there are the following problems in realizing such a technique.

(1) The Internet has already become a social infrastructure, and it is difficult to expand new functions at the network layer that greatly change the existing network structure.
(2) The Internet is formed by a plurality of ASs (Autonomous Systems) having different management entities, and it is difficult to extend new functions to all ASs simultaneously.

  Under such circumstances, as an effective technique that can improve end-to-end QoS without changing a lower network layer, a QoS management technique using an overlay network described in Non-Patent Document 1, for example, has attracted attention.

  An overlay network is, for example, a network configured by forming a logical (virtual) link according to the purpose in an upper layer using an existing link as described in Non-Patent Document 2, for example. .

  The basic concept of QoS management by such an overlay network is illustrated in FIG. In FIG. 1, it is assumed that traffic flows from x to y along a route of IP router 2a → IP router 2d → IP router 2e. Further, it is assumed that the route between the IP routers 2d and 2e is congested, and as a result, the QoS between x and y is lowered.

  At this time, using the overlay network 1 formed by the overlay nodes 1a, 1b, and 1c, the traffic is detoured along the route of x → IP router 2a → overlay node 1a → overlay node 1b → overlay node 1c → IP router 2e → y. If this can be achieved, the above congestion can be avoided.

  In fact, Non-Patent Documents 3, 4, and 5 show that there are a number of such detour routes in the real network based on actual measurements. However, the technologies of Non-Patent Document 3 and Non-Patent Document 5 evaluate the case where the topology of the overlay network is a full mesh and ideal communication path calculation is performed using quality information measured between all overlay nodes. It has become. Non-Patent Document 4 evaluates the case where the number of relay node candidates that provide a detour route is limited, but this also uses the topology of the overlay network as a full mesh and requires quality information between all overlay nodes. . In other words, the techniques of Non-Patent Document 3 to Non-Patent Document 5 do not take into consideration a decrease in scalability (system expandability) when the total number of overlay nodes increases.

  That is, when the total number of overlay nodes is large, there is a problem that it is difficult to configure a network with a full mesh from the viewpoint of scalability.

L. Zhi and P.M. Mohapra, “QRON: QoS-aware routing in overlay net-works,” IEEE J. MoI. Select. Areas Commun. , Vol. 22, pp. 29-40, January 2004. WIDE Project, “Integrated Distributed Environment with Overlay Network,” WIDE Project Research Report, Part 17, 2002. Kamei, Kawahara, “Traffic engineering by end-host overlay network and its effectiveness,” Shingaku Sodai, BS-5-3, 2004. S. Rewaskar and J.H. Kaur, “Testing the Scalability of Overlay Infrastructures,” Proc. PAM2004. April 2004. S. Banerjee, T .; G. Grifin and M.M. Pias, “The Interdomain Connectivity of PlanetLab Nodes,” Proc. PAM 2004, April 2004.

  The problem to be solved is that in the conventional technique, when the total number of overlay nodes is large, it is difficult to configure an overlay network with a full mesh from the viewpoint of scalability.

  The object of the present invention is to solve these problems of the prior art and to form a scalable overlay network.

  In order to achieve the above object, according to the present invention, in an overlay network, which is a logical network constructed by several overlay nodes connected to an IP network such as the Internet, among the overlay nodes, for example, sufficient resources (bandwidth, Some nodes that do not frequently join or leave the network are super nodes, and the super nodes are logically connected to form a super node network layer, and other overlay nodes are generally used. For example, each general node belongs to a super node having the closest distance from itself (round-trip delay time RTT: Round-Trip-Time, etc.) and logically connects the super node and the general node. The hierarchy in which the node group exists is It formed as a work layer, on the overlay network, and forming a second layer of the super node network layer and general node network layer. Each node is mapped to a two-dimensional geometric coordinate space, that is, each node is given a two-dimensional coordinate position, and each super node has an angle θ = π / 3 divides the space into 6 sectors, and logically connects the super node closest to itself among other super nodes located in the space of sector j (j = 1 to 6) as an adjacent super node. To form a super node layer.

  According to the present invention, it is possible to configure a scalable overlay network without configuring a network with all meshes between all nodes, and traffic engineering that can improve QoS is possible.

  The best mode for carrying out the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a network configuration example in which an overlay network is provided on an IP network, and FIG. 2 is an explanatory diagram showing a network configuration example in which a super node layer according to the present invention is formed on the overlay network in FIG. 3 is an explanatory diagram showing an example of a route selection procedure in the super node layer in FIG. 2, FIG. 4 is an explanatory diagram showing an example of a route control procedure in the super node layer in FIG. 2, and FIG. 6 is a block diagram showing a configuration example of a super node forming the super node layer according to FIG. 6, FIG. 6 is a block diagram showing a configuration example of a general node forming the general node layer according to the present invention, and FIG. FIG. 8 is an explanatory diagram showing a first example of an adjacent super node update procedure in the node layer, and FIG. 8 shows an adjacent super node in the super node layer in FIG. Explanatory view showing a second example of the update procedure, Fig. 9 is a flowchart showing a forming procedure example of an overlay network according to the present invention.

  Each of the overlay nodes 1a, 1b, and 1c in FIG. 1 has a computer configuration including a CPU (Central Processing Unit), a main memory, a display device, an input device, and an external storage device, and a CD-ROM via an optical disk drive device or the like. After the programs and data recorded in the storage medium such as the above are installed in the external storage device, they are read from the external storage device into the main memory and processed by the CPU, thereby the super node 50 in FIG. 5 and the general node 60 in FIG. It becomes the structure which comprised each processing means shown by these.

  For example, when functioning as the super node 50 in FIG. 5, processing by the coordinate management unit 50a, the adjacent node determination unit 50b, the packet transfer unit 50c, and the general node management unit 50d is executed as a programmed computer processing function. When functioning as the general node 60 in FIG. 6, processing by the super node selection unit 60a, coordinate holding unit 60b, shortcut route determination unit 60c, and packet transfer unit 60d is executed as a programmed computer processing function.

  As shown in FIG. 1, in an IP network (described as “IP-NW” in the figure) 3 composed of IP routers 2a to 2e, traffic is routed from x to y along the routes of IP routers 2a, 2d, and 2e. When congestion occurs between the IP routers 2d and 2e on this route, the QoS between x and y decreases as a result.

  At this time, using the overlay network (described as “Overlay-NW” in the figure) 1 formed by overlay nodes (hereinafter also simply referred to as nodes) 1a, 1b, 1c, x → IP router 2a → overlay node a The above-described congestion can be avoided by detouring traffic through the route of overlay node c → IP router 2e → y.

  However, at this time, the delay time when the data is directly transferred from the overlay node 1a to the overlay node 1c is temporarily transferred from the overlay node 1a to the overlay node 1b and then transferred from the overlay node 1b to the overlay node 1c. There are cases where the delay time is smaller. This is the case, for example, due to an interdomain routing policy on the Internet.

  In this case, if the traffic can be detoured to the route (overlay node 1a → overlay node 1b → overlay node 1c) represented by the solid line arrow using the overlay network formed by the overlay nodes 1a, 1b, and 1c, the QoS can be reduced. Can be improved.

  Therefore, in this embodiment, as illustrated in FIG. 2, in such an overlay network, several overlay nodes are selected as super nodes among the overlay nodes based on a predetermined selection condition. The super node forms a super node network layer (in the figure and hereinafter referred to as “super node layer”) 21 by logically connecting each super node, and other overlay nodes that do not meet the selection conditions are generally used. Each general node is logically connected to any one of the super nodes to form a general node network layer (in the figure and hereinafter referred to as “general node layer”) 22, and the overlay network is used as the super node layer 21. And the general node layer 22.

  Here, as conditions for deciding which overlay node to select as a super node, for example, there are sufficient resources (bandwidth, CPU performance, etc.), and frequent participation and withdrawal from the network A condition for selecting a super node is not to be performed.

  Such a super node may be selected by an administrator of this network, and when each overlay node connects to the overlay network, it selects whether or not it becomes a super node. If it becomes a super node, this may be notified to each already connected overlay node.

  In addition, the general node identifies a super node having the best communication efficiency including the shortest RTT (Round-Trip-Time) from itself among a plurality of super nodes, and logically identifies the identified super node. Connect and belong to the super node. In addition, instead of the shortest RTT, a one-way delay time or a packet loss rate mapped to a distance (a function that increases the distance if the packet loss rate is large is prepared in advance to identify the super node with the best communication efficiency. Alternatively, the distance obtained from the coordinate position by the processing operation of the super node selection unit 60a in FIG. 6 described later may be used.

  Further, in the overlay network, each overlay node is mapped to a two-dimensional geometric coordinate space, that is, each overlay node is given a two-dimensional coordinate position and each overlay node stores its own two-dimensional coordinate position. In particular, each super node reads out its own two-dimensional coordinate position information stored therein, divides the two-dimensional space into the six sectors at an angle θ = π / 3 around itself, and sets sector j (j = 1-6) from the other super nodes located in the space, to obtain the two-dimensional coordinate position information of the super node, specify the super node closest to itself as the adjacent super node in the sector j, Logically connect to the specified adjacent super node.

  That is, as shown in FIG. 2, the super node layer 21 is divided into six sectors around each super node. An adjacent network configured in this way is called a Yao graph, and it is known that a spanning tree (branching tree) having the shortest path in a geometric space can be configured.

  When performing connection between nodes in the super node layer 21, each super node is triggered by the occurrence of an event such as a certain period or when an event such as a new super node joins or an existing super node leaves. Comparing the distance between the current adjacent supernode in each sector j centered on itself and other supernodes in sector j, and if the other supernode is closer to the current adjacent supernode. If it exists, the nearest super node is updated as a new adjacent super node.

  Each general node notifies its logical two-dimensional coordinate position information stored in the storage device in advance to the logically connected super node, and acquires the two-dimensional coordinate position information of this super node, Calculates the distance (first distance) between itself and the super node using its own two-dimensional coordinate position information, and other general nodes connected to the same super node from the logically connected super node The two-dimensional coordinate position information of the node is acquired, and the other general nodes and the supermarket are obtained using the acquired two-dimensional coordinate position information of the other general nodes and the two-dimensional coordinate position information of the supernode. The distance between the nodes (second distance) is calculated, and two-dimensional coordinate position information of other general nodes connected to the same super node is obtained from the logically connected super nodes. Then, using the acquired two-dimensional coordinate position information of the other general node and the own two-dimensional coordinate position information, a distance (third distance) from each other general node is calculated, and The calculated first distance and the second distance are added, the added value is compared with the third distance, and the value obtained by adding the first distance and the second distance is larger than the third distance. If there is another general node, a shortcut path is created by logically connecting the other general node and itself. Thus, in the general node layer 22, the general node i_x belonging to a certain super node x measures the distance d (i_x, j_x) from another general node j_x belonging to the same super node x, and the value d The value obtained by adding (i_x, x) (distance between the general node i_x and the super node x) and d (j_x, x) (distance between the general node j_x and the super node x) is compared, and if “d (i_x, j_x ) <D (i_x, x) + d (j_x, x) ”, a shortcut path is created by logically connecting the general node i_x and the general node j_x.

  In such a configuration, when a general node x_i belonging to a certain super node x performs packet communication with a general node y_j connected to another super node y, the general node x_i first sends a packet to the super node x. The super node x that receives the packet transfers the packet to the super node y through the super node layer, and the super node y sends the packet to the general node y_j.

  That is, as shown in FIG. 3, when a packet is transferred from the general node (1) to the general node (4), the path is ADF. For the route D here, the super node (1) and the super node (2) do not necessarily have to be directly logically connected, and there may be several hops between them. The relay technique at that time will be described later with reference to FIG.

  Also for the above-described shortcut route, as shown in FIG. 3, the general node (1) passes through the shortcut route of the route C from the general node (2). In addition, the route from the general node (3) to the general node (4) passes through the route (EF) because the route via the super node (2) is closer than the shortcut.

  In such a super node layer, it is assumed that super node x is a source super node and super node y is a destination super node, and a packet transferred from super node x to super node y is now as shown in FIG. , The relay super node u has a line u connecting its own coordinate u and the coordinate y of the destination super node in the coordinate space among the adjacent super nodes v_1, v_2,. One adjacent super node v_j with the smallest angle v_j-u-y and one adjacent super node v_k with the smallest angle yu-v_k are searched one by one from both sides across -y, and the neighboring super nodes are randomly selected. The packet is transferred with either v_j or the adjacent super node v_k as the next hop node.

  In the following, the configuration and processing operation of the super node 50 and the general node 60 that form an overlay network composed of two layers of the super node layer and the general node layer will be described with reference to FIGS.

  As shown in FIG. 5, the super node 50 includes a coordinate management unit 50a, an adjacent node determination unit 50b, a packet transfer unit 50c, and a general node management unit 50d as programmed computer processing functions. , Holding its own coordinate position in a storage device (not shown), inquiring the coordinate position to other super nodes currently participating in the network, obtaining and holding the information, and storing each super node The coordinate information is notified to the adjacent node determination unit 50b.

  Upon receiving the notification, the adjacent node determination unit 50b divides the two-dimensional space into the six sectors at an angle θ = π / 3 centered on itself, and is positioned in the space of the sector j (j = 1 to 6). Each coordinate information is acquired from each super node, a distance calculated from each coordinate information is compared, a super node closest to itself is selected as an adjacent super node, and logically connected.

  The adjacent node determination unit 50b updates the selection of the adjacent super node according to the following procedure. That is, the distance between the current adjacent supernode in each sector j centered on itself and the other supernodes in the sector j is compared, and if another supernode is closer to the current adjacent supernode If it exists, the nearest super node is updated as a new adjacent super node.

  This operation is performed at regular intervals or every time an event is detected (joining a new super node, leaving an existing super node, etc.). This will be described with reference to FIGS.

  FIG. 7 shows an example of adjacent super node selection update processing operation when a new super node participates. It is assumed that super nodes N1, N2, and N3 already exist, and N1, N2, and N3 are solid lines. Assume that they are logically connected as shown.

  Here, consider a case where a logical link between N1 and N2 is disconnected and a new super node L is simultaneously participating. In this case, the adjacent node table of each super node is updated as shown in FIG.

  For example, N1's adjacent supernodes are L and N3. In FIG. 7, both in and out are described, but this means that there is a table for each direction. In both in and out, L and N3 are newly entered as adjacent supernodes. Yes.

  FIG. 8 shows an example of the adjacent super node selection update processing operation when the existing super node L leaves. In this case, the adjacent node table is updated in the opposite manner to the example shown in FIG. Become.

  The general node management unit 50d in the super node 50 in FIG. 5 is logically connected to itself, and acquires and manages the addresses and coordinate positions of the general nodes belonging to the subnode.

  When the packet transfer unit 50c in the super node 50 in FIG. 5 receives a packet from its own general node or from an adjacent super node, it reads the destination of the packet, and if it is for another general node under its own Referring to the address and coordinate position of a general node belonging to its own managed by the general node management unit 50d, the packet is sent to the node, and if the packet is for another super node, the following procedure is performed. To forward the packet.

  Assuming that the packet to be transferred to the super node y is currently at its own location (assumed to be a super node u), the super node u has the own super node v_1, v_2,. The adjacent super node v_j having the smallest angle v_j-u-y and the angle yu-v_k being the smallest, one each from both sides of the line u-y connecting the coordinate u and the coordinate y of the destination super node The adjacent super node v_k is searched, and the packet is randomly transferred with either the adjacent super node v_j or the adjacent super node v_k as the next hop node.

  Next, the operation of the general node will be described with reference to FIG. The general node 60 holds its own coordinate position in the coordinate management unit 60a. At the time of joining the network, for example, the super node selection unit 60a obtains the coordinate position of each super node from each super node, and has the closest distance compared to its own coordinate position held by the coordinate management unit 60a. A super node is searched and selected as a super node to which the super node belongs, and this is notified to the general node management unit 50d of the selected super node.

  Further, the shortcut route determination unit 60c reads the coordinate position of another general node belonging to the super node x from the super node x to which the self (general node i_x) logically connects and belongs as a subordinate, The third distance d (i_x, j_x) between the general node i_x) and the other general node j_x is calculated, and the value “d (i_x, j_x)” and the self (general node i_x) and the super node x The value obtained by adding the second distance “d (j_x, x)” between the other general node j_x and the super node x to the first distance “d (i_x, x)” is compared. , J_x) <d (i_x, x) + d (j_x, x) ”, and if the value obtained by adding the first distance and the second distance is larger than the third distance, self (general node i_x) And general node j_x are logically connected to To create a route of Tokatto.

  The packet transfer unit 60d checks whether or not the destination is under the same super node when sending the packet, and if so, inquires of the shortcut route determination unit 60c to determine whether or not there is a shortcut route. If there is a shortcut route, the packet is sent directly to the destination. If there is no shortcut route, the packet is transferred to the super node to which it belongs. Also, when a packet is transmitted to a general node belonging to another super node, the packet is transferred to the super node belonging to itself.

  Next, a procedure for forming an overlay network including such a super node layer and a general node layer will be described with reference to FIG.

  First, a super node is selected from each overlay node constituting the overlay network (step S901). Here, for example, some nodes that have sufficient resources (bandwidth, CPU, etc.) and do not frequently participate in or leave the network are selected as super nodes. The network administrator may select it, and when each overlay node connects to the network, it chooses whether or not it will become a super node. It may be notified to the existing node.

  Next, each node is mapped to a two-dimensional geometric coordinate space, that is, each overlay node is given and held with a two-dimensional coordinate position, and each super node has an angle in the two-dimensional space around itself. The space is divided into six sectors with θ = π / 3, the coordinate position of the super node is obtained from other super nodes located in the space of sector j (j = 1 to 6), The super node closest to itself is calculated as the adjacent super node and logically connected to form a super node layer (step S902).

  Then, the other overlay node is identified as a general node, for example, a super node having the shortest RTT with itself as a super node having the shortest distance from itself, and logically connected to the identified super node, A general node layer is formed (step S903). In this way, the overlay network is formed in two layers of the super node layer and the general node layer.

  As described above with reference to FIGS. 1 to 9, in this example, the overlay network is formed in two layers of the super node layer and the general node layer. As a result, a scalable overlay network can be configured without configuring a full mesh network between all nodes, and traffic engineering that can improve QoS is possible.

  In addition, this invention is not limited to the example demonstrated using FIGS. 1-9, In the range which does not deviate from the summary, various changes are possible. For example, in this example, each processing unit in the super node 50 (coordinate management unit 50a, adjacent node determination unit 50b, packet transfer unit 50c, general node management unit 50d) and each processing unit in the general node 60 (super node selection unit 60a). The coordinate holding unit 60b, the shortcut route determination unit 60c, and the packet transfer unit 60d) are configured by a computer that includes a CPU and a storage device and can execute each function by a program. Some may have a hardware configuration including logic element circuits.

  Also, regarding the computer configuration example in this example, a computer configuration without a keyboard or optical disk drive may be used. In this example, an optical disk is used as a recording medium. However, an FD (Flexible Disk) or the like may be used as a recording medium. As for the program installation, the program may be downloaded and installed via a network via a communication device.

It is a block diagram which shows the network structural example which provided the overlay network on the IP network. It is explanatory drawing which shows the network structural example which formed the super node layer which concerns on this invention on the overlay network in FIG. It is explanatory drawing which shows the example of a route selection procedure in the super node layer in FIG. It is explanatory drawing which shows the example of a path control procedure in the super node layer in FIG. It is a block diagram which shows the structural example of the super node which forms the super node layer which concerns on this invention. It is a block diagram which shows the structural example of the general node which forms the general node layer which concerns on this invention. FIG. 3 is an explanatory diagram showing a first example of an adjacent super node update procedure in the super node layer in FIG. 2. It is explanatory drawing which shows the 2nd example of the adjacent super node update procedure in the super node layer in FIG. It is a flowchart which shows the example of a formation procedure of the overlay network which concerns on this invention.

Explanation of symbols

  1: Overlay network (Overlay-NW), 1a-1c: Overlay node, 2a-2e: IP router, 3: IP network (IP-NW), 21: Super node layer, 22: General node layer, 50: Super node 50a: Coordinate management unit, 50b: Adjacent node determination unit, 50c: Packet transfer unit, 50d: General node management unit, 60: General node, 60a: Super node selection unit, 60b: Coordinate holding unit, 60c: Shortcut path determination Part, 60d: Packet forwarding part.

Claims (10)

A method of forming an overlay network, which is a logical network composed of a plurality of overlay nodes connected to an IP network,
Among the overlay nodes constituting the overlay network, a plurality of super nodes selected under a predetermined condition form a super node network layer by performing a logical connection between the super nodes.
Each general node that does not meet the above conditions is logically connected to any one of the super nodes to form a general node network layer,
An overlay network forming method comprising: forming two layers of the general node network layer and the super node network layer. The overlay network forming method according to claim 1,
An overlay network forming method, wherein the general node specifies a super node having the best communication efficiency including the shortest RTT among the plurality of super nodes and logically connects to the specified super node. An overlay network forming method according to claim 1 or 2,
Each super node reads out its own two-dimensional coordinate position information stored in advance in a storage device,
Divide the space into 6 sectors at an angle θ = π / 3,
The two-dimensional coordinate position information of the super node is obtained from other super nodes located in the space of the sector j (j = 1 to 6), and the super node having the closest coordinate position is specified as the adjacent super node in the sector j And an overlay network forming method characterized by logically connecting to the specified adjacent super node. The overlay network forming method according to claim 3,
Each super node updates an adjacent super node connected in sector j at every fixed period or triggered by the occurrence of an event including the participation of a new super node or the detachment of an existing super node. . An overlay network forming method according to claim 3 or 4, wherein:
The super node is
When relaying and forwarding the packet forwarded from the source super node to the destination super node,
One of its own coordinates u and the coordinates of the adjacent super node, one on each side of the line yy connecting the coordinate u of its own and the coordinate y of the destination super node in the coordinate space. identifying the first and second adjacent supernodes having the smallest angle formed by the line u-vj connecting vj and the line u-y,
A method for forming an overlay network, comprising: transferring a packet by randomly using either one of the identified first and second adjacent super nodes as a next hop node. An overlay network forming method according to any one of claims 1 to 5,
Each general node notifies its logical two-dimensional coordinate position information stored in the storage device in advance to the logically connected super node, and acquires the two-dimensional coordinate position information of the super node. Using the two-dimensional coordinate position information, calculate the distance (first distance) between itself and the super node,
The two-dimensional coordinate position information of another general node connected to the same super node is acquired from the logically connected super node, and the acquired two-dimensional coordinate position information of the other general node and 2 of the super node are acquired. Using the coordinate position information of the dimension, the distance (second distance) between each other general node and the super node is calculated,
The two-dimensional coordinate position information of other general nodes connected to the same super node is acquired from the logically connected super node, and the acquired two-dimensional coordinate position information of the other general nodes is stored in the storage device. Using its own two-dimensional coordinate position information, the distance (third distance) to each other general node is calculated,
Add the calculated first distance and second distance, compare the added value with the third distance,
If there is another general node in which the value obtained by adding the first distance and the second distance is larger than the third distance, the other general node is logically connected to itself and the shortcut path is connected. An overlay network forming method comprising: creating an overlay network. An overlay network forming method according to any one of claims 1 to 6,
When the above general node performs packet communication with a general node connected to another super node, it sends a packet to the super node to which it is connected,
Transfer the packet through the super node network layer from the super node that received the packet to the super node connected to the general node of the communication destination,
A method for forming an overlay network, wherein a super node connected to a general node of a communication destination transmits a packet to the general node of the communication destination. A plurality of overlay nodes forming an overlay network on the IP network,
The overlay comprising the means for forming the super node layer or the general node layer as either a super node or a general node in the overlay network forming method according to any one of claims 1 to 7. node. An overlay network, which is a logical network composed of a plurality of overlay nodes connected to an IP network,
The overlay network according to claim 8, wherein the overlay node is formed from two layers of the super node layer and the general node layer. The program for making a computer perform the process as either a super node or a general node in the overlay network formation method in any one of Claims 1-7.

Images