CN113810287B

CN113810287B - Data retrieval and pushing method based on NDN and SDN

Info

Publication number: CN113810287B
Application number: CN202111062292.8A
Authority: CN
Inventors: 郭柳; 安建平; 李彤; 张宇; 薛艳明; 刁文澜; 李梦珂
Original assignee: Beijing Institute of Technology BIT
Current assignee: Beijing Institute of Technology BIT
Priority date: 2021-09-10
Filing date: 2021-09-10
Publication date: 2022-07-01
Anticipated expiration: 2041-09-10
Also published as: CN113810287A

Abstract

The invention discloses a data retrieval and pushing method based on NDN and SDN, and belongs to the technical field of computer networks. The method fuses an original distributed NDN architecture and an SDN architecture, introduces a central controller, realizes push-based data packet sending in a global view through the central controller and a source routing technology, removes PIT and FIB tables, thoroughly liberates the performance of the router, reduces the flow in a network, and greatly reduces the processing delay of the router; the router timely informs the content of the change of the CS to the central controller, the central controller selects the routing node with the lowest cost away from the consumer, the content notice is flexibly and effectively distributed, the CS is utilized to the maximum extent, and the data acquisition delay is reduced. The invention can be applied to the field of information retrieval and the like, and solves the technical defects of long matching search processing time delay in FIB and incapability of flexibly and effectively distributing content announcement of the traditional NDN architecture.

Description

Data retrieval and pushing method based on NDN and SDN

Technical Field

The invention relates to a Data retrieval and pushing method based on a Named Data Networking (NDN) and a Software-Defined Networking (SDN), and belongs to the technical field of computer networks.

Background

The Named Data Networking (NDN) takes Data content as a center, is different from an IP address-based addressing mode of the existing TCP/IP network, sets a content name as an addressing mode, establishes a routing table for communication based on the content name, and decouples the relationship between the Data content and the position. Compared with a server-centric communication mode, in the NDN architecture, data content is independent of location, and thus is very suitable for data retrieval.

The transmission mechanism of the NDN network is used for carrying out data communication in a mode that a user sends an interest packet to drive a content producer to reply a data packet, and the interest packet and the data packet both contain a content name to identify data content which needs to be acquired by the user. The router of the NDN needs to maintain 3 tables: content store CS (content store), pending Interest table PIT (pending Interest Table), and forwarding Information table FIB (forwarding Information base). When the router forwards the interest packet from the user, firstly, whether the content matched with the data name carried by the current interest packet exists is searched in the CS, and if the content matched with the data name carried by the current interest packet exists, the matched data packet is directly returned to the node initiating the request; if not, the router records an inflow interface (downstream interface) of the interest packet corresponding to the name of the current request data in the PIT, and meanwhile, the router searches a record matched with the current name in the FIB table to determine a forwarding interface (upstream interface). When a certain node has a data packet matching with the name, the data packet is returned as an original route according to the information of an inflow interface (a downstream interface) recorded by the PIT.

The current NDN architecture has the following defects in the aspect of data retrieval:

(1) in terms of route forwarding, since the NDN network is a unique addressing mode of content names, and the content names are unique and layered in a name space, the content names are much more complex than IP addresses, may consist of dozens of hundreds or even hundreds of characters, and are much longer than IPv4/IPv6 addresses; in addition, the NDN name table may be orders of magnitude larger than today's IP forwarding tables. This requires that a lot of resources are occupied for finding and matching the router in the FIB, so that the processing delay is greatly increased;

(2) under the action of the existing NDN Routing protocol named Data Link State Routing protocol NLSR (named Data Link State Routing protocol), the router only sends the name prefix statically configured or dynamically registered by the local producer. This means that when a content is cached by the router CS, no advertisement update FIB is sent, which results in increased transmission overhead when the content is requested by the consumer, because the route information of the cached content is lacking and is still routed to the producer for fetching;

the root of the above disadvantages is that the current network architecture of the NDN is distributed, the computing capability of the router is generally low, and the efficiency and the speed are low when the complicated computing task of processing and searching the FIB is processed; the whole network lacks a global view angle, and other nodes cannot be notified when the content of the node CS is updated, so that the whole network is stiff, rigid and inflexible.

Disclosure of Invention

The invention aims to provide a flexible data retrieval method based on an NDN-SDN fusion framework, which is used for the field of information retrieval and solves the problems of the traditional NDN framework, aiming at the technical defects of long matching and searching processing time delay in FIB and incapability of flexibly and effectively distributing content announcement in the traditional NDN framework.

In order to achieve the purpose, the technical scheme of the invention is as follows:

the network architecture on which the flexible data retrieval method depends comprises a plurality of nodes; the nodes comprise users, routers and central controllers. The method fuses an original distributed NDN framework and an SDN framework and introduces a central controller. Because in the traditional NDN architecture, the router for processing and searching FIB has lower performance and lacks a global view angle, after the SDN with the high-computation-performance central controller architecture is introduced into the traditional NDN architecture, a control plane is separated from a forwarding plane, the forwarding plane directly sends an interest packet to the central controller, the central controller with strong processing performance analyzes, processes and calculates the interest packet and sends the interest packet to a router, so that the NDN router only needs to be responsible for forwarding, the forwarding router is completely liberated, the forwarding speed is fully improved, and the processing time delay is reduced; when the CS content of the node is updated, the change is informed to the central controller, so that the central controller can select the node with the lowest cost to the consumer when calculating the route, and the flexibility of the network is greatly improved; the control plane uses a Source Route (SR) to issue a route instruction for the forwarding router, the FIB matching process in the traditional NDN architecture is cancelled, PIT and FIB tables in the forwarding router are removed, and therefore a data flow mode based on pushing in the forwarding plane is achieved.

The specific implementation steps of the data retrieval and pushing method based on the NDN and the SDN are as follows:

step 1, setting a network central controller and announcing a whole network;

step 1.1, placing and selecting a central controller in a network central area;

the central controller is used for summarizing the global topology of the network, is responsible for name resolution, routing calculation, instruction issuing and the like in the data communication process, is used as a controller and plays a role of a control plane;

step 1.2, the central controller broadcasts an authentication interest packet to the whole network, and all nodes receiving the authentication interest packet use an interface receiving the authentication interest packet as a forwarding interface after receiving the authentication interest packet so as to ensure that all nodes except the central controller in the network can reach the central controller;

the interface refers to a hardware interface, namely an external interface of a host or a router, and is accessed to other hardware equipment through the interface; the other hardware devices are hardware devices relative to the current node host, namely hardware devices except the current node host;

it should be noted that if different interfaces receive multiple authentication interest packets, the first interface that receives the "authentication interest packet" is selected as the default forwarding interface, and the subsequent interface is not changed unless the central controller issues an instruction to change;

step 2, each router establishes a local link state database and a CS content library and informs a central controller, namely, a 'state update packet' is sent to a central server;

step 2.1, every two adjacent routers in the network mutually send a 'hello interest packet' to each other;

wherein neighboring routers mean "two routers are neighboring if communication between them does not need to pass through the other router";

wherein, the content in the hello interest packet is "the name of the router and the route cost for sending data to the router";

said step 2.1 is used to determine the name and link status of the neighboring routers in the network;

step 2.2, each router records the names of all adjacent routers of the router and the routing cost of the corresponding link according to all received hello interest packets, and forms a link state database of the router;

wherein, the link state refers to "the routing cost of the adjacent router of the router and the corresponding link", and the content of the link state database includes: the names of all adjacent routers of the router and the routing cost of the corresponding link;

step 2.3, each router forms a CS content library of the router according to the content stored in the CS;

wherein, the content of the CS content library comprises: the router CS stores the content names of all contents;

step 2.4, all routers except the central controller in the network send 'status update packets' to the central controller;

wherein, the content of the "status update package" includes: the name of the router, the names of all adjacent routers of the router, the routing cost of the corresponding link, and a content directory stored in the CS of the router, namely all information of a link state database and a CS content library of the local router;

step 3, the central controller establishes a full-network topological structure and a content distribution diagram;

wherein, the contents of the topology structure and the content distribution diagram of the whole network include: the names of all routers in the network, the content names stored in the CS corresponding to the routers, the connection relation among the routers and the routing cost of the corresponding link;

step 4, the user sends the data request interest packet to the adjacent edge router;

the "adjacent edge router" refers to a router directly connected to the user who sends the "data request interest packet"; the contents of the "data request interest package" include: requesting a data name of the content;

step 5, the router processes the data request interest packet and sends the processed data request interest packet to the central controller;

step 5.1, the router judges whether the received data request interest packet contains the access router name, if not, the interest packet is directly sent from the user, and step 5.2 is executed; otherwise, the interest packet is forwarded by other routers, and step 5.3 is executed;

wherein "access router" refers to a router directly connected to a user.

Step 5.2, the router writes the name of the router and the interface receiving the data request interest packet into the data request interest packet, and executes step 5.3;

the step 5.2 is used for enabling the central controller to know the name of the router sending the request and the interface to which the last data packet is sent after receiving the data request interest packet so as to calculate the route, and because the central controller only has a topology structure diagram of the router in the network, the name of the router receiving the data request interest packet and the interface receiving the data request interest packet are necessary information;

step 5.3, the router forwards the interest packet to the central controller;

in step 5.3, when all routers except the central controller in the network receive the data request interest packet containing the router name, the data request interest packet is sent to the central controller through an interface receiving the authentication interest packet by default;

step 6, the central controller calculates a route according to the whole network topological structure and the content distribution diagram;

step 6.1, the central controller performs name resolution on the received data request interest packet, and extracts the required content name, router name and interface information;

the required name content is a content name in a data request interest packet, the name of the router is a router which adds a name in the data request interest packet, the router is also a first router which receives an interest packet sent by a user, the interface information is interface information in the data request interest packet, and the interface information is a downstream interface of the first router which receives the data request interest packet;

6.2, the central controller carries out matching search on the content name in the content distribution graph to obtain a candidate node;

the process of matching and searching is similar to the process of searching FIB, and as long as the CS of the node stores the required content name, the node is a candidate node;

6.3, the central controller calculates the candidate nodes according to the node connection condition and the link routing cost in the whole network topology structure chart by using a Dijkstra optimal path algorithm to obtain the routing cost, and selects the candidate nodes and the paths with the minimum routing cost as target routes;

the central controller calculates route costs for candidate nodes respectively by using a Dijkstra optimal path algorithm, calculates the candidate routes as a starting point and the route names in the interest packets as an end point, then selects the candidate nodes with the minimum route costs and uses the minimum cost path from the candidate nodes to the downstream interface of the route names in the interest packets as the route;

step 7, the central controller issues the calculated route to a content owner router in the form of a source route;

7.1, the central controller converts the calculated route into a source route instruction in a form of interface forwarding and hop-by-hop;

the source routing instruction in the form of interface forwarding and hop-by-hop means that each hop is routed by using a forwarded interface as an identifier, and the central controller designates a forwarding interface of each hop router on a data forwarding path to guide routers along the path to forward data; the source routing instruction specifies that a forwarding interface of each hop router starts from an initial router, namely a next hop router is specified, and data can be sent to a user according to the forwarding interface of each hop router;

step 7.2, the central controller forwards the data command packet to the starting point router;

the content of the "data instruction packet" includes: the converted source routing instruction and the user-required content name in the received interest package; the starting point router refers to a candidate node with the lowest routing cost calculated by a Dijkstra optimal path algorithm;

the data command packet is sent to the starting point router through the central controller;

step 8, the router sends the data packets with the corresponding names to the user hop by hop according to the received source routing instruction;

step 8.1, the command issued by the central controller of the router is encapsulated into a 'push data packet' and forwarded;

wherein, the content of the "push data packet" includes: the source routing instruction sent to the user and the data content requested by the user are calculated by the central controller;

step 8.2, the router receives the sent push data packet, stores the data content in the push data packet in the CS, then checks the source routing instruction aiming at the router in the push data packet, and forwards the push data packet from the specified interface until the push data packet is forwarded to the user;

step 9, the router of the path caches the content in the CS, and notifies the central controller of the change of the content library;

step 9.1, the router of the path caches the content in the CS;

step 9.2 the router arranges the change content of the CS into a 'state update packet';

wherein, the content of the 'status update package' comprises: a content name directory that changes, specifically, a content name that is newly added in the CS and a content name that is deleted in the CS;

step 9.3, the router sends the 'status update packet' to the central controller;

the router sends the state updating packet to the central controller through the interface receiving the authentication interest packet;

step 9.4, the central controller updates the content distribution map according to the status update package;

so far, from step 1 to step 9, a data retrieval and pushing method based on the NDN and the SDN is completed.

Advantageous effects

1. Compared with the traditional NDN network which needs to use a router with low computational power to carry out FIB lookup matching based on name prefix in each step at present, the data retrieval and pushing method based on the NDN and the SDN disclosed by the invention realizes data packet sending based on pushing in a global view through a central controller and a source routing technology, removes PIT and FIB tables, thoroughly liberates the performance of the router, reduces the flow in the network and greatly reduces the processing delay of the router;

2. compared with the existing traditional NDN network in which the content notification only sends the locally registered content, the data retrieval and pushing method based on the NDN and the SDN disclosed by the invention has the advantages that the router notifies the content with the changed CS to the central controller in time, the content notification is flexibly and effectively distributed, the CS is utilized to the maximum extent, and the data acquisition delay is reduced.

Drawings

Fig. 1 is a diagram illustrating an example of a network topology and a central controller configuration in a first step in a data retrieval and push method based on NDN and SDN according to the present invention;

fig. 2 is a diagram illustrating an example of transmission routes of "data request interest packet", "data instruction packet" and "push data packet" in steps four to eight in the data retrieving and pushing method based on the NDN and the SDN according to the present invention;

FIG. 3 is a flowchart of the calculation of Dijkstra optimal path algorithm used in step six of the method for establishing and using an NDN-based edge calculation routing table according to the present invention;

Detailed Description

The present invention will be described in detail with reference to the accompanying drawings and embodiments, and the technical problems and advantages solved by the technical solutions of the present invention will be discussed at the same time. It should be noted that the described embodiments are intended to facilitate the understanding of the present invention and do not limit the present invention in any way.

The embodiment discusses the application of the data retrieval and pushing method based on the NDN and the SDN in data retrieval.

In this embodiment, based on the NDN-SDN network, a user sends an interest packet to request content with a data content name/wan, a router forwards the interest packet to a central controller, the central controller completes name resolution and routing calculation, and issues an instruction to a forwarding plane by using a source route, and finally the forwarding plane pushes data to the user according to the source instruction. The network topology and node arrangement in this embodiment are shown in fig. 1; the whole process of the user sending the interest packet, the router forwarding the interest packet, the central controller calculating and issuing the routing instruction to the forwarding plane, and the forwarding plane forwarding the data to the user according to the instruction is shown in fig. 2. In this embodiment, each node of the network is named "node (x ═ 0,1,2,3 … …)", where node0 is a user and node5 is a central controller. The embodiment of the data retrieval and pushing method based on the NDN and the SDN is applied as follows:

step 1, setting a network center controller and announcing the whole network, and specifically comprises the following substeps:

step 1.1, a central controller is placed and selected in a central area of a network, a node5 is set as the central controller in the embodiment, and the network topology and the central server are set as shown in fig. 1;

step 1.2, the central controller broadcasts an authentication interest packet with the name of "/Server" to the whole network, and all nodes receiving the authentication interest packet use an interface receiving the authentication interest packet as a forwarding interface after receiving the interest packet;

step 2, each router establishes a local link state database and a CS content database and informs a central controller, and the method specifically comprises the following substeps:

step 2.1, every two adjacent routers in the network send a Hello interest packet to each other, where the name of the Hello interest packet is "/Hello/nodex" (nodex is the name of the router, and x is 1,2,3 … …); the contents of the "hello interest package" include: the router name information and the routing cost of sending data to the router;

step 2.2, each router records all adjacent router names of the router and the routing cost of the corresponding link according to all received hello interest packets, and forms a link state database of the router; in this embodiment, router node7 receives hello interest packets from router node4 and node8, and the link state database list of router node4 is shown in table 1;

table 1 link state database listing for router node7

Neighboring router name	Route cost of link
		Node4	10
Node8	15

Step 2.3, each router forms a CS content library of the router according to the content stored in the CS; the CS content library of the router node6 in this embodiment is shown in table 2;

table 2 CS content repository for router node6

Storing content
	/want

Step 2.4, all routers except the central controller in the network send 'status update packets' to the central controller; the name of the "status update packet" is "/Server/nodex" (nodex is the name of the router, and x is 1,2,3 … …), and the contents of the "status update packet" include: the name of the router, the names of all adjacent routers of the router, the routing cost of the corresponding link, and a content directory stored in the CS of the router, namely all information of a link state database and a CS content library of the local router; the router forwards the authentication interest packet through an interface receiving the authentication interest packet, and the status update packet is forwarded to the central server;

the central controller collects all the received state updating packets to establish a whole network topological structure and a content distribution diagram; the data information of the topology and the content distribution map of the whole network in this embodiment is shown in table 3;

TABLE 3 Total network topology and content distribution map data information Table

in this embodiment, the user node0 sends a "data request interest packet" to the neighboring edge router node1, and the format of the "data request interest packet" is shown in table 4;

TABLE 4 "data request interest Package" content sent by user node0 to node1

Name of content	Router name
		/want	Is free of

Step 5, the router sends the processed data request interest packet to the central controller, and the method specifically comprises the following substeps:

step 5.1, the router judges whether the received 'data request interest packet' contains the router name, in the embodiment, after receiving the interest packet, the node1 judges that the router name does not exist, and proves that the interest packet is directly sent from the user;

step 5.2, the router writes the name of the router and the interface receiving the data request interest packet into the data request interest packet, in this embodiment, the node1 writes the name of the router into the data request interest packet, and at this time, the content of the data request interest packet is shown in table 5;

TABLE 5 data request interest Package content processed by node1

Name of content	Router name	Downstream interface
			/want	node1	0

Step 5.3, the router forwards the interest packet to the central controller, in the subsequent forwarding process of this embodiment, the node1 has written the name of the router into the "data request interest packet", and other routers only forward the "data request interest packet" according to the default interface, and do not process it until sending it to the central controller; the route forwarded to the central controller node5 is shown in figure 2;

step 6, the central controller calculates the route according to the whole network topology structure and the content distribution diagram, and the method specifically comprises the following substeps:

step 6.1, the central controller performs name resolution on the received data request interest packet, and extracts the required content name, router name and interface information; in this embodiment, the central controller node5 extracts all the contents of the "data request interest package" shown in table 5;

6.2, the central controller carries out matching search on the content name in the content distribution graph to obtain a candidate node; in an embodiment, as shown in table 3, two nodes with required/wait contents can be found, namely node6 and node9, and are selected as candidate nodes;

the method comprises the following specific steps: with the candidate node6 as the target, according to the link state information between nodes provided by the full network topology structure diagram, the minimum routing cost from node6 to node0 is 45 calculated by utilizing Dijkstra algorithm, the corresponding path is node6 → node5 → node4 → node1 → node0, and the forwarding interfaces are respectively: 1 → 1 → 0 → 0;

with the candidate node9 as the target, according to the link state information between nodes provided by the full network topology structure diagram, the minimum routing cost from node6 to node1 is calculated by utilizing Dijkstra algorithm to be 40, the corresponding path is node9 → node8 → node7 → node4 → node1 → node0, and the forwarding interfaces are respectively: 1 → 1 → 0 → 0;

therefore, in this embodiment, the target route is node9 → node8 → node7 → node4 → node1 → node0, and the forwarding interface is: 1 → 1 → 0 → 0 → 0.

And 7, the central controller issues an instruction to the content owner router in the form of a source route according to the calculated route, and the method specifically comprises the following substeps:

7.1, the central controller converts the calculated route into a source route instruction in a form of interface forwarding and hop-by-hop; in this embodiment, the central controller converts the calculated route into a forwarding interface instruction: 1 → 1 → 0 → 0 → 0, ready for delivery to node 9;

step 7.2, the central controller forwards the data command packet to the starting point router; in this embodiment, the central controller sends the "data command packet" to node9, where the route is as shown in fig. 2;

and 8, the router sends the data packets with the corresponding names to the user hop by hop according to the received source routing instruction, and the method specifically comprises the following substeps:

step 8.1, the router encapsulates the data into a 'push data packet' according to the instruction issued by the central controller and forwards the data packet; in this embodiment, the "push packet" is encapsulated by node9 as shown in table 6;

TABLE 6 push packet content after node9 processing

Source routing instructions: 1 → 1 → 0 → 0 → 0
	/want Data Packet

Step 8.2, the router receives the sent push data packet, stores the data content in the push data packet in the CS, then checks the source routing instruction aiming at the router in the push data packet, and forwards the push data packet from the specified interface until the push data packet is forwarded to the user; in this embodiment, a path for returning the "push packet" to the user is shown in fig. 2;

step 9, the router of the path caches the content in the CS and notifies the central controller of the change of the content library, specifically including the following substeps:

step 9.1, the router of the path caches the content in the CS, and node8, node7, node4, and node1 in this embodiment store/wait in the CS;

step 9.2, the router arranges the change content of the CS into a status update packet, in this example, node8, node7, node4, node1 store/wait in the CS, and then arranges the change content into a "status update packet", as shown in table 7;

TABLE 7 contents of "status update package" sent by node8, node7, node4, node1

Node name	Adding storage content	Deleting stored content
			node8\node7\node4\node1	/want	Is free of

step 9.4, the central controller updates the content distribution map according to the status update package; in this example, the updated topology and distribution of content is shown in table 8;

table 8 updated topology and content distribution map data information table of the whole network

It can be seen from the results of this embodiment that, in the conventional NDN data retrieval, the most time-consuming name resolution is required to be performed many times, but after the present invention is applied, the name resolution is completed only once in the central router (Node5) with high computation power, which greatly reduces the processing delay; in addition, compared with the traditional data retrieval mode, the invention establishes the index of the node and the name and the whole network topology in the central router, can effectively carry out route planning, avoids trapping in a local optimal trap and effectively improves the flexibility of the network.

The above detailed description is intended to illustrate the objects, aspects and advantages of the present invention, and it should be understood that the above detailed description is only exemplary of the present invention and is not intended to limit the scope of the present invention, and any modifications, equivalents, improvements and the like made within the spirit and principle of the present invention should be included in the scope of the present invention.

Claims

1. A data retrieval and pushing method based on NDN and SDN is characterized in that: the method comprises the following steps:

step one, setting a network center controller and announcing a whole network;

step two, each router establishes a local link state database and a CS content library and informs a central controller, namely, a state updating packet is sent to a central server;

step three, the central controller establishes a full-network topological structure and a content distribution diagram;

step four, the user sends the data request interest packet to the adjacent edge router;

step five, the router processes the data request interest packet and then sends the processed data request interest packet to the central controller;

the implementation method of the fifth step is as follows:

step 5.1, the router judges whether the received data request interest packet contains the name of the access router, if not, the interest packet is directly sent from the user, and step 5.2 is executed; otherwise, the interest packet is forwarded by other routers, and step 5.3 is executed;

wherein, the access router refers to a router directly connected with a user;

step 5.3, the router forwards the interest packet to the central controller;

step six, the central controller calculates a route according to the whole network topological structure and the content distribution diagram;

step seven, the central controller issues the calculated route to the content owner router in the form of a source route;

the implementation method of the seventh step is as follows:

the source routing instruction in the form of interface forwarding and hop-by-hop means that each hop of the route takes a forwarded interface as an identifier, and the central controller designates a forwarding interface of each hop of router on the data forwarding path to guide routers along the route to forward data; the source routing instruction specifies that a forwarding interface of each hop router starts from a starting router, namely a next hop router is specified, and data can be sent to a user according to the forwarding interface of each hop router;

the content of the "data instruction packet" includes: the translated source routing instruction and the user-required content name in the received interest package; the starting point router refers to a candidate node with the lowest routing cost calculated by a Dijkstra optimal path algorithm;

step eight, the router sends the data packets with the corresponding names to the user hop by hop according to the received source routing instruction;

the implementation method of the step eight comprises the following steps:

step nine, the router of the path caches the content in the CS and informs the central controller of the change of the content library.

2. The method of claim 1, wherein the NDN and SDN based data retrieval and pushing method comprises: the implementation method of the first step comprises the following steps:

step 1.1, a central controller is placed and selected in a network central area;

the central controller is used for summarizing the global topology of the network, is responsible for name resolution, routing calculation and instruction issuing in the data communication process, is used as a controller and plays the function of a control plane;

it should be noted that, if a plurality of authentication interest packets are received by different interfaces, the first interface receiving the "authentication interest packet" is selected as the default forwarding interface, and the subsequent interface is not changed unless the central controller issues an instruction change.

3. The method of claim 1, wherein the data retrieval and pushing method based on NDN and SDN is characterized in that: the implementation method of the second step is as follows:

wherein, the content of the "status update package" includes: the name of the router, the names of all neighboring routers of the router, the routing cost of the corresponding link, and the content directory stored in the CS of the router, that is, all information of the link state database and the CS content database of the local router.

4. The method of claim 1, wherein the data retrieval and pushing method based on NDN and SDN is characterized in that: the implementation method of the sixth step is as follows:

the central controller calculates route costs for the candidate nodes by using a Dijkstra optimal path algorithm, calculates the candidate routes as a starting point and the route names in the interest packets as an end point, then selects the candidate nodes with the minimum route costs, and uses the minimum cost path from the candidate nodes to the downstream interface of the route names in the interest packets as the route.

5. The method of claim 1, wherein the data retrieval and pushing method based on NDN and SDN is characterized in that: the implementation method of the ninth step is as follows:

step 9.1, the router of the path caches the content in the CS;

the router sends the state updating packet to the central controller through the interface which receives the authentication interest packet;

and 9.4, the central controller updates the content distribution diagram according to the status update package.