CN115865799B - General segment route representational slice design method based on IPv6 forwarding plane - Google Patents

Abstract

The invention relates to a general segment route representational slice design method based on an IPv6 forwarding plane, which reduces the influence of cross-domain interaction on network performance by limiting the network slice design to a single G-SRv domain. Meanwhile, due to the limitation of traditional network traffic scheduling, the flexible and accurate control of traffic scheduling in the VTN network slice is solved. Meanwhile, the invention is suitable for SDN software defined networks, and under the centralized or distributed software network controller, the network has higher flexibility and operability, and the flow sensing and control of the network slice by the user becomes simpler and more feasible. Meanwhile, the software open design can reduce the cost and difficulty of upgrading the hardware equipment, and the deployment and feasibility of the descriptive invention are greatly improved.

Description

General segment route representational slice design method based on IPv6 forwarding plane

Technical Field

The invention relates to a general segment route representational slice design method based on an IPv6 forwarding plane.

Background

The Segment Routing (SR) technology is generated under the influence of SDN thinking, and the core idea is to Segment (Segment) the paths and arrange and combine the paths at the initial node to determine the travel paths.

The SR technology supports two forwarding planes, where the SR based on the internet protocol version 6 (Internet Protocol Version, IPv 6) forwarding plane is referred to as SRv, and its SID is an IPv6 address. The IPv6 message is composed of an IPv6 standard header, an extension header and a load, and the implementation of SRv depends on an IPv6 extension header called SRH. The extension header formulates an IPv6 explicit path, stores the Segment list (SEGMENT LIST) information of the specified path of IPv6, the source endpoint of SRv can combine segments representing different functions according to the requirements, write the segments into SEGMENT LIST, and the intermediate node can forward the path information contained in the SRH extension header, which is the source routing principle of SRv 6.

As one of important members of the IPv6 plus technology, SRv can realize granularity and precision flow control on network services, has good principle support on virtual bearer network slicing, and the SR technology is generated under the influence of SDN thinking, and has the core ideas of segmenting paths and arranging and combining the paths at initial nodes to determine travel paths. However, SRv has excessive data overhead, and the consumption of slice node resources is added, which causes heavy load to the network. And as the number of network slices increases, there is a need to provide more optimized inventions to improve the performance of VTN slices.

Disclosure of Invention

The invention provides a general segment route representational slice design method based on an IPv6 forwarding plane, which aims to solve the problems of node resource efficiency and data packet load efficiency in a SRv-based virtual bearer network slice, realizes resource isolation of the network slice, and provides stronger feasibility for actual deployment of G-SRv6 in a VTN slice.

The technical solution for realizing the purpose of the invention is as follows:

a general segment route representational slice design method based on an IPv6 forwarding plane comprises the following steps:

s1, enabling traffic to enter a virtual bearer network (VTN) slice, enabling a source node to calculate a routing strategy based on the constraint of the network slice, namely, calculating a routing table according to a routing algorithm through communication work of a plurality of nodes, dynamically selecting a data transmission path, determining a traffic forwarding path, and expressing the traffic forwarding path in a form of a segmented routing Local segment identifier (SRv Local SID) list;

s2, carrying out representation of re-slicing on the VTN slice according to the SRv Local SID list;

S3, mapping SRv Local SID list into a general segment identifier (G-SID) list in a controller or a data plane;

s4, the transit node and the segment end node guide the forwarding work of the data message in the flow according to the path expression of the G-SID list.

Further, after the network slice source node receives the data message, calculating a segment route (SRv 6) path meeting specific requirements such as network transmission rate and bandwidth according to a preset routing strategy, encapsulating a series of SRv Local SIDs in a segment route message header (SRH), and mapping SRv Local SIDs into a path expression display of G-SIDs through a mapping relation to guide message forwarding.

Further, in S1, the source node needs to express the forwarding path as a G-SID list through mapping of the controller or the data plane, in addition to performing calculation of the routing policy.

Further, SRv Local SIDs of the VTN slices are the basis for routing decisions, and G-SIDs under the slices are the basis for routing execution behavior.

Further, in S3, for the mapping relationship between SRv Local SID and G-SID, in the scenario where the controller management and control capability is high (reaching the limit value), mapping is directly performed in the controller, and when the node needs to query the mapping relationship, the node communicates with the controller; in case the controller has not high control capacity (limit is not reached), the mapping table is directly used in the data plane, and mapping information is reserved at each node.

Further, in S3, mapping the slice path essentially decouples both routing decisions and routing behavior expressions of the IP address.

Further, in S4, after receiving the data, the transit node and the segment end node identify and parse the sequence of the G-SID, locate the SID list by using the double fingers of the next hop (SL) and the compressed next hop (SI), find the next destination G-SID as the destination address of the next hop, and then, according to the address information matched with the forwarding table entry, confirm the forwarding port of the data at the node, so as to realize the forwarding of the traffic under the guiding path.

Further, in S4, the transit node and the segment end node calculate and generate a routing table by using SRv Local SID information, make a routing decision, and generate a forwarding table by using 32bit G-SID calculation, so as to guide the actual forwarding of the message at the device port.

Compared with the prior art, the invention has the remarkable advantages that:

1. after a slice source node calculates a routing strategy, the routing execution behavior is expressed by the mapping of the G-SID, and under the condition that the mapping change of an SID list does not influence the actual flow state, the cost of network slice node data and forwarding table items is reduced, the cost problem of nodes and data in a VTN slice of SRv is solved, and the network bandwidth utilization rate and node resource utilization rate are improved;

2. The slice is restated by the G-SRv technology, so that more accurate scheduling and control of network slice flow are realized, and a more optimized and higher service quality bearing network slice is provided;

3. The VTN slice based on SRv maintains the original characteristics of simplicity in cross-domain, traffic scheduling isolation and strong expandability of SRv.

Drawings

FIG. 1 is a SRv message forwarding flow diagram;

FIG. 2 is a diagram of a carrier network slicing architecture;

FIG. 3 is an exemplary diagram of G-SRv6 compressing a list of SIDs;

FIG. 4 is an exemplary diagram of a G-SRv tag map;

FIG. 5 is an illustration of a SID forwarding column representation of G-SRv;

fig. 6 is an exemplary diagram of a G-SRv network topology application.

Detailed Description

Embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

SRv the message forwarding flow of SRv is shown in fig. 1, and assuming that there is a message to be forwarded from host 1 to host 2, host 1 sends the message to node a for processing. Node A, B, D, E supports SRv6, node C does not support SRv, only IPv6. We perform network programming on source node A, hopefully the message passes through the B-C, C-D link to node E, from which it is sent to host 2. The message forwarding flow comprises the following steps:

1. the source node A encapsulates SRv path information in SRH, designates SID of B-C, C-D link, and encapsulates SID A5:10 issued by E point (this SID corresponds to one IPv4 VPN of node E), total 3 SIDs, and presses SID sequence in reverse order form. At this time SL (Segment Left) =2, copy SEGMENT LIST value to the DA field of the destination address, find the IPv6 routing table according to the longest match rule, and forward it to the node B.

2. The message arrives at the node B, the node B searches a local SID table (storing SRv SID information generated by the node) and hits the SID (end.X SID) of the node B, and the instruction action corresponding to the SID is executed. The SL value is decremented by 1 and SEGMENT LIST values are copied to the DA field, while the message is sent out of the SID-bonded link (B-C).

3. The message reaches the node C, the node C has no SRv capacity and cannot identify SRH, and according to the normal IPv6 message processing flow, the IPv6 routing table is searched according to the longest matching principle, and the node D represented by the current destination address is forwarded.

4. After receiving the message, the node D searches a local SID table according to the destination address A4, namely 45, hits the SID (end.XSID) of the node D. And subtracting 1 from the SL value of the node B, taking A5:10 as DA, and sending out the message.

5. After receiving the message, the node E searches a local SID table according to A5:10, hits the self SID (end.DT4SID), executes corresponding instruction action, decapsulates the message, removes the IPv6 message header, and forwards the inner layer IPv4 message by looking up the table in the IPv4 routing table of the VPN example bound by the SID, and finally sends the message to the host 2.

Network slicing is a logical end-to-end connection that does not change the underlying hardware physical infrastructure, connecting users using shared physical resources and isolated logical connections, which are used to meet specific service level objectives (SERVICE LEVEL Objectives, SLS) and service level expectations (SERVICE LEVEL Expectations, SLEs). From the user's perspective, with 5G network slicing, users can easily manage and use proprietary network virtual resources to meet specific connection needs. Network slicing is based on resource management, network virtualization and abstract concepts, and can be deployed and designed by using SDN, NFV and other technologies.

The bearing network slice mainly has a three-layer structure, comprising a network infrastructure layer responsible for physical resource reservation, a network slice instance layer for establishing logical connection and a network slice management layer for managing the life cycle of the network slice, and the structural relationship is shown in fig. 2. Wherein the network slice instance layer is composed of a virtual service network (VPN) of an upper layer (Overlay) and a virtual bearer network (Virtual Transport Network, VTN) of a lower layer (Underlay), the focus of the study is the VTN of the network slice instance layer. In the network slicing example layer, a virtual bearing network with resource assurance is built based on the resources allocated by each bearing network device for network slicing, and VPN services are deployed on the virtual bearing network according to requirements. This includes both data plane schemes and control plane scheme considerations. The data plane function of the network slice example layer is to carry the identification information of the network slice in the data service message, and is used for distinguishing the messages belonging to different network slices and indicating the bottom layer network resources reserved for the network slices, so as to guide the messages of different network slices to be forwarded according to the topology, the resources and other constraints defined by the network slices. The data plane needs to provide a generic identification to decouple from the specific resource isolation techniques of the network infrastructure layer.

In the invention, the encapsulation mode of the network slice data plane of the carrier network is that each network device allocates a special SRv locator (position identifier) for different network slices, and allocates corresponding SRv SID (segment routing identifier, segment route identifier) for different network functions in the network slices under the address space of the locator.

SRv6 is based on the original IPv6 address, one problem which is not negligible is that 128bit SID packages resources are wasted, namely overhead is too large, so that not only are load efficiency and transmission efficiency reduced, but also requirements on the switch chip for bearing and processing data are higher, and the cost of hardware facilities is increased. To solve the compression problem of the SRv SID list, researchers have proposed a generalized SRv6 (G-SRv, generalized SRv 6) solution for optimizing SRv6 data headers.

Each node of the native SRv Local SIDs is independently allocated from its own Locator address space, while most of the locators of nodes in the network are allocated step by step from the same large segment of address space, which is called Common Prefix (Common Prefix). G-SRv is used for extracting prefix information shared by SID sequences, the SID is represented by a 32-bit unique information field, namely, the G-SID, so that the compression of a SRv SID list is realized, and the utilization rate of data resources is improved. The G-SRv can support SRv compression, reduce the header overhead of SRv6 messages, and can be mixed with the traditional SRv6 SID to be programmed in one SRH, so that good SRv6 compatibility is achieved.

G-SRv6 defines COC (Continuation of Compression) Flavor, which indicates that the next SID is a compressed G-SID if it contains COC information, and a 128-bit SRv SID if it does not have a COC Flavor identification. Meanwhile, G-SRv utilizes SI (Compressed SID left) to locate G-SID information, and a next hop address updating mechanism under a compressed and non-compressed mixed editing scene is realized through SL and SI double pointers.

Fig. 3 illustrates an example of G-SRv6 compressing a list of SIDs. Under the forwarding path of the node a-B-C, as shown in (a) of fig. 3, the list of 3 SIDs occupies 3×128 bits of the IP packet header. The original message header content is improved by using the G-SRv technology, because all SIDs in (a) in fig. 3 have public prefix information 3001, the public prefix 3001 can be extracted, the length of the extracted public prefix is 96 bits, the rest of SIDs are represented in the form of G-SIDs, namely, the length of each G-SID is 32 bits, 3G-SIDs are 3 x 32 bits in total, and 128 bits are added after padding, as shown in (b) in fig. 3. By comparing the list packages, it can be seen that the overhead of the G-SRv technology to the data messages is greatly reduced.

The following description is made here of the above concepts, respectively:

SRv6 Local SID: is an IPv6 address explicitly representing SRv functions for use in routing decisions, computation to generate a routing table. In the virtual bearer network slice, a node may be allocated with multiple Local SIDs, which not only indicates the network slice to which the node belongs, but also can determine resources allocated to the network slice by the node, including related parameters such as node link resources in the network slice topology and bandwidth queues corresponding to the network slice, so as to implement resource reservation and resource isolation of the underlying network slice. Under this scheme SRv Local SID is the basis for SRv routing table generation.

G-SID: the G-SID is a carrier of unique information of SID in G-SRv, has a word length of 32 bits, and has the same function as SID. Under the scheme, the G-SID is the basis for generating a forwarding table. The G-SID identification is used for route expression, and a forwarding table is generated through calculation.

Source node: the node that generated SRv's 6 message is referred to as the SRv source node. At a source node of the network, segment list information is encapsulated into a message. The network converts service and service demands into policies for data forwarding through a source node calculation path, encapsulates SID list into data message to realize control of forwarding instructions of data, and the SID list can guide network traffic to forward according to corresponding constraint by calculation of the source node policies.

Transit node and Duan Duandian node: and routing decisions are not carried out, and forwarding and receiving processing work of SRv messages are respectively responsible.

The functionality of an IP address in a network is summarized in two ways: addressing and routing. Addressing refers to the use of the globally unique identifying characteristic of an IP address by which a device in a network can be uniquely determined by other devices can find the device. And the routing function refers to that the IP address can be used for guiding the routing behavior of the data, and the node can determine the forwarding target of the data packet through the IP address so as to guide the end-to-end transmission of the data from the source address to the destination address. Since the data packet in SRv is based on the policy transmission by the routing source node, at the ingress node the forwarding path of the data packet in the network has been determined, i.e. the router entity through which the message route needs to pass has been determined at the path start point. Whereas the role of SRv SID list is to show the routing decision in a behavioral way only and does not correspond to the router itself. Thus, if synchronized mapping modifications are made to the already decided SID list and the router's label, no behavioral impact is imposed on the packets that already encapsulate the SID list.

In the conventional IPv6 network, since the routing decision and the routing behavior of the data packet are coupled, when the IP address is allocated, each hop must correctly couple the addressing function and the routing function of the IP address to ensure that the forwarding of the network proceeds smoothly, so when the routing behavior is executed, it must be ensured that the next hop address in the routing forwarding packet is the same as the next hop address pointed by the routing decision, so as to ensure that the data packet can be correctly sent to the decided router entity. In order to avoid the influence of dynamic update route on the addressing and routing decision behavior of nodes, we separate the addressing function and routing function of IP addresses, which are respectively represented by programming labels and routing labels, so as to ensure that changing SID list will not affect the entity path. Meanwhile, the separation addressing function and the routing function can also facilitate the compression of SID lists.

After receiving the data message, the network node mainly performs two aspects of processing: routing decisions and routing behavior. The former relates to the topology change condition of the whole network, a plurality of nodes are required to communicate, a routing table is calculated according to a routing algorithm, and a data transmission path is dynamically selected. The latter is realized by a forwarding table of the node, only the local node is involved, and the message is guided to enter from one port of the node and exit from the other port by the information of the forwarding table item. Therefore, after each node receives the data message, the forwarding table is really queried, and the forwarding work of the ingress and egress guiding data of the port is utilized according to the content of the forwarding table, so that the traffic transmission is realized.

Whereas in SRv networks, data is transported over the network based on source routing mechanisms. The routing decision of the data completes the corresponding calculation and logic determination at the entrance of the network, and the source node expresses the routing policy by encapsulating SRv Local SID list. Therefore, under the condition that the source node is not influenced to calculate the corresponding forwarding path, mapping expression is carried out on the determined SRv Local SID list, and the routing forwarding behavior of SRv network data is not influenced.

Based on the above principle, the invention restates SRv slice problems of the virtual bearer network: firstly, after the traffic enters the VTN slice, the source node calculates a routing strategy based on the constraint of the network slice, determines the forwarding path of the traffic, and expresses the forwarding path in the form of SRv Local SID list. And according to SRv Local SID list, re-slicing the VTN slice, mapping SRv Local SID list into G-SID list in controller or data plane, and forwarding the data message by transfer node and segment end node according to path expression of G-SID list. The data is calculated at the source node by the routing policy, and the forwarding behavior and rules of the data are logically determined, so that the SID list for routing expression is mapped in a correlated manner, and the routing planning and the traffic state of the network are not affected.

Fig. 4 shows the mapping between addressing labels and routing labels in the G-SRv expression, where the following two traffic flows calculate the corresponding forwarding paths based on SRv Local SID (e.g. 2001:: 1), and the actual forwarding is performed on each node on the path based on G-SID (e.g. a:: 1).

The mapping of VTN slice paths essentially decouples both routing decisions and routing behavior expressions for IP addresses. SRv6 Local SIDs of the VTN slices are the basis for routing decisions, and G-SIDs under the slices are the basis for routing execution behaviors. Nodes in the network utilize SRv Local SID information to calculate and generate a routing table, carry out routing decision, and utilize 32bit G-SID to calculate and generate a forwarding table to guide the actual forwarding of the message at the device port.

Under this scheme, the source node needs to express the forwarding path as a G-SID list through mapping of the controller or the data plane, as shown in fig. 5, in addition to performing the calculation of the forwarding policy. After receiving the data, the transit node and the segment end node identify and analyze the sequence of the G-SID, locate the SID list by utilizing the SL and the SI double fingers, and find the next destination G-SID as the destination address of the next hop. And then, according to the corresponding forwarding table item matching address information, confirming the forwarding port of the data on the node, thereby realizing forwarding of the traffic under the guiding path.

On the basis of not influencing the flow decision and behavior of the source node, the compression efficiency of the G-SRv scheme is optimized by carrying out G-SRv description on the VTN slice. The content of the high SID list of the original G-SRv scheme, which depends on the address information, will affect the compression efficiency of G-SRv, the actual compression efficiency of which is affected by the addressing policy, and in most cases, the prefix allocation of the router is not aggregated, so the efficiency is low. Under the scheme, the G-SID is used as the expression basis of the re-slicing, and the main function is to express the routing information and not bind with a router entity, and if the allocation of the G-SID is optimized, the addressing of the network is not affected. Therefore, in order to guarantee the compression efficiency of the G-SRv under the dynamically changing network, the performance slice is dynamically mapped, that is, the mapping relationship changes along with the change of the network traffic state. The final network can introduce the same network prefix into the expressive slice in the path with large flow according to different network quality demands such as time delay, reliability, packet loss rate, bandwidth and the like, and optimize the compression efficiency of the SID list.

Meanwhile, under the G-SRv descriptive slice, the flow of a plurality of slices can be expressed as the same G-SID list, and the expression of the path information is carried out. Therefore, considering the isolation problem between slices, an APN6 technology can be introduced, and more refined and more granular control of the network on the traffic can be realized by applying the sensing capability. And by utilizing a path expression mechanism, the flow between the slices is ensured not to be affected, and the perceptibility and isolation capability of the slices to the service are enhanced.

In the invention, after the source node receives the data, the related routing decision is not affected, the slicing path meeting the condition is calculated by a static designated path, a head node path calculation or a controller path calculation mode, a SID list is inserted into the data message, the message is explicitly guided to be forwarded according to the specific path and the condition, and the end-to-end fine granularity control of the network flow is realized. And then, mapping the SID list into the expression of the G-SID through a mapping relation, so as to realize the compression of SRv data packets, reduce the network load borne by nodes and links, and improve the SRv network data packet load efficiency.

The effectiveness of the present invention is mainly embodied in the following aspects:

1. Improving SRv data packet load efficiency

After receiving the data, the source node calculates a slicing path meeting the conditions in a static designated path, head node path calculation or controller path calculation mode, and inserts an SID list for the data message. By re-expression of the VTN slices, the list of SIDs is mapped to an expression of G-SIDs, while the determined routing decisions are not affected in any way. And the rest nodes on the path guide the hop-by-hop forwarding of the data according to the displayed G-SID list, so that the end-to-end scheduling and control of the network traffic are realized. By the G-SRv6 descriptive slice design scheme, SRv data packets are compressed, network loads borne by nodes and links are reduced, and SRv network data packet load efficiency is improved.

2. Reducing isolation overhead of slice nodes

For the data overhead problem of a node, each network node in a network slice needs to calculate forwarding entries generated to other nodes within the network slice. For the same destination node, separate forwarding entries are required for different network slices. Under the descriptive scheme, a 32bit G-SID is taken as the basis of a forwarding table. Giving out a network topology, as shown in fig. 6, the node forwarding table is shown in table 1, and compared with the original multiple forwarding table entries, the node forwarding table is established based on shorter G-SID information, so that the overhead of the node for the size of the network slice forwarding table entry is greatly reduced, and the node resource efficiency is improved.

TABLE 1

3. Optimized G-SRv6 compression efficiency

And introducing the same network prefix to the restated slice in the path with large flow aiming at different network quality demands such as time delay, reliability, packet loss rate, bandwidth and the like, and optimizing the compression efficiency of the SID list. Because the network traffic state is continuously changed, the mapping relation between SRv Local SID identifications and G-SID identifications of the virtual bearer network slice nodes is dynamically changed, and the mapping relation can be changed along with the change of service requirements, so that better compression efficiency and network performance are realized.

Under the G-SRv descriptive slice, the traffic of multiple slices may be expressed as the same G-SID list, and the path information is expressed. Therefore, considering the isolation problem between slices, an APN6 technology can be introduced, and more refined and more granular control of the network on the traffic can be realized by applying the sensing capability. And by utilizing a path expression mechanism, the traffic among the network slices is ensured not to be affected, and the perceptibility and isolation capability of the slices to the service are enhanced.

The invention relates to a processing of data by a node, which comprises the following steps:

After the network slice source node receives the data, SRv paths meeting specific requirements are calculated according to a certain strategy, a series of SRv Local SIDs are packaged in the SRH, and then the SRv Local SIDs are mapped into path expression display ground guide messages of the G-SIDs through corresponding mapping relations to carry out corresponding forwarding. And for the mapping relation between SID and G-SID, mapping is directly carried out in the controller under the scene of higher control capacity of the controller, and the node is communicated with the controller when the mapping relation is required to be inquired. The mapping table can be directly used in the data plane when the control capability of the controller is not high, and mapping information is reserved at each node.

After receiving the data, the transit node and the segment end node identify and analyze the sequence of the G-SID, locate the SID list by utilizing the SL and the SI double fingers, and find the next destination G-SID as the destination address of the next hop. And then, according to the address information matched with the forwarding table entry, confirming the forwarding port of the data at the node, thereby realizing the forwarding of the traffic under the guiding path.

In summary, in order to provide a more optimized solution to improve the performance of VTN slicing, the present invention uses the advantages of source routing to essentially decouple the routing decision and routing behavior expression of IP addresses for mapping slice paths on the basis that the traffic decision and behavior of source nodes are not affected. SRv6 Local SIDs of the VTN slices are the basis for routing decisions, and G-SIDs under the slices are the basis for routing execution behaviors. Nodes in the network calculate and generate a routing table by utilizing SRv Local SID information, carry out routing decision, generate a forwarding table by utilizing G-SID calculation, and guide the actual forwarding of a message at a device port.

The invention reduces the influence of cross-domain interaction on network performance by limiting the network slice design to a single G-SRv domain. Meanwhile, due to the limitation of traditional network traffic scheduling, the flexible and accurate control of traffic scheduling in the VTN slice is solved. Meanwhile, the scheme is suitable for SDN software defined networks, the network has higher flexibility and operability under the centralized or distributed software network controller, and the flow sensing and control of the VTN slice by the user becomes simpler and more feasible. Meanwhile, the software open design can reduce the cost and difficulty of upgrading the hardware equipment, and the deployment and feasibility of the expressive scheme are greatly improved.

Claims (7)

1. The general segment route representational slice design method based on the IPv6 forwarding plane is characterized by comprising the following steps of:

S1, enabling traffic to enter a virtual bearer network VTN slice, enabling a source node to calculate a routing strategy based on the constraint of the network slice, determining a forwarding path of the traffic, and expressing the forwarding path in the form of a segmented routing Local segment identifier SRv and a Local SID list;

s2, carrying out representation of re-slicing on the VTN slice according to the SRv Local SID list;

S3, mapping SRv Local SID list into a general segment identifier G-SID list in a controller or a data plane;

S4, the transit node and the segment terminal node guide the forwarding work of the data message in the flow according to the path expression of the G-SID list;

The calculation of the routing policy in S1 specifically includes: through the communication work of a plurality of nodes, a routing table is calculated according to a routing algorithm, and a data transmission path is dynamically selected;

S3, for the mapping relation between SRv Local SIDs and G-SIDs, mapping is directly carried out in the controller under the scene that the control capacity of the controller reaches the limit value, and the node is communicated with the controller when the mapping relation needs to be inquired; under the condition that the control capacity of the controller does not reach the limit value, the mapping table is directly used in a data plane, and mapping information is reserved at each node;

And S4, after receiving the data, the transit node and the segment end point node identify and analyze the sequence of the G-SID, locate the SID list by utilizing the double fingers of the next hop SL and the compressed next hop SI, find the next destination G-SID as the destination address of the next hop, and then confirm the forwarding port of the data at the node according to the address information matched with the forwarding table item so as to realize the forwarding of the flow under the guiding path.

2. The general segment routing declarative slice design method based on an IPv6 forwarding plane of claim 1, wherein: after the network slice source node receives the data message, calculating a segmented route SRv path meeting specific requirements according to a preset routing strategy, encapsulating a series of SRv Local SIDs in a segmented route message header SRH, and mapping SRv Local SIDs into a path expression display of G-SIDs through a mapping relation to guide the message to be forwarded.

3. The general segment routing declarative slice design method based on an IPv6 forwarding plane according to claim 2, wherein: specific requirements include specific network transmission rates and bandwidths.

4. The general segment routing declarative slice design method based on an IPv6 forwarding plane of claim 1, wherein: in S1, the source node needs to express the forwarding path as a G-SID list through mapping of the controller or the data plane, in addition to performing calculation of the routing policy.

5. The general segment routing declarative slice design method based on an IPv6 forwarding plane of claim 1, wherein: SRv6 Local SIDs of the VTN slices are the basis for routing decisions, and G-SIDs under the slices are the basis for routing execution behaviors.

6. The general segment routing declarative slice design method based on an IPv6 forwarding plane of claim 1, wherein: the mapping in S3 essentially decouples both the routing decision and the routing behavior expression of the IP address.

7. The general segment routing declarative slice design method based on an IPv6 forwarding plane of claim 1, wherein: and S4, the transit node and the segment end point node calculate and generate a routing table by utilizing SRv Local SID information, make a routing decision, generate a forwarding table by utilizing G-SID calculation, and guide the actual forwarding of the data message at the equipment port.