Method for establishing a signaling connection based on dns

Abstract

The application provides a method for establishing a signaling connection based on DNS, which constructs a DNS domain name resolution system suitable for a 5GC architecture, wherein a source network element only needs to provide domain name information of a target network element to a DNS network element, the DNS network element feeds back two optimal and suboptimal connection information to the source network element as a response according to the provided information through optimization, and the source network element can establish signaling communication with the target network element. When the connection is interrupted, the two network elements can continue to communicate at the moment of failure, thereby reducing the influence of the failure on the signaling communication. Since all regional DNSs save the network element state of the whole network broadcast by the root DNS, if a certain regional DNS fails, the regional DNSs in the adjacent region can process it, thereby significantly reducing the requirement for the reliability of the regional DNS and reducing the cost. In addition, when the burdens of the regional DNSs in adjacent regions differ by more than a set critical value, the regional DNSs can cooperate with each other, thereby improving the overall processing capacity of the system.

Description

Technical Field

[0001] This invention relates to a method, apparatus, and computer-readable storage medium for establishing a signaling connection based on DNS. Background Technology

[0002] As is well known, the signaling network is a crucial component of mobile communications, and its performance and functions play a key role in the entire mobile communication network. Signaling communication in 2G / 3G networks is based on Signaling System No. 7 (SS7) carried by TDM. The 4G core network packet domain (EPC) uses Diameter signaling, primarily applied to the PCC policy and charging control system and the IMS domain. 5G's NSA (Non-Standalone) architecture is built using EPC+ (EPC upgrade), with signaling types and networking methods similar to 4G. 5G's SA (Standalone) core network adopts a new network architecture based on service-oriented interfaces, with communication between various NFs (Network Functions, also known as network elements) based on new HTTP 2.0 signaling messages.

[0003] With the large-scale commercialization of 5G, the connections between network elements are becoming increasingly complex, and signaling networks have become an important direction for the evolution of 5G.

[0004] The 5GC architecture introduces NRF (Network Repository Function), which supports the following functions: 1) Supports service discovery, receiving network element discovery requests from network element instances and providing the discovered network element instance (discovered) information to the network element instance; 2) Maintains network element configuration files of available network element instances and their supported services, supports network element capability registration and service discovery, and data configuration and route query are all completed through NRF.

[0005] In Release 16, 3GPP introduced a new network element, SCP (Service Communication Proxy). SCP is similar to STP in Signalling No. 7 and DRA in Diameter signaling network; it is a relay proxy device for HTTP signaling messages within the 5G SA core network, used to simplify signaling routing and network organization within network elements.

[0006] 3GPP Release 16 provides the following four networking modes (Option AD) for inter-network element communication in Appendix E:

[0007] Option A (without NRF, SCP) uses the network element configuration service to provide network element profiles and communicate directly with the selected network element;

[0008] Option B (with NRF, without SCP) uses network elements for service discovery, queries the NRF to obtain the list of network elements providing the service, and directly sends the request to the selected service provider.

[0009] Option C (NRF Query, SCP Routing) uses network elements for service discovery, queries the NRF to obtain a list of service-providing network elements. The service uses the network element to send the address of the service-providing network element to the SCP (or sends a list of service-providing network elements to the SCP, where the SCP selects the service-providing network element; the SCP can also obtain parameters such as the location and capacity of the service-providing network element from the NRF). The SCP is responsible for routing the request to the service-providing network element.

[0010] Option D (SCP queries NRF and routes): The service-using network element does not perform any service discovery or selection. It needs to add the relevant parameters required for discovering the service-providing network element to the service request message. SCP uses these parameters to route the message to the applicable service-providing network element. SCP and NRF jointly complete service discovery and obtain the discovery result.

[0011] As 5G networks expand, the number of users increases, and 5G signaling connection management faces growing challenges. Currently, the 5G core network (5GC) based on NRF and SCP faces the following issues:

[0012] In Option A mode, the source network element needs to be configured with the addressing data of the target network element. When the values ​​of the source network element and the target network element in the network are large, a large number of source network elements need to be configured and maintained with a large amount of routing data. Moreover, if a source network element is misconfigured, the workload for troubleshooting will be enormous.

[0013] While Option B simplifies the routing office data configuration for network elements, it requires network elements to configure data belonging to the NRF, register with the NRF, and configure data for target network element service discovery and selection. This means configuring the query parameters provided to the NRF for discovering target network elements. However, the NRF cannot obtain real-time dynamic data about the network elements.

[0014] In addition, for Option B mode, NRF matches a set of available target network elements based on the query parameters provided by the source network element and returns the results to the source network element. The source network element needs to select the final target network element from multiple available target network elements. Since this selection function is distributed across each network element, the algorithm complexity for selection cannot be too high, which affects the optimization effect of target network element selection, limiting the selection to low-complexity algorithms.

[0015] Compared to Option B, Option C has the same data volume configured for both the network element and the NRF, and the same method for handling HTTP signaling. The TCP connection between the source and destination network elements is now aggregated by the SCP to which the network element belongs, significantly reducing the number of TCP connections maintained by the network element. However, due to the proxying role of the SCP, signaling connection latency increases, and the reliability requirements for the SCP are high. If the SCP to which the network element belongs fails, even if the network element and the NRF are still functioning normally, normal communication between network elements will be impossible.

[0016] Option D maintains the mechanism of network elements registering with their home NRF, and the HTTP path can be organized through the SCP, but it transfers the network element service discovery work of the source network element to the SCP. When the source network element needs to interact with the target network element via HTTP signaling, the source network element sends the HTTP signaling request message directly to the home SCP, carrying the query parameters required for target network element service discovery. The home SCP then executes the target network element service discovery process with the home NRF and selects the target network element. Option D further simplifies the data configuration of network elements, requiring only the data registered with the home NRF and the data to the home SCP to be retained, while the NRF's configuration data remains unchanged. However, compared to Option C, SCP requires the addition of a target network element service discovery process and data configuration for selecting the target network element. The number of TCP connections for the network element is the same as in Option C, leading to greater complexity in connection management.

[0017] The existing Option AD in 5GC has an optimization problem in selecting target network elements because the existing mechanism does not have a way to obtain real-time information about the target network elements.

[0018] The existing Option AD in 5GC has the problem of high complexity in source network element connection management. Due to the large number of network elements in the entire network, the cumulative complexity of all nodes in the entire system will increase significantly, which not only increases the cost of the entire system, but also increases the energy consumption of the entire system, as well as subsequent operation and maintenance and management.

[0019] In response to the needs of intelligent and green networks in the evolution of 5G, and based on the existing 5G SA networking mode, a signaling connection mechanism with low complexity of network element connection management and more intelligent connection optimization capabilities is constructed. This is of positive significance for adapting to the needs of large-scale 5G applications, promoting the development of 5G networks, and accumulating technology for the future 6G. Summary of the Invention

[0020] In view of this, the present invention provides a method for establishing signaling connections based on DNS. This method constructs a DNS domain name resolution system suitable for the 5GC architecture. Network Functions (NFs) only need to provide the target network element's domain name information to the DNS network element, which is either the root DNS or a zone DNS. Based on the provided information, the DNS network element optimizes and selects the optimal and second-best connection information as responses to the source network element. The source network element can then use the DNS mechanism's fast connection mode to establish a connection with the target network element for signaling communication. When a connection is interrupted, a message is automatically sent to the target network element pointed to by the backup connection. If the target network element responds, the completed signaling records of the primary connection's target network element are packaged and sent. The target network element of the backup connection restores the completed signaling process, allowing both network elements to continue during a failure, reducing the impact of the failure on signaling communication. Since all zone DNS stores the network element state broadcast by the root DNS for the entire network, if a zone DNS fails, it can be handled by a neighboring zone DNS, meaning adjacent zone DNS serve as backups for each other. This significantly reduces the reliability requirements of zone DNS and lowers costs. In addition, when the difference in DNS load between adjacent areas exceeds the set threshold, they can coordinate with each other (e.g., coordination between office areas and residential areas at different times), thereby improving the overall processing capacity of the system.

[0021] The method according to the present invention includes the following steps: establishing a two-layer signaling DNS structure including a root DNS and regional DNS; the root DNS establishing a real-time status information registry of all network elements and a domain name resolution table of all regional DNS, and the root DNS periodically broadcasting the real-time status information registry of all network elements to all regional DNS; the regional DNS recording the real-time status information registry of all network elements, and the regional DNS periodically collecting the network element status information of network elements in its own region, and sending the aggregated network element status information of network elements in its own region and its own status information to the root DNS; the source network element establishing a signaling connection with the target network element through the regional DNS of its own region, and initiating a domain name resolution request to the regional DNS, the request carrying the target network element information and subsequent service information; the regional DNS selecting the optimal and second-best target network elements according to the domain name resolution request through the real-time status information registry of all network elements, and sending the connection information of the target network element to the source network element; in the initial state, the regional DNS receives a preset connection optimization scheme issued by the root DNS, and finds the optimal connection scheme based on this preset connection optimization scheme using the following formula.

[0022]

[0023] In this formula, π'(s) is the better optimization scheme under state s, and a is the action under state s. The regional DNS performs reinforcement learning by periodically broadcasting the states of all network elements by the root DNS to find the optimal optimization scheme for action a under state s. The source network element establishes a signaling connection with the optimal target network element based on the connection information, and uses the second-best target network element as a backup signaling connection.

[0024] According to an advantageous design of the present invention, the source network element receives the target network element's area identifier, network address, port number, service identifier, and first certificate generated by the area DNS, and then combines the network address, port number, and first certificate into a header data block, adds signaling request data, encapsulates it into a signaling request data packet, and sends it to the target network element.

[0025] According to an advantageous design of the present invention, after receiving a signaling request data packet, the target network element verifies the first certificate. After passing the verification, it processes the signaling request to generate response data, adds its own second certificate to generate a response data packet, and sends the response data packet to the source network element.

[0026] According to the advantageous design of the present invention, after the source network element receives the response data packet, it verifies the second certificate. After passing the verification, it extracts the signaling data and hands it over to the network element instance for processing. The network element instance determines whether it is the corresponding response signaling requested by itself. If so, the connection has been established and the regional DNS is notified.

[0027] According to an advantageous design of the present invention, when the connection between the source network element and the optimal target network element is interrupted, the source network element establishes a signaling connection with the suboptimal target network element and sends the connection interruption information to the regional DNS.

[0028] According to the advantageous design of the present invention, after receiving the connection interruption information, the regional DNS determines the suboptimal target network element as the new optimal target network element, and reselects the suboptimal target network element through the real-time status information registry of all network elements, and sends the new suboptimal connection information to the source network element.

[0029] According to an advantageous design of the present invention, after the source network element fails to communicate with the regional DNS of its own region, it requests information from the neighboring regional DNS from the root DNS and establishes a signaling connection with the target network element through the neighboring regional DNS.

[0030] According to an advantageous design of the present invention, when the domain name resolution burden of a regional DNS exceeds a threshold, it requests the neighboring regional DNS to share the domain name resolution requests within its own region from the root DNS. The source network element establishes a signaling connection with the target network element through the neighboring regional DNS.

[0031] According to another aspect of the present invention, an electronic device is provided, comprising: at least one processor; and a memory storing at least one program, wherein when the at least one program is executed by the at least one processor, the at least one processor implements the method according to the present invention.

[0032] According to another aspect of the invention, a computer-readable storage medium is provided, on which a computer program is stored, which, when executed by a processor, implements the method according to the invention.

[0033] Specifically, the present invention includes the functions and structure of the signaling DNS network element and the fast TCP connection mechanism for network elements supported by the signaling DNS (S-DNS).

[0034] The Signalling DNS (S-DNS) network element provides the source with the desired target's domain name (a domain name pattern of network element name@zone name or information pattern that identifies the target, such as a code number). The Signalling DNS resolves the target's network information (network address, port number) and sends it to the target. The Signalling DNS network element achieves dynamic information acquisition of network elements through a mechanism of periodically reporting information from each network element in its zone and receiving periodic broadcasts of network element status from across the network. This eliminates the need for the NRF (Network Request for Information) reporting mode, making the dynamic information of network element status more efficient and comprehensive, while also providing conditions for connection management and optimization. When a source network element needs to find a target, it sends a resolution request to the local Signalling DNS network element in the form of a domain name. The Signalling DNS network element, based on its local dynamic registry and through an optimization algorithm, provides a primary connection and a backup connection that meet the optimization target. The connection includes the target's address and port, and is returned to the source as a response to the source's request. The source network element uses the certificate-based fast connection management mode in the Signalling DNS mechanism to conduct signaling communication with the target.

[0035] The signaling DNS is structured in two layers: the root DNS network element and the zone DNS network element. The root DNS has a fixed IP address and port, and is deployed across the network with two geographically separated primary and backup nodes, each configured with a fixed primary IP address and backup IP address. When the primary IP address fails, the system switches to the backup IP address. The zone DNS's IP address is dynamic, dynamically resolved by the root DNS based on the zone DNS's name.

[0036] The functions of the root DNS include: 1) Zone DNS management, namely, the registration, deregistration, and health management of zone DNS; 2) Network element status information forwarding, namely, each zone's network elements periodically send their status to their respective zone DNS, and the zone DNS summarizes the received network element status information and periodically sends it to the root DNS. After receiving the network element status parameters from all zone DNS, the root DNS encapsulates, cleans, filters, and packages them to form the network-wide network element status information, which is periodically broadcast to all zone DNS. After receiving the network-wide status information, the zone DNS updates its local registry (network element status registry) that records real-time information of all network elements, for use in network element domain name resolution, connection management, and fault handling; 3) Zone DNS domain name resolution, namely, the number, address, and port of zone DNS are dynamically configured, and the root DNS is responsible for the resolution of zone DNS domain names. The root DNS registry stores the addresses and status information of all zone DNS registered with the root DNS, and periodically updates the real-time status based on the zone DNS's heartbeat information. When a network element needs to query the IP address and port of its local DNS, it sends a domain name query request to the root DNS. Upon receiving the request, the root DNS looks up the IP address and port of the corresponding local DNS in the registry and sends it to the requesting network element. The requesting network element caches the local DNS address locally and communicates directly with its local local DNS in subsequent calls. If the local local DNS does not respond, it sends another domain name resolution request to the root DNS.

[0037] The functions of a regional DNS include: 1) Network element domain name resolution, i.e., the regional DNS provides domain name resolution services for network elements within the region. Source network elements within the region provide the target network element's domain name information to the regional DNS network element according to their business logic. This domain name information includes two types: one is a domain name pattern of network element name@region name, and the other is a context information pattern that identifies the target. The regional DNS network element, based on the received domain name resolution request information, searches for a list of target network elements that meet the requirements. Then, according to a multi-dimensional optimization algorithm, it selects the optimal and second-best two connections (primary connection and backup connection) as a response and sends it back to the source network element. Both the primary and backup connections contain the actual network element address information, such as the target network element's network address and port number, thus completing the conversion from the target network element's domain name to its actual network address. After receiving the domain name resolution response from the regional DNS, the source network element establishes a connection with the source network element using a fast connection mode to communicate signaling; 2) Regional network element health management, i.e., implemented through a mechanism where network elements within the region periodically send heartbeat information to the regional DNS. Heartbeat information includes the area identifier where the network element is located, the network element name, the network element identifier, the network address, the network port, the device parameters (device function code, processor configuration, storage configuration, network bandwidth, maximum service processing capacity, energy efficiency ratio), and the real-time working status of the device (processor utilization, storage resource utilization, network bandwidth utilization, number of connections, number of processes, number of services processed, energy consumption); 3) Network element status information reporting, that is, the area DNS will summarize the status information of the network elements in the area it receives and periodically send it to the root DNS, reporting the information of the network elements in the area, so that the root DNS can form the whole network. 4) Dynamic updates of network element information: After the regional DNS periodically receives the network status information broadcast by the root DNS, it updates the network element status data in the local registry for dynamic domain name resolution and connection management of the regional network element; 5) Network element connection management: The local source network element sends a connection request to the regional DNS to request a connection with the target network element. The regional DNS returns the response data of the connection request to the source network element using the fast signaling TCP mode. The requesting source network element establishes a connection with the target using the fast signaling TCP mode according to the connection parameters in the response.After a connection is successfully established, the source network element sends a connection establishment confirmation message to the regional DNS. The regional DNS saves this connection in the primary connection column of the connection queue. Simultaneously, it calculates a backup connection using a multi-dimensional optimization algorithm, places it in the backup connection column, and dynamically updates the backup connection column based on the network element's status. 6) Connection failure handling: The regional DNS's connection management maintains both primary and backup fast TCP connection information. If a network element fails to receive a response during communication, it retryes once and requests a backup connection from the regional DNS. The second retry uses the backup connection. Upon receiving a response, it switches to the backup connection and notifies the regional DNS to replace the backup connection information in the primary connection column. Then, it recalculates the backup connection information, thus achieving seamless connection switching during failures. 7) Dynamic connection updates: The regional DNS periodically receives broadcasts from the root DNS regarding all network elements. The system collects status data and dynamically labels the performance status information of the primary and backup connection columns based on the received network-wide status data. When the performance difference between the primary and backup connections exceeds the threshold set by the system, a connection switching request is sent to the source network element in this region. After the source network element in this region completes the connection switching, it notifies the regional DNS to replace the backup connection column information of the connection with the primary connection column, and then recalculates the backup connection column information to achieve seamless connection switching; 8) Connection optimization, that is, based on the dynamic backup connection of the regional DNS, multi-dimensional optimization technology is used to dynamically update the backup connection and dynamically switch the primary connection; 9) Mutual backup, that is, the registry of the regional DNS stores the status information of all network elements broadcast by the root DNS. When the regional DNS of a certain region fails, the network elements of that region change to request services from the regional DNS of the adjacent region, realizing mutual backup of multiple regional DNS in adjacent regions.

[0038] The core technologies of the network element fast TCP connection mechanism supported by Signaling DNS (S-DNS) include the network element fast TCP connection mechanism algorithm, the certificate body of fast TCP connection, and the connection management optimization algorithm.

[0039] Regarding the fast TCP connection mechanism algorithm supported by Signalling DNS (S-DNS), since the regional DNS stores real-time dynamic information of all network elements, and based on the high reliability of 5GC links and network elements, the fast TCP connection mechanism adopts a model where the regional DNS assigns a certificate to each connection. The certificate carries information about the client (source) network element, making it easier for the server (target) to distinguish between different clients. Because the source and target network elements in 5GC themselves have a request-response working mode, data can be directly transmitted through the fast connection mechanism, further reducing TCP connection time and enabling fast TCP connections and dynamic connection switching between network elements.

[0040] The specific connection algorithm of the network element fast TCP connection mechanism is as follows: 1) The source network element initiates a domain name resolution request to the regional DNS, carrying target information and subsequent business information (such as the target's region, service network element type, and the code number of this service); 2) The regional DNS extracts the target's region and network element type based on the information carried in the source's domain name resolution request, and searches for all targets that meet the requirements in the local network element registry; 3) The regional DNS finds the optimal target network element information based on the status markers in the target list, including the region identifier, network address, port number, service identifier, and the information of the second-best target network element, and then adds the certificate generated by the regional DNS for this connection, returning it to the source network element as a response message to the source network element's request; 4) The source network element, based on the response message received from the regional DNS, assembles the target's network address, port number, and certificate into a header data block, adds the first set of signaling request data, encapsulates it into a data packet, and sends it to the target network element; 5) The target network element receives the connection request data packet from the source. Afterwards, the certificate is extracted and verified to determine the legitimacy of the request. If it is legitimate, the source area identifier, network address, and port number are obtained, and signaling data is received. The signaling data is processed, and the corresponding signaling response data is added to the target certificate, encapsulated into a response data packet, and sent as response data to the source network element; 6) After receiving the data packet, the source network element extracts the certificate for verification. If it is legitimate, the signaling data is extracted and handed over to the signaling processing instance for processing. The signaling processing instance determines whether it is the response signaling corresponding to the corresponding signaling request in the signaling process. If it is the expected response signaling, the connection has been established, and the area DNS is notified; 7) If the source network element does not receive response data within the set time, it uses the backup connection provided by the area DNS to send a data request data packet to the target host address and port of the backup connection. If the correct response signaling is received, the backup connection is set as the primary connection, and the area DNS is notified. The area DNS removes the primary connection from the list, sets the backup connection as the primary connection, and recalculates the backup connection.

[0041] Regarding the certificate system for fast TCP connections in Signalling DNS (S-DNS), a Certificate Authority (CA) is deployed in the root DNS. This CA serves as a trusted and authoritative security center for the entire network, responsible for authenticating and verifying the legitimacy of public keys, acting as the "root certificate," and confirming the legitimacy of regional DNS certificates. Regional DNS is responsible for generating, distributing, and managing the public key certificates required for signaling connections between network elements within the same region. These certificates can prove connection requests between network elements in different regions, possessing the public key listed in the certificate and ensuring that third parties cannot forge or tamper with the certificate.

[0042] The certificate contains the source network element's area identifier, device identifier, network addresses and port numbers for data transmission and reception, and the public key belonging to this network element, and is digitally signed by the CA of the area DNS. The target network element uses the public key certificate to confirm that the area DNS of the target network element recognizes that the public key indeed belongs to the source network element, allowing for subsequent signaling processing. The target network element's certificate is also attached when sending the response. Similarly, the source network element can verify that the response originated from the peer network element that sent the previous service request based on the certificate.

[0043] Regarding the optimization algorithm for signaling DNS connection management, the regional DNS periodically receives data packets broadcast by the root DNS showing the status of all network elements, updates its local network element status table, and selects the optimal and suboptimal connections for similar network elements within different regions based on the status table. Since optimization needs to address multiple objectives, including service balancing among similar network elements in pool mode, network element processing capacity, and energy consumption, a multi-network approach using deep learning is employed for multi-dimensional optimization. Each network completes optimization for one dimension, and communication between different dimensions enables collaborative optimization, thereby achieving joint optimization of multiple objectives.

[0044] Each dimension is evaluated using a value function. The specific algorithm is as follows: 1) The root DNS collects connection optimization schemes from the regional DNS and distributes them to each regional DNS. Initially, the root DNS distributes a preset optimization scheme; 2) Each region uses a deep qualitative network (DQN) to evaluate different optimization schemes; 3) The regional DNS uses an action value function to evaluate different optimization schemes, and the results of the optimization schemes are judged by the regional DNS based on local network element dynamic information; 4) The regional DNS performs optimization for each dimension; 5) The regional DNS uses the states of all network elements periodically broadcast by the root DNS to perform reinforcement learning through functions of different dimensions, and finds the optimal optimization scheme for actions under each state.

[0045] In summary, with the development of 5G networks and their continuous expansion, signaling between network elements places higher demands on connection management, latency, and reliability. Simultaneously, the evolution of the signaling system to construct a more intelligent network has become an important direction for 5G network evolution. This invention, based on the existing 5G core network architecture, constructs a new signaling resolution and connection management model through a domain name resolution system, which has positive significance for promoting the development of 5G networks and future 6G networks. Attached Figure Description

[0046] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0047] Figure 1 This is a schematic diagram of a DNS domain name resolution system suitable for 5GC architecture constructed using the method of the present invention. Detailed Implementation

[0048] To enable those skilled in the art to better understand the technical solutions of the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0049] Figure 1 This is a DNS domain name resolution system suitable for 5GC architecture constructed according to the method of the present invention.

[0050] The method according to the present invention includes the following steps: establishing a two-layer signaling DNS structure including a root DNS and regional DNS; the root DNS establishing a real-time status information registry of all network elements and a domain name resolution table of all regional DNS, and the root DNS periodically broadcasting the real-time status information registry of all network elements to all regional DNS; the regional DNS recording the real-time status information registry of all network elements, and the regional DNS periodically collecting the network element status information of network elements in its own region, and sending the aggregated network element status information of network elements in its own region and its own status information to the root DNS; the source network element establishing a signaling connection with the target network element through the regional DNS of its own region, and initiating a domain name resolution request to the regional DNS, the request carrying the target network element information and subsequent service information; the regional DNS selecting the optimal and second-best target network elements according to the domain name resolution request through the real-time status information registry of all network elements, and sending the connection information of the target network element to the source network element; in the initial state, the regional DNS receives a preset connection optimization scheme issued by the root DNS, and finds the optimal connection scheme based on this preset connection optimization scheme using the following formula.

[0051]

[0052] In this formula, π'(s) is the better optimization scheme under state s, and a is the action under state s. The regional DNS performs reinforcement learning by periodically broadcasting the states of all network elements by the root DNS to find the optimal optimization scheme for action a under state s. The source network element establishes a signaling connection with the optimal target network element based on the connection information, and uses the second-best target network element as a backup signaling connection.

[0053] According to an advantageous design of the present invention, the source network element receives the target network element's area identifier, network address, port number, service identifier, and first certificate generated by the area DNS, and then combines the network address, port number, and first certificate into a header data block, adds signaling request data, encapsulates it into a signaling request data packet, and sends it to the target network element.

[0054] According to an advantageous design of the present invention, after receiving a signaling request data packet, the target network element verifies the first certificate. After passing the verification, it processes the signaling request to generate response data, adds its own second certificate to generate a response data packet, and sends the response data packet to the source network element.

[0055] According to the advantageous design of the present invention, after the source network element receives the response data packet, it verifies the second certificate. After passing the verification, it extracts the signaling data and hands it over to the network element instance for processing. The network element instance determines whether it is the corresponding response signaling requested by itself. If so, the connection has been established and the regional DNS is notified.

[0056] According to an advantageous design of the present invention, when the connection between the source network element and the optimal target network element is interrupted, the source network element establishes a signaling connection with the suboptimal target network element and sends the connection interruption information to the regional DNS.

[0057] According to the advantageous design of the present invention, after receiving the connection interruption information, the regional DNS determines the suboptimal target network element as the new optimal target network element, and reselects the suboptimal target network element through the real-time status information registry of all network elements, and sends the new suboptimal connection information to the source network element.

[0058] According to an advantageous design of the present invention, after the source network element fails to communicate with the regional DNS of its own region, it requests information from the neighboring regional DNS from the root DNS and establishes a signaling connection with the target network element through the neighboring regional DNS.

[0059] According to an advantageous design of the present invention, when the domain name resolution burden of a regional DNS exceeds a threshold, it requests the neighboring regional DNS to share the domain name resolution requests within its own region from the root DNS. The source network element establishes a signaling connection with the target network element through the neighboring regional DNS.

[0060] According to another aspect of the present invention, an electronic device is provided, comprising: at least one processor; and a memory storing at least one program, wherein when the at least one program is executed by the at least one processor, the at least one processor implements the method according to the present invention.

[0061] Furthermore, the implementation of the technical solution of the present invention is described in detail. The signaling connection method based on DNS in a 5G network disclosed in the present invention includes the function and structure of the signaling DNS (S-DNS) network element and the fast TCP connection mechanism of the network element supported by the signaling DNS (S-DNS).

[0062] Regarding the functions and structure of the Signaling DNS (S-DNS) network element, the 5GC architecture introduces NRF (Network Repository Function). The signaling process for network element connections based on NRF is as follows: When the target network element starts up, it registers its own code information with the home NRF for service. The NRF stores the mapping between the code and the registered network element instance, completing the registration function of the target network element. The source sends a service discovery request message to the NRF (containing the required query parameters, such as the source network element type, the target network element type, the services required by the target, and the user SUPI required to match the target network element). After receiving the service discovery request, the NRF queries the list of target network elements that meet the requirements (the IP address corresponding to the device) based on the SUPI and returns it to the requesting source network element. After receiving the request, the source network element selects an appropriate network element to initiate a connection. In cross-domain applications, the NRF supports registering information to another NRF, which is used to forward or redirect service discovery requests.

[0063] Compared to NRF, Signaling DNS (S-DNS) network elements achieve service discovery through DNS resolution. The function is described as follows: The source provides the desired target's domain name (a domain name pattern of network element name@zone name or information pattern that identifies the target, such as a code number). The Signaling DNS resolves the target's network information (network address, port number) and sends it to the target. The Signaling DNS network element achieves dynamic information acquisition of network elements through a mechanism of periodically reporting information from each network element in its zone and receiving periodic broadcasts of network element status from across the network. This eliminates the network element reporting mechanism of the NRF model, making the dynamic information of network element status more efficient and comprehensive, while also providing conditions for connection management and optimization. When a source network element needs to find a target, it sends a resolution request to the local Signaling DNS network element in the form of a domain name. The Signaling DNS network element, based on its local dynamic registry and through an optimization algorithm, provides a primary connection and a backup connection that meet the optimization target. Both the primary and backup connections include the target's address and port, which are returned to the source as a response to the source's request. The source network element uses the certificate-based fast connection management mode in the signaling DNS mechanism to communicate with the target via signaling, thereby reducing connection latency and improving the efficiency of the entire signaling network.

[0064] The signaling DNS is structured in two layers: the root DNS and the zone DNS. The root DNS has a fixed IP address and port, and is deployed across the network with two geographically separated primary and backup nodes, each configured with a fixed primary IP address and backup IP address. When the primary IP address becomes unavailable, the system switches to the backup IP address. The zone DNS's IP address is dynamic, dynamically resolved by the root DNS based on the zone DNS's name.

[0065] The functions of the root DNS include: Zone DNS management, where zone DNS registers with the root DNS service upon startup and deregisters upon shutdown. The root DNS maintains zone DNS through the registry. The root DNS dynamically manages the status of zone DNS through health checks. Health management is implemented through a heartbeat mechanism of the zone DNS. The heartbeat mechanism involves the zone DNS periodically sending a data packet to the root DNS and then starting a thread to continuously check for the root DNS's response. If no response is received from the root DNS within a set time, a heartbeat is simultaneously sent to the root DNS's backup address, and the response from the backup address is checked. If a response is received from the primary address while forwarding to the backup address, sending to the backup address stops. If the root DNS does not receive heartbeat information from the zone DNS within a set period, the corresponding zone DNS is deleted from the registry. Network element status information forwarding: all network elements in each area of ​​the 5G network periodically send their status parameters to the zone DNS of their respective area. The zone DNS receives these parameters. After the status information of the regional network elements is collected, it is periodically sent to the root DNS. After receiving the network element status parameters from all regional DNS, the root DNS encapsulates and cleans the data. The main tasks are to collect and sort the data according to the region to which it belongs and the category of the network elements within the region, simplify and filter the data (for example, data with the same pattern in several periods is sent only once in one period), remove data with errors after verification, remove data that does not conform to consistency, encapsulate all data (encrypt the data blocks, add header information and check information at the end), and form the status information of the entire network. After the above processing, the root DNS forms the network element status information of the entire network and periodically broadcasts it to all regional DNS in the entire network. After receiving the overall network status information, the regional DNS updates its local registry (network element status registry) which records real-time information about all network elements. This registry is used for network element domain name resolution, connection management, and fault handling. Regional DNS domain name resolution is crucial because the structure of the 5G network is dynamically changing as it is built and evolved. The number and structure of regional DNS servers are also constantly evolving. To meet the needs of this dynamic network structure, the number, address, and port of regional DNS servers are dynamically configured. The root DNS is responsible for resolving regional DNS domain names. The root DNS registry stores the addresses and status information of all regional DNS servers registered with it and periodically updates their real-time status based on the regional DNS's heartbeat information. When a network element needs to query the IP address and port of its regional DNS server, it sends a domain name query request to the root DNS server. Upon receiving the query request, the root DNS server looks up the IP address and port of the regional DNS server corresponding to the domain name in the registry and sends it to the requesting network element. The requesting network element caches the regional DNS address locally and subsequently communicates directly with the local regional DNS server through this cache.If the local zone DNS does not respond, the request for domain name resolution is sent to the root DNS again, repeating the above process. The message sent by the querying network element to the root DNS for the zone DNS query request is as follows: {Message header + destination address + source address + zone domain name + querying network element network identifier + querying network element certificate + check field}. After receiving the query request, the root DNS verifies the network element's network identifier and corresponding certificate, extracts the query domain name, queries the local registry for the IP address and port number corresponding to the domain name, and the additional description information of the zone DNS, as the DNS query result. The query result is then encrypted using the querying network element's certificate. Finally, the destination address and source address are swapped to form the response data, which is sent to the querying network element. The root DNS's response message is as follows: {Message header + destination address + source address + DNS query result + root DNS network identifier + certificate + check field}.

[0066] The functions of a zone DNS include: domain name resolution for network elements, that is, providing domain name resolution services for network elements within the zone. Source network elements within the zone provide the zone DNS network element with the domain name information of the target network element according to their business logic. This domain name information includes two types: one is a domain name pattern of network element name@zone name, and the other is a contextual information pattern that identifies the target, such as the SUPI code resource. Since the zone DNS registry stores dynamic information about all network elements across the entire network, periodically broadcast by the root DNS, the zone DNS network element searches for a list of target network elements that meet the requirements based on the received domain name resolution request information. Then, according to a multi-dimensional optimization algorithm, it selects the optimal and second-best two connections (primary connection and backup connection) as responses and sends them back to the source network element. Both the primary and backup connections contain the actual network element address information, such as the network address and port number of the target network element, thus completing the conversion from the domain name of the target network element to its actual network address. After receiving the domain name resolution response from the regional DNS, the source network element establishes a connection with the source network element using the fast connection mode to conduct signaling communication; regional network element health management, that is, the regional DNS obtains the real-time status of all network elements in the region through the health management of regional network elements, which is achieved through the mechanism of network elements in the region periodically sending heartbeat information to the regional DNS. Heartbeat information includes the area identifier, network element name, network element identifier, network address, network port, device parameters (device function code, processor configuration, storage configuration, network bandwidth, maximum service processing capacity, energy efficiency ratio), and real-time device operating status (processor utilization, storage resource utilization, network bandwidth utilization, number of connections, number of processes, number of services processed, energy consumption); network element status information reporting, that is, the area DNS summarizes the status information of network elements received within the area and periodically sends it to the root DNS, reporting the information of network elements in the area, so that the root DNS can form the network element status information of the entire network; dynamic updating of network element information, that is, after the area DNS periodically receives the network status information broadcast by the root DNS, it updates the status data of network elements in the local registry, which is used for the dynamic resolution of domain names and connection management of the entire network of the area network elements; network element connection management, that is, the signaling between network elements in 5GC SA mode uses the HTTP 2.0 protocol for communication and supports multiplexing.Multiple requests can be executed in parallel on a single connection (multi-stream parallelism, with different requests identified by different IDs). It also has server-side push capabilities. There are two signaling addressing methods between network elements: addressing of local network elements independent of user IDs (e.g., AMF addressing SMF); and addressing of cross-province / regional network elements related to user IDs (e.g., AMF addressing UDM). Since the regional DNS stores dynamic information on the status of all network elements, it can obtain the status data of all network elements for both addressing methods. When a local source network element sends a connection request to the regional DNS to establish a connection with the target network element, the regional DNS returns the connection request response data to the source network element using fast signaling TCP mode. The requesting source network element then establishes a connection with the target using fast signaling TCP mode based on the connection parameters in the response. After a connection is successfully established, the source network element sends a connection establishment confirmation message to the regional DNS. The regional DNS saves this connection in the primary connection column of the connection queue and simultaneously calculates a backup connection using an optimization algorithm, placing it in the backup connection column. The backup connection column is dynamically updated based on the network element's status. For connection failure handling, since there are a large number of TCP connections between network elements, a failure in one network element will trigger TCP link alarms on all connected network elements, making maintenance and management difficult. The regional DNS connection management maintains both primary and backup fast TCP connection information. When a network element sends a message but does not receive a response, it retryes once and requests a backup connection from the regional DNS. The second retry uses the backup connection. Upon receiving a response, it switches to the backup connection and notifies the regional DNS to replace the backup connection information in the primary connection column. Then, it recalculates the backup connection information, thus achieving seamless connection switching during failures. Dynamic connection updates are also implemented. In other words, because the status of target network elements is constantly changing, such as device resource utilization and real-time processing capabilities, a large number of inefficient connections will appear in the system if the connection duration is long, thus affecting the performance of the entire system. The regional DNS periodically receives the status data of all network elements broadcast by the root DNS and dynamically marks the performance status information of the primary and backup connection columns based on the received network status data. When the performance difference between the primary and backup connections exceeds the system-set threshold, a connection switching request is sent to the source network element in the region. After the source network element in the region completes the connection switching, it notifies the regional DNS to replace the backup connection column information of the connection with the primary connection column, and then recalculates the backup connection column information to achieve seamless connection switching. Connection optimization, that is, because the network elements in the network are handling various services in a dynamic manner, for the Pool working mode (multiple network elements share the load proportionally or N+1 backup), dynamic load balancing can improve the processing capacity of network elements. Based on the dynamic backup connection of the regional DNS, multi-dimensional optimization technology is used to dynamically update the backup connection and dynamically switch the primary connection.

[0067] Zone DNS plays a crucial role in the normal operation of the entire zone and has extremely high reliability requirements. However, because the zone DNS registry stores the status information of all network elements broadcast by the root DNS, when the zone DNS of a certain zone fails, network elements in that zone can request services from the zone DNS of neighboring zones. Since there are multiple neighboring zones, these zone DNS systems can act as backups for each other, improving system reliability while reducing costs.

[0068] Next, we will further describe the fast TCP connection mechanism for network elements supported by Signaling DNS (S-DNS).

[0069] In 5G's SA core network, network elements connect using the HTTP / 2.0 communication protocol. Compared to HTTP / 1.x, HTTP / 2.0 offers more efficient transmission performance and supports multiplexing. The underlying HTTP protocol establishes connections via TCP. The TCP communication process includes three steps: establishing a TCP connection channel, transmitting data, and closing the TCP connection channel. Establishing a TCP connection involves a three-way handshake, as follows: First handshake: The client sends a SYN packet (seq = x) to the server and enters the SYN_SEND state, waiting for server confirmation. Second handshake: Upon receiving the SYN packet, the server must acknowledge the client's SYN (ack = x + 1) and simultaneously send its own SYN packet (seq = y), i.e., a SYN+ACK packet. At this point, the server enters the SYN_RECV state. Third handshake: Upon receiving the server's SYN+ACK packet, the client sends an ACK packet (ack = y + 1) to the server. After this packet is sent, both the client and server enter the ESTABLISHED state, completing the three-way handshake.

[0070] The packets transmitted during the handshake do not contain data. After the three-way handshake is complete, the client and server officially begin transmitting data. Ideally, once a TCP connection is established, it will remain open until either party actively closes the connection.

[0071] The TCP three-way handshake is primarily used to prevent the server from opening unnecessary connections, increasing server overhead, and to prevent expired connection request segments from being unexpectedly transmitted to the server, thus causing errors. Therefore, a "third handshake" is needed to confirm this process, allowing both the client and server to promptly detect connection creation failures caused by network issues, so that the server's port can be closed instead of waiting indefinitely.

[0072] TCP's three-way handshake is designed for unreliable networks. While it increases reliability, it reduces transmission efficiency, especially for short-connection protocols like HTTP. The 5GC network environment is a highly reliable, low-latency network, and the large number of HTTP signaling connections based on TCP's three-way handshake would significantly degrade overall system performance. TCP Fast Open (TFO) is a fast TCP protocol designed to bypass the three-way handshake, improving efficiency, or in other words, completing data transmission simultaneously with the handshake, resulting in a significant improvement in HTTP efficiency.

[0073] Compared to the traditional TCP three-way handshake mechanism, TFO clients include the first data packet sent in the final ACK, which is one of TFO's optimizations. Additionally, TFO references the HTTP login state process, using cookies to let the server know that a client has previously "logged in," allowing direct reception of its data without requiring another three-way handshake. When a client first connects to the server in TFO, it doesn't have a cookie, so it sends an empty cookie to request one from the server. The server returns the cookie to the client via the SYN+ACK path. The client saves this cookie and simultaneously sends the final ACK of the three-way handshake to the server. When the client disconnects and requests the same server again, it includes the previously stored cookie and the data to be sent, sending them together via the SYN path. Subsequent handshakes involve sending data information simultaneously, allowing for earlier data transmission. The server receives the data only if it verifies the cookie; otherwise, it discards the data and returns a new cookie via the SYN+ACK path (usually due to cookie expiration).

[0074] Because the regional DNS stores real-time dynamic information of all network elements, and based on the high reliability of 5GC links and network elements, the network element fast TCP connection mechanism adopts a certificate allocation model for each connection using the regional DNS. This reduces the cookie process in TFO, and the certificate carries information about the client (source) network element, making it easier for the server (target) to distinguish between different clients. Furthermore, since the source and target network elements in 5GC inherently operate in a request-response mode, data can be directly transmitted through the fast connection mechanism, further reducing TCP connection time and enabling fast TCP connections and dynamic connection switching between network elements.

[0075] The specific connection algorithm of the network element fast TCP connection mechanism is as follows: The source network element initiates a domain name resolution request to the regional DNS, carrying target information and subsequent business information (such as the target's region, service network element type, and the code number of this service); the regional DNS extracts the target's region and network element type based on the information carried in the source's domain name resolution request, and searches for all targets that meet the requirements in the local network element registry; the regional DNS finds the optimal target network element information based on the status flags in the target list, including the region identifier, network address, port number, service flag, and the information of the second-best target network element, and then adds the certificate generated by the regional DNS for this connection, returning it to the source network element as a response message to the source network element's request; the source network element, based on the response message received from the regional DNS, assembles the target's network address, port number, and certificate into a header data block, adds the first set of signaling request data, encapsulates it into a data packet, and sends it to the target network element; after receiving the connection request data packet from the source, the target network element... The system retrieves and verifies the certificate to determine the legitimacy of the request. If legitimate, it obtains the source's zone identifier, network address, and port number, and receives signaling data. The signaling data is processed, and the corresponding response data is appended with the target's certificate, encapsulated into a response data packet, and sent as response data to the source network element. Upon receiving the data packet, the source network element extracts and verifies the certificate. If legitimate, it extracts the signaling data and submits it to the signaling processing instance. The signaling processing instance determines if it corresponds to the expected response signaling request in the signaling process. If it is the expected response signaling, the connection is established, and the zone DNS is notified. If the source network element does not receive response data within a set time, it uses the backup connection provided by the zone DNS to send a data request data packet to the target host address and port of the backup connection. If a correct response signaling is received, the backup connection is set as the primary connection, and the zone DNS is notified. The zone DNS removes the primary connection from the list, sets the backup connection as the primary connection, and recalculates the backup connection.

[0076] The following example illustrates the implementation of the fast TCP connection mechanism for network elements supported by the Signaling DNS (S-DNS): The AMF of the N8 interface (AMF-UDM) obtains subscription data from the UDM after completing UE authentication. The network element status data packet broadcast by the root DNS to the entire network contains status information of all UDMs in each region, including host address, port number, code number information, and operating status. After receiving this data, the regional DNS periodically updates its local registry, which records the status of all network elements. The regional DNS stores the mapping relationship between code numbers and UDMs. The AMF sends a domain name resolution request, which includes a SUPI. The regional DNS uses the SUPI to calculate a list of UDMs that meet the requirements based on the mapping relationship between code numbers and UDMs. Then, it uses an optimization algorithm to determine the UDM information (region identifier, address, port number) for the primary and backup connections, and adds this information to the connection generated by the regional DNS. The certificate is returned to the AMF network element as a response message to the AMF domain name resolution request. Based on the response message received from the regional DNS, the AMF combines the network address and port number of the main connection UDM, along with the certificate, into a header data block, adds the UE's context (message name: Nudm_UEContextManagement_Registration), encapsulates it into a data packet, and sends it to the main connection UDM. Upon receiving the connection request data packet from the AMF, the UDM extracts the certificate for verification to determine the validity of the request. If valid, the UDM extracts the Nudm_UEContextManagement_Registration message. If the UE's registration information exists, the UDM replaces the previous registration information with the received amf-3gpp-access and returns a 200 OK or 204 NO Context response. The UDM then calls Nudm_UEContextManagent_DerigistrationNotify to notify the old AMF to delete the UE Context. If the UE's registration information is not available, the received information will be saved and a 201 Created response will be returned. After the AMF obtains the UDM's response data, it will obtain the UE's subscription data and notify the regional DNS that the connection has been established. If no response data is received from the UDM within the set time, the first retry will use the UDM information (network address, port number) of the backup connection provided by the regional DNS to resend the connection request data packet. After receiving the response data, the regional DNS will be notified to replace the primary connection with the backup connection, and the regional DNS will recalculate the backup connection.

[0077] A Certificate Authority (CA) is deployed in the root DNS to serve as a trusted and authoritative security center for the entire network. It is responsible for authenticating and verifying the legitimacy of public keys, acting as the "root certificate," and validating the legitimacy of regional DNS keys. The root DNS stores the key pairs for the entire network and issues key pairs and certificates to regional DNS servers. Regional DNS servers are responsible for generating, distributing, and managing the public-key certificates (PKCs) required for signaling connections between network elements within their respective regions. These certificates can prove that connection requests between network elements in different regions possess the public keys listed in the certificate, ensuring that third parties cannot forge or tamper with the certificates.

[0078] The certificate contains the source network element's area identifier, device identifier, network address and port number for data transmission and reception, and the public key belonging to this network element, and is digitally signed by the CA of the area DNS. By using the public key certificate, the target network element can verify that the area DNS of that network element recognizes the public key as belonging to the source network element, and can then process subsequent signaling by attaching the target network element's certificate to its response. Similarly, the source network element can use the certificate to confirm that the response originated from the peer network element that sent the previous service request.

[0079] The regional DNS periodically receives data packets broadcast by the root DNS showing the status of all network elements, updates its local network element status table, and selects the optimal and suboptimal connections for similar network elements within different regions based on the status table. Since optimization needs to address multiple objectives, including service balancing among similar network elements in pool mode, network element processing capacity, and energy consumption, a deep learning-based multi-Q network is employed for multi-dimensional optimization. Each Q network completes optimization for one dimension, and different dimensions communicate with each other to achieve collaborative optimization. For example, communication and collaboration between the service balancing Q network optimization and the energy consumption Q network optimization, thereby achieving joint optimization of multiple objectives.

[0080] Each Q-network is evaluated using a value function, and the specific algorithm is as follows:

[0081] The root DNS collects the connection optimization scheme from the zone DNS and distributes it to each zone DNS. Initially, the root DNS distributes a preset optimization scheme.

[0082] Since optimization schemes for other regional DNS may not be suitable for this region, a DeepQ-network (DQN) is used, where the regional DNS evaluates different optimization schemes.

[0083] The regional DNS uses an action-value function (Q-function) to evaluate different optimization schemes. The results of the optimization schemes are judged by the regional DNS based on local network element dynamic information. The input to the value function is a state-action pair. Taking a certain action in a certain state and using a certain optimization scheme (such as setting a few network element devices to sleep instead of starting all of them to distribute tasks evenly in a pool mode when the overall task is small) yields the expected value of the cumulative energy consumption (Q-function) and processing capacity (Q-function) rewards.

[0084] The regional DNS is optimized for each dimension using Equation 1.

[0085] Where π'(s) is the better optimization scheme under state s, and a is the action under state s;

[0086] The regional DNS uses Q-functions at different dimensions to perform reinforcement learning on the states of all network elements periodically broadcast by the root DNS, in order to find the optimal optimization scheme for action a under state s.

[0087] In summary, this application provides a method for establishing signaling connections based on DNS. This method constructs a DNS domain name resolution system suitable for the 5GC architecture. Network Functions (NFs) only need to provide the target network element's domain name information to the DNS network element, which can be the root DNS or a zone DNS. Based on the provided information, the DNS network element optimizes and selects the optimal and second-best connection information as responses to the source network element. The source network element can then use the DNS mechanism's fast connection mode to establish a connection with the target network element for signaling communication. When a connection is interrupted, a message is automatically sent to the target network element pointed to by the backup connection. If the target network element responds, the completed signaling records of the primary connection's target network element are packaged and sent. The target network element of the backup connection restores the completed signaling process, allowing both network elements to continue during a failure, reducing the impact of the failure on signaling communication. Since all zone DNS stores the network element state broadcast by the root DNS, if a zone DNS fails, it can be handled by a neighboring zone DNS, meaning adjacent zone DNS serve as backups for each other. This significantly reduces the reliability requirements of zone DNS and lowers costs. In addition, when the difference in DNS load between adjacent areas exceeds the set threshold, they can coordinate with each other (e.g., coordination between office areas and residential areas at different times), thereby improving the overall processing capacity of the system.

[0088] It will be understood by those skilled in the art that all or some of the steps, systems, or apparatuses disclosed above, and their functional modules / units, can be implemented as software, firmware, hardware, or suitable combinations thereof. In hardware implementations, the division between functional modules / units mentioned in the above description does not necessarily correspond to the division of physical components; for example, a physical component may have multiple functions, or a function or step may be performed collaboratively by several physical components. Some or all physical components may be implemented as software executed by a processor, such as a central processing unit, digital signal processor, or microprocessor, or as hardware, or as an integrated circuit, such as an application-specific integrated circuit (ASIC). Such software may be distributed on a computer-readable medium, which may include computer storage media (or non-transitory media) and communication media (or transient media). As is known to those skilled in the art, the term computer storage media includes volatile and non-volatile, removable and non-removable media implemented in any method or technology for storing information (such as computer-readable instructions, data structures, program modules, or other data). Computer storage media include, but are not limited to, RAM, ROM, EEPROM, flash memory or other memory technologies, CD-ROM, digital versatile disc (DVD) or other optical disc storage, magnetic cartridges, magnetic tape, disk storage or other magnetic storage devices, or any other medium that can be used to store desired information and can be accessed by a computer. Furthermore, it is well known to those skilled in the art that communication media typically contain computer-readable instructions, data structures, program modules, or other data in modulated data signals such as carrier waves or other transmission mechanisms, and may include any information delivery medium.

[0089] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for establishing a signaling connection based on DNS, the method constructing a DNS domain name resolution system suitable for 5GC architecture, comprising the following steps: Establish a two-tier signaling DNS structure including root DNS and zone DNS; The root DNS establishes a real-time status information registry of all network elements and a domain name resolution table of all regional DNS, and the root DNS periodically broadcasts the real-time status information registry of all network elements to all regional DNS. The regional DNS records the real-time status information of all network elements in the network. The regional DNS periodically collects the network element status information of the network elements in its own region and sends the network element status information of the network elements in its own region and its own status information to the root DNS. The source network element establishes a signaling connection with the target network element through the regional DNS of its own region, and initiates a domain name resolution request to the regional DNS, which carries the information of the target network element and subsequent service information. The regional DNS selects the optimal and second-best target network elements based on the domain name resolution request through the real-time status information registry of all network elements, and sends the connection information of the target network elements to the source network element. In the initial state, the regional DNS receives a preset connection optimization scheme from the root DNS, and finds the optimal connection scheme based on this preset connection optimization scheme using the following formula. In this formula, π'(s) is the better optimization scheme under state s, and a is the action under state s. The regional DNS performs reinforcement learning through the states of all network elements periodically broadcast by the root DNS to find the optimal optimization scheme for action a under state s. The source network element establishes a signaling connection with the optimal target network element based on the connection information, and uses the second-best target network element as a backup signaling connection.

2. The method according to claim 1, characterized in that, The source network element receives the target network element's area identifier, network address, port number, service identifier, and the first certificate generated by the area DNS, and then combines the network address, port number, and first certificate into a header data block, adds signaling request data, encapsulates it into a signaling request data packet, and sends it to the target network element.

3. The method according to claim 2, characterized in that, After receiving the signaling request data packet, the target network element verifies the first certificate. If the verification is successful, it processes the signaling request, generates response data, adds its own second certificate to generate a response data packet, and sends the response data packet to the source network element.

4. The method according to claim 3, characterized in that, After receiving the response data packet, the source network element verifies the second certificate. After passing the verification, it extracts the signaling data and hands it over to the network element instance for processing. The network element instance determines whether it is the corresponding response signaling for its own request. If so, the connection has been established and the region DNS is notified.

5. The method according to claim 1, characterized in that, When the connection between the source network element and the optimal target network element is interrupted, the source network element establishes a signaling connection with the second-best target network element and sends the connection interruption information to the regional DNS.

6. The method according to claim 5, characterized in that, After receiving the connection interruption information, the regional DNS determines the suboptimal target network element as the new optimal target network element, and reselects the suboptimal target network element through the real-time status information registry of all network elements, and sends the new suboptimal connection information to the source network element.

7. The method according to claim 1, characterized in that, After the source network element fails to communicate with the regional DNS of its own region, it requests information from the neighboring regional DNS from the root DNS and establishes a signaling connection with the target network element through the neighboring regional DNS.

8. The method according to claim 1, characterized in that, When the domain name resolution burden of the regional DNS exceeds a threshold, it requests the neighboring regional DNS to share the domain name resolution requests within its own region from the root DNS. The source network element establishes a signaling connection with the target network element through the neighboring regional DNS.

9. An electronic device, comprising: At least one processor; A memory storing at least one program that, when executed by the at least one processor, causes the at least one processor to implement the method of any one of claims 1 to 8.

10. A computer-readable storage medium having a computer program stored thereon, the program, when executed by a processor, implementing the method of any one of claims 1 to 8.

Images