Search

EP-4170990-B1 - INTENTION NEGOTIATION METHOD AND APPARATUS

EP4170990B1EP 4170990 B1EP4170990 B1EP 4170990B1EP-4170990-B1

Inventors

  • LI, XIANMING
  • YU, YIJUN
  • JU, Min

Dates

Publication Date
20260506
Application Date
20210728

Claims (13)

  1. An intent negotiation method in a hierarchical intent system scenario, wherein the method comprises: sending (S501), by a first network element, a first sub-intent of a first intent to a second network element; and sending (S502), by the first network element, a second sub-intent of the first intent and indication information to the second network element, wherein the second sub-intent is sent to the second network element after sending the first sub-intent, and the first sub-intent and the second sub-intent are sub-intents that are obtained by the first network element through two consecutive times of decomposition of the first intent and that are sent by the first network element at two consecutive times, and wherein the indication information is used to indicate that the first sub-intent and the second sub-intent have a same action and a same object, for a purpose that the second network element omits translation of the second sub-intent based on the indication information.
  2. The method according to claim 1, further comprising: sending, by the first network element, an estimated quantity of times to the second network element, wherein the estimated quantity of times is used to indicate a quantity of negotiation times of a second intent that has a same action and a same object as the first intent, and the quantity of negotiation times is used to indicate a quantity of decomposition times of the second intent in a process of executing the second intent.
  3. The method according to claim 1 or 2, further comprising: storing, by the first network element, a quantity of negotiation times of the first intent when the first network element determines that the first intent is achieved, or negotiation for the first intent times out, or the first intent is deleted.
  4. An intent negotiation method in a hierarchical intent system scenario, wherein the method comprises: receiving (S501), by an intent translation module in a second network element, a first sub-intent of a first intent from a first network element; receiving (S502), by the intent translation module, a second sub-intent of the first intent and indication information from the first network element, wherein the second sub-intent is received by the intent translation module after receiving the first sub-intent, and the first sub-intent and the second sub-intent are sub-intents that are obtained by the first network element through two consecutive times of decomposition of the first intent and that are sent by the first network element at two consecutive times, and wherein the indication information is used to indicate that the first sub-intent and the second sub-intent have a same action and a same object, which means that the second sub-intent does not need to be translated; and omitting (S503a), by the intent translation module, translation of the second sub-intent based on the indication information.
  5. The method according to claim 4, further comprising: determining, by the intent translation module, a target value of the second sub-intent based on the second sub-intent; and sending, by the intent translation module, an intent update request to a third network element in the second network element, wherein the intent update request comprises an identifier of a third sub-intent and the target value of the second sub-intent, and the third sub-intent is a 1 st sub-intent that is for the first intent and that is received by the intent translation module from the first network element.
  6. The method according to claim 4 or 5, further comprising: receiving, by the intent translation module, an estimated quantity of times from the first network element, wherein the estimated quantity of times is used to indicate a quantity of negotiation times of a second intent that has a same action and a same object as the first intent, and the quantity of negotiation times is used to indicate a quantity of decomposition times of the second intent in a process of executing the second intent.
  7. The method according to claim 6, further comprising: determining, by the intent translation module based on a quantity of policies comprised in an available execution policy set, that the estimated quantity of times is greater than the quantity of policies, wherein the available execution policy set is determined based on the third sub-intent; and sending, by the intent translation module, a global intent conflict detection request to a fourth network element in the second network element, wherein the global intent conflict detection request is used to request to detect, for each policy in the available execution policy set, whether a sub-intent that conflicts with the policy exists.
  8. An intent negotiation method in a hierarchical intent system scenario, wherein the method comprises: receiving (S501), by a policy management module in a second network element, a first sub-intent of a first intent from a first network element; receiving (S502), by the policy management module, a second sub-intent of the first intent and indication information from the first network element, wherein the second sub-intent is received by the policy management module after receiving the first sub-intent, and the first sub-intent and the second sub-intent are sub-intents that are obtained by the first network element through two consecutive times of decomposition of the first intent and that are sent by the first network element at two consecutive times, and wherein the indication information is used to indicate that the first sub-intent and the second sub-intent have a same action and a same object, which means that the second sub-intent does not need to be translated; and omitting (S503b), by the policy management module, translation of the second sub-intent based on the indication information.
  9. The method according to claim 8, further comprising: determining, by the policy management module, a target value of the second sub-intent based on the second sub-intent; and sending, by the policy management module, an intent update request to a third network element in the second network element, wherein the intent update request comprises an identifier of a third sub-intent and the target value of the second sub-intent, and the third sub-intent is a 1 st sub-intent that is for the first intent and that is received by an intent translation module from the first network element.
  10. The method according to claim 8 or 9, further comprising: receiving, by the policy management module, an estimated quantity of times from the first network element, wherein the estimated quantity of times is used to indicate a quantity of negotiation times of a second intent that has a same action and a same object as the first intent, and the quantity of negotiation times is used to indicate a quantity of decomposition times of the second intent in a process of executing the second intent.
  11. The method according to claim 10, further comprising: determining, by the policy management module based on a quantity of policies comprised in an available execution policy set, that the estimated quantity of times is greater than the quantity of policies, wherein the available execution policy set is determined based on the third sub-intent; and detecting, by the policy management module for each policy in the available execution policy set, whether a sub-intent that conflicts with the second sub-intent exists under a condition of the policy, and determining a conflict sub-intent set, wherein the conflict sub-intent set comprises a sub-intent that conflicts with at least one policy in the available execution policy set.
  12. The method according to claim 11, further comprising: determining, by the policy management module, a first policy in the available execution measurement set based on the available execution policy set and the target value of the second sub-intent when the second sub-intent is not achieved; determining, by the policy management module based on the conflict sub-intent set, a sub-intent that conflicts with the first policy, and invoking a preset method to resolve the sub-intent that conflicts with the first policy; and sending, by the policy management module, the first policy to a fourth network element in the second network element.
  13. An apparatus (1500, 1600), comprising a module configured to perform the method according to any one of claims 1 to 12.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS TECHNICAL FIELD Embodiments of this application relate to the field of network management technologies, and in particular, to an intent negotiation method and apparatus. BACKGROUND In February 2017, an intent-based network white paper was released. The white paper holds that an intent-based network can help reduce operating expense (operating expense, OPEX), build a simplified intelligent network, and support operation and maintenance of a network with massive connections. The white paper also proposes that an intent-based network system needs to provide four functions: translation and validation (translation and validation), automated implementation (automated implementation), awareness of a network state (awareness of network state), and dynamic optimization assurance and remediation (assurance and dynamic optimization/remediation). FIG. 1 shows a Gartner intent-driven network. In a hierarchical intent system scenario, after receiving an intent from a management personnel or a higher-level (level) intent-driven network, a high-level intent-driven network (high-level intent-driven networking, H-IDN) decomposes the intent and sends a sub-intent obtained through decomposition to a low-level intent-driven network (low-level intent-driven networking, L-IDN). Operation objects (for example, network elements) corresponding to the intent are managed by different L-IDNs, as shown in FIG. 2(a). After executing the intent, the L-IDN sends feedback information to the H-IDN. The H-IDN re-decomposes the intent based on feedback information reported by each L-IDN. A specific procedure is shown in FIG. 2(b). Step 1: The H-IDN receives the intent. Step 2: The H-IDN decomposes a target value of the intent into target values of sub-intents of L-IDNs. An optimization intent is used as an example. During intent decomposition, it only needs to be ensured that a comprehensive result (for example, a sum or an average value) of target values of sub-intents obtained through decomposition meets a target value of the intent. Therefore, a target value of each sub-intent may be any value. Steps 3 to 5: The H-IDN sends a plurality of sub-intents to the corresponding L-IDNs respectively. Steps 6 to 8: After receiving the sub-intent, the L-IDN executes the sub-intent, and sends feedback information to the H-IDN. If the sub-intent is not achieved, the feedback information may include a currently reached measurement value. Alternatively, if the sub-intent is achieved, the feedback information may include a measurement value that can be reached and that is estimated by the L-IDN based on a network state and historical data. Step 9: The H-IDN determines, based on the feedback information reported by each L-IDN, whether the intent is achieved. If the intent is not achieved, the H-IDN re-decomposes the intent, and determines a target value of an updated sub-intent of each L-IDN. If the intent is achieved, the procedure ends. Steps 10 to 12: The H-IDN sends a plurality of updated sub-intents to the corresponding L-IDNs respectively. Steps 13 to 15 repeat steps 6 to 8. Subsequently, steps 9 to 15 are repeated until the intent is deleted, or intent negotiation times out, or the intent is achieved. Step 16: The H-IDN reports that the intent is achieved. In this case, each sub-intent is achieved. Step 17: The H-IDN reports that the intent is not achieved. In this case, each sub-intent is not achieved. Alternatively, a part of the sub-intents are achieved, but the other part of the sub-intents are not achieved. It can be learned from the foregoing procedure that, an existing intent negotiation solution causes a large waste of resources on an L-IDN side. Specifically, in an intent negotiation process, a sub-intent delivered each time differs only in a target value, but the L-IDN needs to perform complete lexical and syntactic parsing on an intent expression of the sub-intent each time. In addition, the sub-intent delivered each time may correspond to a same available execution policy set in an intent knowledge repository. However, the L-IDN needs to perform intent translation each time, in other words, needs to determine the available execution policy set in the intent knowledge repository each time. Document "Experiential Networked Intelligence (ENI), Intent Aware Network Autonomicity", ETSI DRAFT SPECIFICATION, ENI 008, EUROPEAN TELECOMMUNICATIONS STANDARDS INSTITUTE (ETSI), 650, ROUTE DES LUCIOLES; F-06921 SOPHIA-ANTIPOLIS; FRANCE, vol. ISG ENI Experiential Networked Intelligence, no. V0.0.14, 4 June 2020 (2020-06-04), pages 1-27, XP014370603, discloses accepting, translating and validating intent statements, determining how intent affects the goals and operation of the ENI system, and how it is used by business users, application developers and network administrators. SUMMARY This application provides an intent negotiation method and apparatus, as defined in the appended set of claims, to resolve a problem