CN115033254A

CN115033254A - Application pre-deployment method, device, equipment and storage medium

Info

Publication number: CN115033254A
Application number: CN202210643285.5A
Authority: CN
Inventors: 王晴天; 刘洋; 邢燕霞; 陈鹏
Original assignee: China Telecom Corp Ltd
Current assignee: China Telecom Corp Ltd
Priority date: 2022-06-08
Filing date: 2022-06-08
Publication date: 2022-09-09

Abstract

The embodiment of the invention provides a pre-deployment method, a pre-deployment device, application pre-deployment equipment and a storage medium, and relates to the technical field of wireless communication. The specific implementation scheme is as follows: determining a service request received by Near-RT RIC in a specified historical time period; each service request is processed by calling one xAPP in each third-party application xAPP; determining the calling frequency of each xAPP in a specified historical time period based on the determined service request; predicting the calling probability corresponding to each xAPP at the target moment based on the calling frequency of each xPP in the appointed historical time period; selecting at least one xAPP to be deployed from each xAPP according to the calling probability corresponding to each xAPP at a target moment; and deploying the selected xAPP to be deployed in the Near-RT RIC. Therefore, the method and the device can improve the hit rate of the service request, thereby reducing the communication overhead of service processing.

Description

Application pre-deployment method, device, equipment and storage medium

Technical Field

The present invention relates to the field of wireless communication technologies, and in particular, to a method, an apparatus, a device, and a storage medium for pre-deployment of an application.

Background

The Open Radio Access Network (Open Radio Access Network) organization proposes a Radio Intelligent Controller (RIC) for RAN virtualization. RIC includes Non-Real-Time Radio interactive controller (Non-Real-Time intelligent controller) and Near-Real-Time Radio interactive controller (Near Real-Time intelligent controller). The Non-RT RIC is deployed in a node with abundant computing power, and an installation file of each xAPP (third party independently deployed application) can be stored in the node where the Non-RT RIC is located in advance, wherein an AI (Artificial Intelligence) model for realizing a specified service can be embedded in the xAPP; the Near-RT RIC is deployed at a node with limited calculation capacity, and is responsible for processing services with higher delay requirements.

In order to ensure the real-time performance of the service processing, xAPP may be deployed in advance in the Near-RT RIC, so that the deployed xAPP is utilized to respond to the service request from the terminal.

However, the number of xAPPs that can be pre-deployed into the Near-RT RIC is limited due to the limited memory space in the Near-RT RIC. Then, when the Near-RT RIC receives a service request, if xAPP required for processing the service request is not pre-deployed in the Near-RT RIC, the service request is not hit, and the service request needs to be responded by other distributed Near-RT RIC or Non-RT RIC, thereby causing a large communication overhead.

Therefore, how to reasonably pre-deploy xAPP in Near-RT RIC to improve the hit rate of service request, thereby reducing the communication overhead of service processing becomes a technical problem to be solved urgently.

Disclosure of Invention

The embodiment of the invention aims to provide an application deployment method to improve the hit rate of service requests, so that the communication overhead of service processing is reduced. The specific technical scheme is as follows:

in a first aspect, an embodiment of the present invention provides an application pre-deployment method, which is applied to Near-real-time intelligent controller Near-RT RIC, and the method includes:

determining a service request received by the Near-RT RIC in a specified historical time period; each service request is processed by calling one xAPP in each third-party application xAPP;

determining the calling frequency of each xAPP in the specified historical time period based on the determined service request;

predicting the calling probability corresponding to each xPP at the target moment based on the calling frequency of each xPP in the specified historical time period; the target time is the time when the calls for the xAPPs in the Near-RT RIC are distributed in a steady state;

selecting at least one xAPP to be deployed from each xAPP according to the calling probability corresponding to each xAPP at the target moment; the calling probability corresponding to each xAPP to be deployed is higher than the calling probabilities corresponding to other xAPs except the at least one xAPP to be deployed;

and deploying the selected xAPP to be deployed in the Near-RT RIC.

Optionally, the determining, based on the determined service request, a calling frequency of each xAPP in the specified historical time period includes:

and counting the number of the service requests which are processed by calling the xAPP in the determined service requests aiming at each xAPP in each xAPP, and calculating the ratio of the counted number to the total number of the determined service requests to be used as the calling frequency of the xAPP in the appointed historical time period.

Optionally, the predicting, based on the invocation frequency of each xAPP in the specified historical time period, the invocation probability corresponding to each xAPP at the target time includes:

constructing a state set aiming at each xPP by using the calling frequency of each xPP in the appointed historical time period; the state set is a set formed by sequencing each xAPP according to the calling frequency;

constructing a state transition matrix aiming at each xAPP based on the state set and the calling frequency of each xAPP; each element in the state transition matrix represents the state transition probability of each xAPP;

and calculating the steady-state probability distribution of each xAPP at the target moment based on the state transition matrix to obtain the calling probability corresponding to each xPP at the target moment.

Optionally, the constructing a state transition matrix for each xAPP based on the state set and the calling frequency of each xAPP includes:

aiming at each element in the state set, calculating corresponding state transition probability when the element is respectively transferred to each sequencing position in the state set when a service request occurs on the basis of the state set by using the calling frequency of each xAPP, and taking the corresponding state transition probability as the state transition probability corresponding to the element;

and constructing a state transition matrix based on the state transition probability corresponding to each element.

Optionally, the calculating, based on the state transition matrix, a steady-state probability distribution of each xAPP at the target time to obtain a call probability corresponding to each xAPP at the target time includes:

and calculating the steady-state probability distribution corresponding to the state transition matrix at the target moment by using a C-K equation of a Markov chain to obtain the calling probability corresponding to each xAPP at the target moment.

Optionally, the selecting, according to the invocation probability corresponding to each xAPP at the target time, at least one xAPP to be deployed from each xAPP includes:

determining the number of xAPPs to be deployed locally based on the local storage space as a specified number;

and selecting a specified number of xAPPs to be deployed from each xPP according to the calling probability corresponding to each xPP at the target moment.

In a second aspect, an embodiment of the present invention provides an apparatus for pre-deploying an application, where the apparatus is applied to a Near-real-time intelligent controller Near-RT RIC, and the apparatus includes:

the first determining module is used for determining the service requests received by the Near-RT RIC in the appointed historical time period; each service request is processed by calling one xAPP in each third-party application xAPP;

a second determining module, configured to determine, based on the determined service request, a call frequency of each xAPP in the specified historical time period;

the prediction module is used for predicting the calling probability corresponding to each xAPP at the target moment based on the calling frequency of each xAPP in the specified historical time period; the target time is the time when the calls for the xAPPs in the Near-RT RIC are distributed in a steady state;

the selection module is used for selecting at least one xPP to be deployed from each xPP according to the calling probability corresponding to each xPP at the target moment; the calling probability corresponding to each xAPP to be deployed is higher than the calling probabilities corresponding to other xAPs except the at least one xAPP to be deployed;

and the deployment module is used for deploying the selected xAPP to be deployed in the Near-RT RIC.

Optionally, the second determining module is specifically configured to:

Optionally, the prediction module includes:

the first construction submodule is used for constructing a state set aiming at each xPP by using the calling frequency of each xPP in the appointed historical time period; the state set is a set formed by sequencing each xAPP according to the calling frequency;

the second constructing submodule is used for constructing a state transition matrix aiming at each xAPP based on the state set and the calling frequency of each xAPP; each element in the state transition matrix represents the state transition probability of each xAPP;

and the calculation submodule is used for calculating the steady-state probability distribution of each xAPP at the target moment based on the state transition matrix to obtain the calling probability corresponding to each xPP at the target moment.

Optionally, the second building sub-module is specifically configured to:

Optionally, the calculation submodule is specifically configured to:

Optionally, the selecting module is specifically configured to:

determining the number of xAPPs to be deployed locally as a specified number based on the local storage space;

In a third aspect, an embodiment of the present invention provides an electronic device, including a processor, a communication interface, a memory, and a communication bus, where the processor and the communication interface complete communication between the memory and the processor through the communication bus;

a memory for storing a computer program;

a processor for implementing the steps of the method for pre-deployment of an application described above when executing a program stored in the memory.

In a fourth aspect, an embodiment of the present invention provides a computer-readable storage medium, in which a computer program is stored, and the computer program, when executed by a processor, implements the steps of the application pre-deployment method described in any one of the above.

The embodiment of the invention has the following beneficial effects:

the method for pre-deploying the application, provided by the embodiment of the invention, determines the called calling frequency of each xAPP in the specified historical time period based on the service request received by the Near-RT RIC in the specified historical time period, and predicts the calling probability of each xPP at the target time based on the calling frequency, so that at least one xPP to be deployed is selected from each xPP for deployment by using the calling probability corresponding to each xPP at the target time. The calling probability corresponding to each xAPP at the target moment is the calling probability when the calling of each xPP is in steady-state distribution, so that the proper xAPP can be selected for deployment by using the calling probability corresponding to each xPP. And when the xAPs to be deployed are selected, the calling probability of the selected xAPs to be deployed is higher than that corresponding to other unselected xAPs, and the xAPs with high calling probability are deployed in the Near-RT RIC, so that the Near-RT RIC can hit more service requests, and the response capability of the service requests is improved. Therefore, the method and the device can improve the hit rate of the service request, thereby reducing the communication overhead of service processing.

Of course, not all of the advantages described above need to be achieved at the same time in the practice of any one product or method of the invention.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art that other embodiments can be obtained by referring to these drawings.

FIG. 1 is a flow chart of a pre-deployment method for an application according to an embodiment of the present invention;

fig. 2 is a flowchart of implementing step S103 according to the embodiment of the present invention;

fig. 3 is a flowchart of implementing step S104 according to an embodiment of the present invention;

FIG. 4 is a timing diagram of one particular example of a method of pre-deployment of an application provided by an embodiment of the present invention;

fig. 5 is a schematic structural diagram of a pre-deployment apparatus for an application according to an embodiment of the present invention;

fig. 6 is a block diagram of an electronic device implementing the pre-deployment method of the application provided by the embodiment of the invention.

Detailed Description

The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived from the embodiments given herein by one of ordinary skill in the art, are within the scope of the invention.

With the depth of cloud network convergence, the core network functional units have started to be gradually clouded, and research on replacing special network devices by general devices is steadily advanced. The network intellectualization is an important network evolution route facing the requirements of diversification, endogenous design and open ecology of network services in the future.

The O-RAN organization proposes an intelligent controller RIC for RAN virtualization, wherein the RIC comprises a Non-RT RIC and a Near-RT RIC. The Non-RT RIC and the Near-RT RIC adopt a distributed architecture when being deployed, namely one Non-RT RIC can be connected with one or more Near-RT RICs, and meanwhile the Non-RT RIC can issue the AI model after training to the Near-RT RIC.

In order to ensure the real-time performance of processing the service, xAPP for processing the relevant service request may be pre-deployed in the Near-RT RIC, so that when the Near-RT RIC receives the service request, the xAPP may be used to quickly complete the processing of the service request. However, the number of xAPPs that can be pre-deployed into the Near-RT RIC is limited due to the limited memory space in the Near-RT RIC. Then, when the Near-RT RIC receives a service request, if xAPP required for processing the service request is not pre-deployed in the Near-RT RIC, the service request is not hit, and the service request needs to be responded by other distributed Near-RT RIC or Non-RT RIC, thereby causing a large communication overhead.

Based on the above, in order to improve the hit rate of the service request and reduce the communication overhead of the service processing, embodiments of the present invention provide a pre-deployment method, apparatus, device, and storage medium for an application.

First, a pre-deployment method applied to the embodiments of the present invention will be described.

The application pre-deployment method provided by the embodiment of the invention is applied to Near-RT RIC, the Near-RT RIC can be deployed in a server, a virtual machine or a container, the Near-RT RIC is used for processing services with higher delay requirements, and installation files for deploying xAPP can be obtained from Non-RT RIC and deployed locally.

The pre-deployment method of the application provided by the embodiment of the invention can comprise the following steps:

determining the calling frequency of each xPP in the specified historical time period based on the determined service request;

predicting the calling probability corresponding to each xPP at the target moment based on the calling frequency of each xPP in the designated historical time period; the target time is the time when the calls for the xAPPs in the Near-RT RIC are distributed in a steady state;

and deploying the selected xAPP to be deployed in the Near-RT RIC.

According to the scheme provided by the embodiment of the invention, based on the service request received by the Near-RT RIC in the appointed historical time period, the called calling frequency of each xPP in the appointed historical time period is determined, and the calling probability of each xPP at the target time is predicted based on the calling frequency, so that at least one xPP to be deployed is selected from each xPP for deployment by using the calling probability corresponding to each xPP at the target time. The calling probability corresponding to each xAPP at the target moment is the calling probability when the calling of each xPP is in steady-state distribution, so that the proper xAPP can be selected for deployment by using the calling probability corresponding to each xPP. And when the xAPs to be deployed are selected, the calling probability of the selected xAPs to be deployed is higher than that corresponding to other unselected xAPs, and the xAPs with high calling probability are deployed in the Near-RT RIC, so that the Near-RT RIC can hit more service requests, and the response capability of the service requests is improved. Therefore, through the scheme, the hit rate of the service request can be improved, and the communication overhead of service processing is reduced.

The pre-deployment method of the application provided by the embodiment of the invention is described below with reference to the accompanying drawings.

As shown in FIG. 1, the pre-deployment method of application provided by the embodiment of the present invention, applied to Near-RT RIC, may include steps S101-S105:

s101, determining a service request received by the Near-RT RIC in a specified historical time period; each service request is processed by calling one xAPP in each third-party application xAPP;

it can be understood that, since the Near-RT RIC is responsible for processing the service with higher latency requirement, and the Near-RT RIC processes the service request by using the xAPP deployed by the Near-RT RIC, different types of service requests are processed by using different xapps. If xAPP required to be utilized by a certain service request is not deployed in the Near-RT RIC, the Near-RT RIC does not hit the service request, so that the Near-RT RIC needs to respond to the service request by distributed other Near-RT RIC or Non-RT RIC, thereby causing a large communication overhead.

In order to improve the hit rate of the service request, in this embodiment, the call probability of each xAPP at the target time is predicted by analyzing the service request received by the Near-RT RIC in the historical time period, so that the xAPP with the higher call probability is deployed preferentially, thereby improving the hit rate of the service request and reducing the communication overhead of service processing. Therefore, in this embodiment, the service request received by the Near-RT RIC within a specified historical time period is first determined. The type of the service request may be a voice service, a video service, a web browsing service, and the like, which is not limited in the embodiment of the present invention; the specified historical period of time may be any length of time of the historical period of time, for example, the specified historical period of time may be a period of one week prior to the current time, or a period of one month prior to the current time, etc.; in addition, each third-party application xAPP may be an xAPP in which an installation file is stored in advance in a node where the Non-RT RIC is located, and at least one of the xapps may have been deployed in the Near-RT RIC within a specified historical time period.

In addition, it can be understood that, since an AI model for implementing a specific service is embedded in each xAPP, each service request can be processed through the xAPP in which the AI model for implementing the corresponding service is embedded. For example, if an AI model for implementing a voice service is embedded in xAPP1, a service request of a voice service type may be processed by calling xAPP 1; if the AI model for implementing the video service is embedded in the xAPP2, the service request of the video service type can be processed by calling the xAPP 2.

Moreover, it should be noted that, within the specified historical time period, at least one xAPP that has been deployed within the Near-RT RIC may be randomly deployed for implementing pre-deployment of the application; of course, the pre-deployment process for the application may be a periodic process, in which case, the specified historical time period may include a historical period, in which case, at least one xAPP that has been deployed in the Near-RT RIC may include the xAPP deployed in the historical period.

S102, determining the calling frequency of each xPP in the appointed historical time period based on the determined service request;

in this embodiment, because each service request is processed by the xAPP embedded with the AI model for implementing the corresponding service corresponding to the service request, after the service request received by the Near-RT RIC in the specified historical time period is determined in step S101, the xAPP required to be called by each service request in the specified historical time period can be obtained by analyzing the service request received by the Near-RT RIC in the specified historical time period, and thus the calling frequency of each xAPP in the specified historical time period is calculated. The embodiment of the present invention is not limited to this, and for example, each service request carries an application identifier of an xAPP that needs to be called, so that the xAPP that needs to be called by each service request can be obtained, which is not limited to this.

Optionally, in an implementation manner, determining, based on the determined service request, a call frequency of each xAPP in the specified historical time period may include:

and counting the number of the service requests processed by calling the xAPP in the determined service requests according to each xAPP in each xAPP, and calculating the ratio of the counted number to the total number of the determined service requests as the calling frequency of the xAPP in the specified historical time period.

In the implementation mode, for the local Near-RT RIC, the calling condition of each xPP is analyzed by counting the calling frequency of each xPP in the specified historical time period, so that the local Near-RT RIC is reasonably pre-deployed with the xPP.

Illustratively, if the Near-RT RIC receives 100 voice service requests and 300 video service requests in a specified historical time period, and the third-party application for processing voice services is xAPP1 and the third-party application for processing video services is xAPP2, the call frequency of xAPP1 in the specified historical time period is 0.25 and the call frequency of xAPP2 is 0.75.

S103, predicting the calling probability corresponding to each xPP at the target moment based on the calling frequency of each xPP in the specified historical time period; wherein, the target time is the time when the calls aiming at each xAPP in the Near-RT RIC are distributed in a steady state;

in this embodiment, after the calling frequency of each xAPP in the specified historical time period is determined through step S102, in order to improve the hit rate of the service request, the calling condition of each xAPP in the future time may be estimated based on the calling frequency of each xAPP in the specified historical time period, and in order to ensure the stability of the hit rate of the service request, the steady distribution of the call to each xAPP may be estimated. Specifically, the Near-RT RIC may predict the call probability corresponding to each xAPP at the target time based on the call frequency of each xAPP in the specified historical time period, thereby implementing prediction of the call condition and ensuring stability of the service request hit rate.

It can be understood that there may be a plurality of implementation manners for predicting the calling probability corresponding to each xAPP at the target time, and for example, the Near-RT RIC may construct a state set for each xAPP by using the calling frequency of each xAPP; then, a state transition matrix of each xAPP at the initial moment is constructed by using the state set, and the Markov chain is used for solving the steady-state distribution probability of each xPP, namely the calling probability of each xPP at the moment when the calling of each xPP is in steady-state distribution, so that the calling probability corresponding to each xPP at the target moment is obtained. It should be noted that, for clarity of layout, a process of predicting a calling probability corresponding to each xAPP at a target time based on a calling frequency of each xAPP in the specified historical time period will be described below, and details are not repeated here.

S104, selecting at least one xPP to be deployed from each xPP according to the calling probability corresponding to each xPP at the target moment; the calling probability corresponding to each xAPP to be deployed is higher than the calling probabilities corresponding to other xAPs except the at least one xAPP to be deployed;

in this embodiment, after the calling probability corresponding to each xAPP at the target time is obtained in step S103, at least one xAPP may be selected from each xAPP according to the calling probability corresponding to each xAPP, and is used as the xAPP to be deployed. It can be understood that, in order to improve the hit rate of the service request, when the at least one xAPP to be deployed is selected, the call probability corresponding to each xAPP to be deployed is higher than the call probabilities corresponding to other xapps except the at least one xAPP to be deployed. For example, the at least one xAPP to be deployed may be selected by selecting the first several xapps as the xapps to be deployed according to the high-to-low order of the calling probability corresponding to each xAPP.

And S105, deploying the selected xAPP to be deployed in the Near-RT RIC.

In this embodiment, an installation file embedded with each xAPP (third party independently deployed application) may be stored in advance in a node where the Non-RT RIC is located, an AI model for implementing a specified service is embedded in each xAPP, after at least one xAPP to be deployed is selected in step S104, the Near-RT RIC may send a request for the installation file of the xAPP to be deployed to the Non-RT RIC, so that after the Non-RT RIC receives the request, the installation file of the xAPP to be deployed is sent to the node where the Near-RT RIC is located, and after the node where the Near-RT RIC is located receives the installation file of the xAPP to be deployed, the node is installed to complete pre-deployment of the xapps. It can be understood that after the xPP to be deployed is deployed, because the pre-deployed xPP is the xPP with a higher invocation probability, the pre-deployed xPP can respond to more service requests, thereby improving the hit rate of the service requests and reducing the communication overhead of service processing.

Optionally, in another embodiment of the present invention, as shown in fig. 2, in the step S103, predicting the calling probability corresponding to each xAPP at the target time based on the calling frequency of each xAPP in the specified history time period may include steps S1031 to S1033:

s1031, constructing a state set aiming at each xPP by using the calling frequency of each xPP in the specified historical time period; the state set is a set formed by sequencing each xAPP according to the calling frequency;

in the implementation manner, a state set for each xAPP is constructed according to the order of the calling frequency of each xAPP from high to low in the specified historical time period. Illustratively, if the invocation frequency of xAPP1 is 0.2, the invocation frequency of xAPP2 is 0.5, and the invocation frequency of xAPP3 is 0.3 within the specified historical time period, the constructed state set is { xAPP2, xAPP3, xAPP1 }.

S1032, constructing a state transition matrix aiming at each xAPP based on the state set and the calling frequency of each xAPP; wherein, each element in the state transition matrix represents the state transition probability of each xAPP;

it can be understood that, since the state set is constructed by using the calling frequency of each xAPP in the specified historical time period, based on the state set and the calling frequency of each xAPP, a state transition matrix at an initial time can be constructed, and each element in the state transition matrix represents the state transition probability of each xAPP. The initial time may be a time when a service request occurs for the first time on the basis of the state set. Illustratively, if the state set is { xAPP2, xAPP3, xAPP1}, the calling frequency of xAPP1 is 0.2, the calling frequency of xAPP2 is 0.5, and the calling frequency of xAPP3 is 0.3, then the position ordering of xAPP1 in the set is first if xAPP1 is called at the initial time, and since the calling frequency of xAPP1 is 0.2, the state transition probability of the xAPP1 from position 3 to position 1 is 0.2, that is, there is a probability of 0.2 at the initial time such that xAPP1 transitions from position 3 to position 1; when xAPP2 or xAPP3 is called at this initial time, the position ranking of xAPP1 in the set is still the last, and the probability that the position of xAPP1 remains unchanged is the sum of the calling frequencies of all xapps ranked before xAPP1, that is, the sum of the calling probability of xAPP2 and the calling frequency of xAPP3 in this state set is 0.8.

For example, in a specific implementation, constructing a state transition matrix for each xAPP based on the state set and the calling frequency of each xAPP may include steps a1-a 2:

a1, calculating the corresponding state transition probability of each element in the state set when the element is respectively transferred to each sequencing position in the state set when a service request occurs on the basis of the state set by using the calling frequency of each xAPP as the state transition probability corresponding to the element;

in this implementation, each element in the state set is each xAPP ordered from high to low according to the call frequency. If an xAPP is called at a certain time, the ordering position of the xAPP in the state set is raised to the first position; if the xAPP before the position of the xAPP in the state set is called, the ordered position of the xAPP in the state set is kept unchanged; if the xPP located after the position of the xPP in the state set is called, the xPP is shifted one bit after the sorting position in the state set. Since the calling frequency of each element in the state set is calculated by using the calling times of the xapps in the specified historical time period, the state transition probability of each element at the initial time can be calculated by using the calling frequency of each xAPP, and the probability that each element is transitioned to the first sorting position in the state set at the initial time is the calling frequency of the element. For example, if the calling frequency of xAPP1 is 0.2, the calling frequency of xAPP2 is 0.5, and the calling frequency of xAPP3 is 0.3, the probability of transition to the first ranking position in the state set at the initial time xAPP1 is 0.2, the probability of transition to the first ranking position in the state set at xAPP2 is 0.5, and the probability of transition to the first ranking position in the state set at xAPP3 is 0.3.

In addition, it can be understood that, for any element except for the first sorting position in the state set, if any xAPP sorted before the element is called at the initial time, the sorting position of the element in the state set sorted by the calling frequency of each xAPP at the initial time remains unchanged, so that the probability that the sorting position of the element remains unchanged is the sum of the calling frequencies of all xapps sorted before the element. For any element except for the last-but-one ordering position in the state set, if any xPP ordered after the element is called at the initial time, the ordering position of the element in the state set ordered by the calling frequency of each xPP at the initial time is shifted backward by one, so that the probability that the ordering position of the element at the initial time is shifted backward by one is the sum of the calling frequencies of all xAPs ordered after the element.

A2, constructing a state transition matrix based on the state transition probability corresponding to each element.

After the state transition probability corresponding to each element in the state set is calculated, a row dimension representing each xAPP can be constructed, and a column dimension representing a probability state transition matrix for transferring the xAPP to each sequencing position in the state set. Illustratively, if the calling frequency of xAPP1 is 0.6, the calling frequency of xAPP2 is 0.3, and the calling frequency of xAPP3 is 0.1, the state set is { xAPP1, xAPP2, xAPP3 }. The state transition matrix

Wherein, P ₁₁ Characterizing the probability that the position of the xAPP1 remains unchanged in this state set, P ₁₂ Characterizing xAPP1 in this state setProbability of position transition from first bit to second bit, P ₂₁ Characterize the probability that the position of xPP 2 is first in the state set, P ₂₂ Characterizing the probability that the position of the xAPP2 remains unchanged in this state set, P ₂₃ Characterize the probability that the xAPP2 is shifted one bit down in the state set, P ₃₁ Characterize the probability that the position of xAPP3 is first in the set of states, P ₃₃ The probability of a bit shift in this state set by the xAPP3 is characterized.

And S1033, calculating the steady-state probability distribution of each xPP at the target moment based on the state transition matrix to obtain the calling probability corresponding to each xPP at the target moment.

It can be understood that, after the initial state transition matrix is constructed in step S1032, the ordering position of each xAPP in the state set changes continuously with the passage of time, the state transition probability of each xAPP changes continuously, and finally the state transition probability of each xAPP tends to be stable, at this time, the steady state probability distribution of each xAPP, that is, the call probability corresponding to each xAPP at the target time, can be obtained.

Illustratively, in a specific implementation manner, calculating a steady-state probability distribution of each xAPP at the target time based on the state transition matrix to obtain a call probability corresponding to each xAPP at the target time includes:

and calculating the steady-state probability distribution corresponding to the state transition matrix at the target moment by using the C-K equation of the Markov chain to obtain the calling probability corresponding to each xAPP at the target moment.

In this implementation, by using the traversability of the markov chain, after n steps of state transition, the state transition matrix has a limit distribution, that is, each element in the state transition matrix has a steady-state probability distribution. And solving the C-K equation to obtain the steady state probability distribution corresponding to the state transition matrix at the target moment. It should be noted that the formula of the C-K equation, also called the schepmann-kolmogoroff equation, is pi P ═ pi, where pi is the steady state probability distribution of each xAPP, P is the n-step state transition matrix of the xAPP, and n is a positive integer.

Therefore, according to the scheme, the Markov chain is utilized to solve the steady-state distribution probability of each xAPP, namely the calling probability at the moment when the calling of each xAPP is in steady-state distribution, and the calling probability corresponding to each xPP at the target moment is obtained. Therefore, at least one xAPP to be deployed can be selected from each xAPP for deployment by utilizing the calling probability corresponding to each xAPP at the target moment, so that the hit rate of the service request is improved, and the communication overhead of service processing is reduced.

Optionally, in another embodiment of the present invention, as shown in fig. 3, in the step S104, selecting at least one xAPP to be deployed from each xAPP according to the calling probability corresponding to each xAPP at the target time may include steps S1041 to S1042:

s1041, determining the number of local xAPPs to be deployed as a specified number based on the local storage space;

in this embodiment, when the xAPP to be deployed is selected, the number of the xAPP to be deployed may be determined according to the local storage space, that is, the local remaining memory space.

And S1042, selecting a specified number of xAPs to be deployed from each xPP according to the calling probability corresponding to each xPP at the target moment.

For example, if two xapps can be deployed in the local storage space, and it is determined that the specified number is 2, the first two xapps with the highest call probability may be selected from the xapps, and used as the xAPP to be deployed.

Therefore, through the scheme, the xAPP in the Near-RT RIC can be reasonably deployed according to the size of the actual storage space in the Near-RT RIC, so that the storage space of the Near-RT RIC is maximally utilized to deploy the xAPP, and the hit rate of the Near-RT RIC for the service request is further improved.

For a better understanding of the contents of the embodiments of the present invention, a specific example of the embodiments of the present invention is described below with reference to fig. 4.

The existing O-RAN organization discusses RIC, considering from the interface and application scene, and the research on Non-RT RIC and Near-RTRIC related to service scheduling is little. The embodiment of the invention uses the discrete Markov chain to calculate the steady-state probability of the business scene for the xPP requirement from the aspect of saving resources, and arranges the corresponding xPP in the Near-RT RIC in advance according to the magnitude sequence of the steady-state probability so as to improve the hit rate of the business request.

Fig. 4 is a timing diagram illustrating a specific example of a pre-deployment method for implementing the application provided by the embodiment of the present invention. The example illustrates the pre-deployment process of the Non-RT RIC to the Near-RT RIC node by taking Near-RT RIC deployed at a certain node as an example. As shown in fig. 4, the method comprises the following steps:

(1) service request collection: the Near-RT RIC collects service requests for a specified historical time period. Each service request is processed by calling one of the xapps. And calculating the frequency of each xPP used in the time period, namely the calling frequency of the xPP according to the number of service requests. The specific calculation process is as follows: setting the total number of service requests received by the Near-RT RIC in the appointed historical time period as m, wherein a certain xPP is called for n times, and then the calling frequency P of the xPP _{Initiation of} ＝n/m。

(2) Constructing an xAPP initial probability distribution: constructing an initial probability distribution of each xAPP, namely a state transition matrix of each xPP at an initial moment, and comprising the steps B1-B2:

b1, based on the calculated calling frequency of each xAPP, ranks each xAPP from high to low according to the calling frequency, and forms a state set C ═ 1, 2, 3, …, t, t +1, …, x }.

B2, constructing a state transition matrix P of each xAPP at the initial moment;

in the obtained ordering of the state set, if the xAPP at the t-th position at the initial time is called, the ordering of the xAPP at the t-th position at the time in the state set C is raised to the first position. Thus, the status of xAPP at position t in set C is: if the xAPP before the xAPP arranged at the t-th position is called at a certain moment, the state of the xAPP at the t-th position is still at the t-th position; if the xAPP following the xAPP ranked at the t-th position is called, the following xAPP ranks at the first position, and the state of the xAPP at the t-th position shifts to the t +1 th position. And, in aiming at the setCombining with the t-th ordered xPP in C, if the xPP is called at a certain time, the calling probability is P _t→1 ＝P _{Initiation of} Wherein, P _t→1 Characterizing the state transition probability, P, of the state of the xPP in the set C from the t-th bit to the first bit _Initial The calling frequency of the xPP in the specified historical time period. If the state of the xPP in the set C is transferred from the t-th bit to the t + 1-th bit at a certain moment, the probability of state down shifting is the sum of the calling frequencies called by all xAPPs positioned after the t-th bit in the set C, and then P _t→t+1 ＝∑P _s ，s∈[t+1，x]Wherein P is _t→t+1 The state transition probability, sigma P, of the state of the xPP in the set C from the t th bit to the t +1 th bit is represented _s The sum of the calling frequencies of all xAPPs after the t bit is called for the initial time. If the state of the xPP in the set C is from the tth bit to the tth bit at a certain moment, the probability that the state remains unchanged is the sum of the calling frequencies called by all xAPPs before the tth bit in the set C, and P _t→t ＝∑P _s ,s∈[1,t-1]Wherein P is _t→t The state transition probability, sigma P, of the state of the xPP in the set C from the t bit to the t bit is represented _s The sum of the calling frequencies of all xAPPs before the t bit is called for the initial time.

In addition, for the xAPP located first in the state set C, there are only two transition states: the ordering is unchanged or shifted down by one bit, and for the xAPP located at the last bit in the state set C, there are also only two transition states: the ordering is either lifted to the first bit or left unchanged. From the above analysis, it can be found that the state transition probability of the tth xPP in the set C always satisfies P _t→1 +P _t→t +P _t→t+1 ＝1。

After the state transition probability of each xAPP is calculated, a state transition matrix P of each xPP at the initial moment can be constructed.

The above process is described below in connection with a 5 x 5 state transition matrix.

Each row in the matrix represents each xAPP, and each element in each row represents a corresponding state transition probability when being transferred to each sequencing position in the state set. For example, P ₂₃ And (3) representing the probability that the state of the xPP currently positioned at the 2 nd bit in the state set is transferred to the 3 rd bit in the state set. As can be seen from the matrix, the first and last rows have only two non-0 elements, and the other rows have three non-0 elements. Taking the third row as an example, P31 ═ P _{Initial 3} ；P33＝P _{Initial 1} +P _{Initial 2} Wherein P is _{Initial 1} For the calling frequency, P, of the xPP 1 in the specified historical time period _{Initial 2} For the calling frequency, P, of the xPP 2 in the specified historical time period _{Initial 3} Is the invocation frequency of the xAPP3 at the specified historical time period.

(3) Calculating the steady-state probability distribution of xAPP: and calculating the steady-state probability distribution of each xAPP according to a C-K equation, namely the calling probability corresponding to each xPP at the target moment. That is, the steady-state probability distribution pi of each xAPP is calculated according to the formula pi P ═ pi. The steady-state probability distribution pi found by the formula pi P ═ pi is a one-dimensional matrix, and sigma pi _i 1, wherein i is the identity of each xAPP.

(4) Determining xAPP to be deployed according to the steady-state probability distribution and the size of a storage space: and selecting xAPPs with the specified number of steady-state probability distributions ordered from high to low as the xAPPs to be deployed according to the steady-state probability distribution calculated in the pi and the residual storage space in the node where the current Near-RT RIC is located.

(5) Sending a request for installation files of xAPP to be deployed: and the Near-RT RIC sends a request aiming at the installation file of the xAPP to be deployed to the Non-RT RIC according to the selected xAPP to be deployed.

(6) In response to a request sent by the Near-RT RIC: the Non-RT RIC responds to requests sent by the Near-RT RIC.

(7) Issuing an image and a configuration file of xAPP to be deployed: and the Non-RT RIC transmits the mirror image and the configuration file of the xPP to be deployed to the node where the Near-RT RIC is located according to the request, and the node where the Near-RT RIC is located carries out xPP installation after receiving the mirror image and the configuration file of the xPP to be deployed so as to complete the pre-deployment of the xPP.

Therefore, according to the scheme, the steady-state probability distribution of the xAPP can be predicted according to the historical data of the service request, so that the xAPP can be deployed in advance according to the size of the actual storage space of the node where the Near-RT RIC is located and the size of the steady-state probability distribution, an xPP pre-deployment mechanism of the Near-RT RIC node is made up, the hit rate of the service request is improved, and the communication overhead of service processing is reduced. And the algorithm complexity is low, the landing deployment is convenient, and the service rate of the business can be improved.

Correspondingly, the embodiment of the foregoing method, an embodiment of the present invention further provides an application pre-deployment apparatus, which is applied to a Near-real-time intelligent controller Near-RT RIC, as shown in fig. 5, where the apparatus includes:

a first determining module 510, configured to determine a service request received by the Near-RT RIC within a specified historical time period; each service request is processed by calling one xAPP in each third-party application xAPP;

a second determining module 520, configured to determine, based on the determined service request, a calling frequency of each xAPP in the specified historical time period;

a prediction module 530, configured to predict, based on the call frequency of each xAPP in the specified historical time period, a call probability corresponding to each xAPP at a target time; the target time is the time when the calls for the xAPPs in the Near-RT RIC are distributed in a steady state;

a selecting module 540, configured to select at least one xAPP to be deployed from each xAPP according to the call probability corresponding to each xAPP at the target time; the calling probability corresponding to each xAPP to be deployed is higher than the calling probabilities corresponding to other xAPs except the at least one xAPP to be deployed;

a deployment module 550, configured to deploy the selected xAPP to be deployed in the Near-RT RIC.

Optionally, the second determining module is specifically configured to:

and for each xAPP in each xAPP, counting the number of the service requests processed by calling the xAPP in the determined service requests, and calculating the ratio of the counted number to the total number of the determined service requests as the calling frequency of the xAPP in the specified historical time period.

Optionally, the prediction module includes:

Optionally, the second building submodule is specifically configured to:

Optionally, the computation submodule is specifically configured to:

Optionally, the selecting module is specifically configured to:

An embodiment of the present invention further provides an electronic device, as shown in fig. 6, including a processor 601, a communication interface 602, a memory 603, and a communication bus 604, where the processor 601, the communication interface 602, and the memory 603 complete mutual communication through the communication bus 604,

a memory 603 for storing a computer program;

the processor 601 is configured to implement the steps of the application pre-deployment method according to any of the above embodiments when executing the program stored in the memory 603.

The communication bus mentioned in the electronic device may be a Peripheral Component Interconnect (PCI) bus, an Extended Industry Standard Architecture (EISA) bus, or the like. The communication bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one thick line is shown, but this is not intended to represent only one bus or type of bus.

The communication interface is used for communication between the electronic equipment and other equipment.

The Memory may include a Random Access Memory (RAM) or a Non-Volatile Memory (NVM), such as at least one disk Memory. Optionally, the memory may also be at least one memory device located remotely from the processor.

The Processor may be a general-purpose Processor, including a Central Processing Unit (CPU), a Network Processor (NP), and the like; but also Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete Gate or transistor logic devices, discrete hardware components.

In yet another embodiment of the present invention, a computer-readable storage medium is further provided, in which a computer program is stored, and the computer program, when executed by a processor, implements the steps of the pre-deployment method of an application described in any of the above embodiments.

In yet another embodiment, the present invention further provides a computer program product containing instructions, which when run on a computer, causes the computer to execute the method for pre-deployment of an application as described in any of the above embodiments.

In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When implemented in software, may be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When loaded and executed on a computer, cause the processes or functions described in accordance with the embodiments of the invention to occur, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another computer readable storage medium, for example, the computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that incorporates one or more of the available media. The usable medium may be a magnetic medium (e.g., floppy Disk, hard Disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., Solid State Disk (SSD)), among others.

It is noted that, herein, relational terms such as first and second, and the like may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in the process, method, article, or apparatus that comprises the element.

All the embodiments in the present specification are described in a related manner, and the same and similar parts among the embodiments may be referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, as for the apparatus embodiment, since it is substantially similar to the method embodiment, the description is relatively simple, and for the relevant points, reference may be made to the partial description of the method embodiment.

The above description is only for the preferred embodiment of the present invention, and is not intended to limit the scope of the present invention. Any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention shall fall within the protection scope of the present invention.

Claims

1. A pre-deployment method of an application, applied to a Near-real-time intelligent controller Near-RT RIC, the method comprising:

and deploying the selected xAPP to be deployed in the Near-RT RIC.

2. The method of claim 1, wherein the determining the calling frequency of each xApP within the specified historical time period based on the determined service request comprises:

3. The method of claim 1, wherein the predicting the calling probability corresponding to each xAPP at a target time based on the calling frequency of each xAPP in the specified historical time period comprises:

constructing a state set aiming at each xAPP by using the calling frequency of each xAPP in the appointed historical time period; the state set is a set formed by sequencing each xAPP according to the calling frequency;

and calculating the steady-state probability distribution of each xPP at the target moment based on the state transition matrix to obtain the calling probability corresponding to each xPP at the target moment.

4. The method of claim 3, wherein constructing a state transition matrix for each xAPP based on the state set and the calling frequency of each xAPP comprises:

5. The method of claim 3, wherein the calculating the steady-state probability distribution of each xAPP at the target time based on the state transition matrix to obtain the call probability corresponding to each xPP at the target time comprises:

6. The method according to any one of claims 1-5, wherein the selecting at least one xAPP to be deployed from each xAPP according to the calling probability corresponding to each xAPP at the target time comprises:

7. An apparatus for pre-deployment of an application, applied to a Near-real-time intelligent controller Near-RT RIC, the apparatus comprising:

the first determining module is used for determining the service request received by the Near-RT RIC in the appointed historical time period; each service request is processed by calling one xAPP in each third-party application xAPP;

8. The apparatus of claim 7, wherein the second determining module is specifically configured to:

9. The apparatus of claim 7, wherein the prediction module comprises:

the second construction submodule is used for constructing a state transition matrix aiming at each xAPP based on the state set and the calling frequency of each xAPP; each element in the state transition matrix represents the state transition probability of each xAPP;

10. The apparatus according to claim 9, wherein the second building module is specifically configured to:

11. The apparatus according to claim 9, wherein the computation submodule is specifically configured to:

12. The apparatus according to any one of claims 7 to 11, wherein the selection module is specifically configured to:

13. An electronic device is characterized by comprising a processor, a communication interface, a memory and a communication bus, wherein the processor and the communication interface are used for realizing mutual communication by the memory through the communication bus;

a memory for storing a computer program;

a processor for implementing the method steps of any of claims 1-6 when executing a program stored in the memory.

14. A computer-readable storage medium, characterized in that a computer program is stored in the computer-readable storage medium, which computer program, when being executed by a processor, carries out the method steps of any one of claims 1 to 6.