CN106487570A - A kind of method and apparatus of assessment network performance index variation tendency - Google Patents

Abstract

The invention discloses a method and equipment for evaluating the change trend of network performance indicators. The method includes: obtaining the index values of n network performance indicators within a set geographic granularity at m consecutive time points, and obtaining n network performance indicators The n×m-dimensional original performance index matrix, and then by calculating the covariance matrix of the original performance index matrix, calculate the Mahalanobis distance between each other index in the n network performance indexes and the predetermined index, and then obtain the change with the predetermined index The index with the closest trend and the index with the largest difference from the change trend of the predetermined index, wherein, the n is a positive integer greater than 1, and the m is a positive integer not less than 1, thereby realizing the change trend of the network performance index The automatic judgment of network optimization improves the analysis efficiency of network optimization and can improve the comprehensiveness and scalability of network optimization analysis.

Description

A method and device for evaluating the change trend of network performance indicators

technical field

The invention relates to the technical field of network communication, in particular to a method and equipment for evaluating the change trend of network performance indicators.

Background technique

After the communication network is completed, continuous optimization is required to ensure the normal operation of the network and make the network quality meet the needs of users. Network optimization is to find out the reasons that affect the quality of the network through data collection, data analysis, parameter analysis, hardware inspection, etc. on the existing network, and through parameter adjustment, network structure adjustment, equipment configuration adjustment and other technical means , to ensure the high-quality operation of the network, so that the existing network resources can obtain the best benefits.

In general, network optimization can include the following steps:

Step 1: Data collection, that is, collect network performance indicators through testing, network management or monitoring systems;

Step 2: Data analysis, that is, analyze the collected network performance indicators, such as the absolute value of the analysis indicators and their changes;

Step 3: Locating the problem, that is, judging the problem of the live network based on the analysis results of the indicators and the optimization experience of the network optimization personnel;

Step 4: Plan implementation, that is, formulate an optimization plan based on the judged possible problems of the live network, and implement the live network;

Step 5: Effect evaluation, that is, evaluate whether the optimization plan is effective by comparing the changes in network performance indicators before and after the implementation of the plan. If the problem is not solved, it needs to be optimized again.

It can be seen from the above that in the process of network optimization, network performance indicators are the basis for network analysis by network optimization personnel. Network optimization personnel can locate network problems by analyzing changes in network performance indicators. The re-analysis of the network performance index evaluates the effect of the optimization scheme.

Specifically, when network optimization personnel analyze network performance indicators, they usually output the predefined network performance indicators into an excel report, and draw a change trend chart for the set of indicators they care about, and then perform Problem analysis; and, after obtaining a preliminary solution through analysis and implementing it, network optimization personnel can again draw a change trend diagram for the set of indicators they care about, and further evaluate the optimization effect.

That is to say, in the existing network optimization process, network optimization personnel need to manually observe the change trends of various indicators by drawing indicator change curves based on a large number of network performance indicator reports. However, due to the many indicators defined in the current network, it is impossible for network optimization personnel to draw the change curves of each indicator in turn for problem analysis. They can only select some key indicators they believe to observe according to the existing network optimization experience. Analyze and form a preliminary optimization plan. Similarly, after the implementation of the optimization plan, it is impossible for the network optimization personnel to observe the impact of the optimization plan on all indicators. They can only observe the change trend of the indicators that they think may be affected according to past experience, and obtain the evaluation results, which will lead to the existence of The following questions:

Problem 1. It consumes a lot of manpower and time, reducing the efficiency of network optimization: Since network optimization is a continuous process that begins after the network is built and operated, network optimization personnel need to continuously select indicators, observe indicators, and optimize indicators. For evaluation and other operations, if you only rely on human eyes to observe indicators, it will be time-consuming and labor-intensive, which will affect efficiency.

Problem 2: Lack of comprehensiveness and accuracy of analysis: Due to the huge number of network performance indicators, it is impossible for network optimization personnel to analyze the changing trends of all indicators, and it is also impossible to fully understand the impact on the quality of the existing network after the implementation of the optimization scheme. It is possible that the set of indicators that network optimization personnel care about changes according to the expected trend, but other indicators have deteriorated but have not been discovered. These situations will affect the accuracy and comprehensiveness of the analysis.

Problem 3: Lack of scalability for analysis: In the process of network optimization, network optimization personnel usually select the index sets they care about for observation based on a large amount of optimization experience accumulated in the past. More and more detailed network performance indicators will be continuously defined. At this time, network optimization personnel have no experience to follow for the relationship between the newly defined indicators and the original indicators, which will greatly affect the scalability of the analysis and reduce the efficiency of the analysis.

That is to say, since the change trend of network performance indicators is usually judged by humans based on experience, there will be problems such as reducing the analysis efficiency of network optimization and making network optimization analysis lack comprehensiveness and scalability.

Contents of the invention

The embodiment of the present invention provides a method and equipment for evaluating the change trend of network performance indicators, which are used to realize the automatic judgment of the change trend of network performance indicators, so as to solve the problems of low analysis efficiency and lack of comprehensiveness in the existing network optimization analysis methods and scalability issues.

An embodiment of the present invention provides a method for evaluating the trend of network performance indicators, including:

Obtain the index values of n network performance indicators within the set geographic granularity at consecutive m time points, and obtain an n×m-dimensional original performance indicator matrix of n network performance indicators;

Calculating the covariance matrix of the original performance index matrix;

For any predetermined index among the n network performance indexes, according to the original performance index matrix and the covariance matrix, calculate the Markov value of each other index among the n network performance indexes and the predetermined index distance;

According to the calculated Mahalanobis distance between each other index and the predetermined index, determine the index that is the closest to the change trend of the predetermined index among the n network performance indexes and has the largest difference from the change trend of the predetermined index. index of;

Wherein, said n is a positive integer greater than 1, and said m is a positive integer not less than 1.

Optionally, the covariance matrix Σ of the original performance index matrix X is calculated by the following formula:

Among them, μi is the expected value of the i-th indicator Xi, and the (i, j)th element Σij in the covariance matrix Σ is the covariance between Xi and the j-th indicator Xj, and i and j are any positive integer.

Optionally, the Mahalanobis distance Dij between any index Xj among the n network performance indexes and the predetermined index Xi is calculated by the following formula:

Wherein, Σ is the covariance matrix of the original performance index matrix X, and i and j are any positive integers not greater than n.

Optionally, according to the calculated Mahalanobis distance between each other index and the predetermined index, determine the index closest to the change trend of the predetermined index among the n network performance indexes and the distance between the predetermined index and the predetermined index. Indicators with the largest gaps in trend, including:

Obtain a 1×(n-1)-dimensional Mahalanobis distance matrix according to the calculated Mahalanobis distance between each other index and the predetermined index;

According to the size of each matrix element in the Mahalanobis distance matrix, sort each matrix element in the Mahalanobis distance matrix in order from large to small;

The highest-ranked indicator is used as the indicator closest to the change trend of the predetermined indicator among the n network performance indicators, and the last-ranked indicator is used as the indicator with the largest difference from the change trend of the predetermined indicator.

Optionally, before calculating the covariance matrix of the original performance index matrix, the method further includes:

A normalization operation is performed on each index value of the obtained n network performance indexes.

Optionally, according to the calculated Mahalanobis distance between each other index and the predetermined index, determine the index closest to the change trend of the predetermined index among the n network performance indexes and the index closest to the predetermined index After the index with the largest variation trend gap, the method further includes:

According to the determined index closest to the change trend of the predetermined index and the index most different from the change trend of the predetermined index, determine whether the influence of the adjustment of the network parameter on the change of the index is reasonable, so as to judge whether the adjustment of the network parameter is reasonable.

Correspondingly, the embodiment of the present invention also provides a device for evaluating the change trend of network performance indicators, including:

The index acquisition module is used to obtain the index values of n network performance indicators within the set geographic granularity at consecutive m time points, and obtain an n×m-dimensional original performance index matrix of n network performance indicators; wherein, the n is a positive integer greater than 1, and the m is a positive integer not less than 1;

A covariance calculation module, configured to calculate the covariance matrix of the original performance index matrix obtained by the index acquisition module;

An indicator distance calculation module, configured to, for any predetermined indicator among the n network performance indicators, according to the original performance indicator matrix obtained by the indicator acquisition module and the calculated covariance calculation module A covariance matrix, calculating the Mahalanobis distance between each of the n network performance indicators and the predetermined indicator;

An indicator analysis module, configured to determine the change trend of the n network performance indicators closest to the predetermined indicator according to the Mahalanobis distance between each other indicator calculated by the indicator distance calculation module and the predetermined indicator indicators and the indicator with the greatest gap from the changing trend of the predetermined indicator.

Optionally, the covariance calculation module is specifically configured to calculate the covariance matrix Σ of the original performance index matrix X by the following formula:

Among them, μi is the expected value of the i-th indicator Xi, and the (i, j)th element Σij in the covariance matrix Σ is the covariance between Xi and the j-th indicator Xj, and i and j are any positive integer.

Optionally, the indicator distance calculation module is specifically configured to calculate the Mahalanobis distance Dij between any indicator Xj among the n network performance indicators and the predetermined indicator Xi through the following formula:

Wherein, Σ is the covariance matrix of the original performance indicator matrix X, and i and j are any positive integers not greater than N, respectively.

Optionally, the index analysis module is specifically configured to obtain a 1×(n-1)-dimensional Mahalanobis distance matrix according to the calculated Mahalanobis distance between each other index and the predetermined index; and according to the Mahalanobis distance According to the size of each matrix element in the Mahanobis distance matrix, the matrix elements in the Mahalanobis distance matrix are sorted according to the order from large to small; For the index closest to the change trend of the predetermined index, the index ranked last is taken as the index with the largest difference from the change trend of the predetermined index.

Optionally, the device also includes a normalization module:

The normalization module is configured to normalize the index values of the n network performance indexes acquired by the index acquisition module before the covariance calculation module calculates the covariance matrix of the original performance index matrix operation.

Optionally, the device also includes an indicator evaluation module:

The indicator evaluation module is used to determine, in the indicator analysis module, which of the n network performance indicators has the best variation trend with the predetermined indicator according to the calculated Mahalanobis distance between each other indicator and the predetermined indicator. After the index that is close to the change trend of the predetermined index and the index that has the greatest distance from the change trend of the predetermined index, determine the network parameters according to the determined index that is the closest to the change trend of the predetermined index and the index that is the largest difference from the change trend of the predetermined index Whether the influence of the adjustment on the index change is reasonable, so as to judge whether the adjustment of the network parameters is reasonable.

The beneficial effects of the present invention are as follows:

The embodiment of the present invention provides a method and equipment for evaluating the change trend of network performance indicators. The method includes: obtaining the index values of n network performance indicators within the set geographic granularity at m consecutive time points, and obtaining n network performance indicators. The n×m-dimensional original performance index matrix of the performance index, and then by calculating the covariance matrix of the original performance index matrix, calculating the Mahalanobis distance between each other index in the n network performance indexes and the predetermined index, and then obtaining the predetermined index The index with the closest change trend and the index with the largest gap with the change trend of the predetermined index, wherein, the n is a positive integer greater than 1, and the m is a positive integer not less than 1, so that the network performance index The automatic judgment of the change trend improves the analysis efficiency of network optimization and can improve the comprehensiveness and scalability of network optimization analysis.

Description of drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings that need to be used in the description of the embodiments will be briefly introduced below. Obviously, the drawings in the following description are only some embodiments of the present invention. For Those of ordinary skill in the art can also obtain other drawings based on these drawings without making creative efforts.

FIG. 1 is a schematic flow diagram of a method for evaluating a change trend of a network performance index described in Embodiment 1 of the present invention;

FIG. 2 is a schematic structural diagram of a device for evaluating a change trend of a network performance index described in Embodiment 2 of the present invention.

detailed description

In order to make the purpose, technical solutions and advantages of the present invention clearer, the present invention will be further described in detail below in conjunction with the accompanying drawings. Obviously, the described embodiments are only some of the embodiments of the present invention, rather than all of them. Based on the embodiments of the present invention, all other embodiments obtained by persons of ordinary skill in the art without making creative efforts belong to the protection scope of the present invention.

Embodiment one:

Embodiment 1 of the present invention provides a method for evaluating the change trend of network performance indicators, as shown in FIG. The method for trending performance indicators may include the following steps:

Step 101: Obtain the index values of n network performance indicators within the set geographic granularity at consecutive m time points, and obtain an n×m-dimensional original performance indicator matrix of n network performance indicators; wherein, the n is greater than 1 positive integer, and the m is a positive integer not less than 1.

Wherein, the set geographical granularity can be flexibly set according to the actual situation, such as a set cell, a set area or the whole network, etc.; in addition, the values of n and m can also be set according to the actual situation Flexible settings, for example, in order to make the network analysis results more accurate, the value of n can usually be set to the total number of all network performance indicators within the set geographic granularity, and the value of m can be set to a relative Larger numerical values, etc.; moreover, the time points among the m time points may take values such as hours or days according to actual conditions, which will not be described in this embodiment of the present invention.

Optionally, according to the obtained index values of the n network performance indicators within the set geographic granularity at consecutive m time points, the obtained n×m-dimensional original performance indicator matrix X of the n network performance indicators Can be shown as follows:

Wherein, all indicator data at each time point can be used as an analysis sample, that is, m samples of n indicators can be obtained. In addition, the (i, j)th element Xij in the original performance index matrix X may refer to the i (i=1,...,n) network performance index at the j (j=1,...,m ) index value at the time point, that is, the i-th index value in the j-th sample.

For example, if you want to analyze the network performance indicators of a certain cell on a daily basis, that is, if the obtained index values of n network performance indicators in the set geographic granularity at m consecutive time points are consecutive 30-day index sampling value (such as a sampling point every 15 minutes), then the obtained n × m-dimensional original performance index matrix X of the n network performance indicators is a large number of index sampling values of each network performance index By processing on a daily basis, the obtained original matrix of index values of all network performance indicators of the cell for 30 consecutive days, with a size of N rows and 30 columns, will not be described in this embodiment of the present invention.

Step 102: Calculate the covariance matrix of the original performance index matrix.

Optionally, in the embodiment of the present invention, the covariance matrix Σ of the original performance indicator matrix X can be calculated by the following formula:

Among them, μi is the expected value of the i-th indicator Xi, that is, μi=E(Xi); the (i, j)th element Σij in the covariance matrix Σ is the covariance between Xi and the j-th indicator Xj, where , i and j are any positive integers not greater than n. In addition, it should be noted that the indexes Xi and Xj described here generally refer to an index vector including corresponding m index values, which will not be described in detail in this embodiment of the present invention.

Step 103: For any predetermined index among the n network performance indexes, calculate each other index among the n network performance indexes and the predetermined index according to the original performance index matrix and the covariance matrix Mahalanobis distance.

Among them, the Mahalanobis distance was proposed by the Indian statistician Mahalanobis, which can represent the covariance distance of the data, and is an effective method for calculating the similarity between two unknown sample sets. Unlike Euclidean distance, Mahalanobis distance can take into account the connection between various characteristics (for example, a piece of information about height will bring a piece of information about weight, because the two are related) and is scale-independent (scale-invariant), that is, independent of the measurement scale. Therefore, compared with the Euclidean distance, which treats each dimension of the sample uniformly and does not consider the connection between features, it will avoid the loss of some information of the sample and make the calculation result of the similarity more accurate.

Optionally, in the embodiment of the present invention, the Mahalanobis distance Dij between any index Xj among the n network performance indexes and the predetermined index Xi can be calculated by the following formula:

Wherein, Σ is the covariance matrix of the original performance index matrix X, and i and j mentioned here are respectively any positive integers not greater than n; in addition, it should be noted that the index Xi described here , Xj may generally refer to an index vector including corresponding m index values, which will not be described in detail in this embodiment of the present invention.

Step 104: According to the calculated Mahalanobis distance between each other index and the predetermined index, determine the index closest to the change trend of the predetermined index among the n network performance indexes and the change with the predetermined index The indicator with the largest trend gap.

Optionally, according to the calculated Mahalanobis distance between each other index and the predetermined index, determine the index closest to the change trend of the predetermined index among the n network performance indexes and the distance between the predetermined index and the predetermined index. Indicators with the largest gaps in trend can include:

According to the Mahalanobis distance of each other index obtained by calculation and the predetermined index, the Mahalanobis distance matrix of 1 * (n-1) dimension is obtained, that is, the Mahalanobis distance vector of 1 * (n-1) dimension is obtained; it needs to be explained More importantly, if the Mahalanobis distance between the predetermined index and the predetermined index itself can be considered when calculating the Mahalanobis distance, the obtained Mahalanobis distance matrix can be a 1×n-dimensional Mahalanobis distance matrix;

According to the size of each matrix element in the Mahalanobis distance matrix, sort each matrix element in the Mahalanobis distance matrix in order from large to small;

The highest-ranked indicator is used as the indicator closest to the change trend of the predetermined indicator among the n network performance indicators, and the last-ranked indicator is used as the indicator with the largest difference from the change trend of the predetermined indicator.

Of course, it should be noted that the matrix elements in the Mahalanobis distance matrix can also be sorted in ascending order, and the index at the end of the sorting can be used as the index among the n network performance indexes that matches the predetermined For the index with the closest changing trend of the index, the index with the highest ranking is taken as the index with the largest difference from the changing trend of the predetermined index.

That is to say, the index with the smaller Mahalanobis distance is closer to the change of the predetermined index, and the change gap between the index with the larger Mahalanobis distance and the predetermined index is larger.

Optionally, since different network performance indicators have different dimensions, in order to compare the changes of each indicator horizontally, a normalization operation can be performed on each indicator data sample first, that is, after obtaining the original performance indicator matrix , and before calculating the covariance matrix of the original performance index matrix, the method may further include:

A normalization operation is performed on each index value of the acquired n network performance indexes, so that when calculating the Mahalanobis distance subsequently, the Mahalanobis distance between each normalized index Xjnorm and the predetermined index Xinorm can be calculated.

Optionally, the following formula can be used to perform a normalization operation on each index value of the obtained n network performance indicators, so as to normalize each index to the interval [0, 1]:

Among them, Xij may refer to the index value of the i (i=1,...,n) network performance index at the j (j=1,...,m) time point, and Xij _norm may refer to the Xij's Normalized value; min(Xi) may refer to the minimum index value corresponding to the i-th (i=1,...,n) network performance index, and max(Xi) may refer to the i-th (i=1,... , n) the maximum index value corresponding to the network performance indexes.

Further, according to the calculated Mahalanobis distance between each other index and the predetermined index, determine the index closest to the change trend of the predetermined index among the n network performance indexes and the distance between the predetermined index and the predetermined index. After the index with the largest variation trend gap is selected, the method may further include the following steps:

According to the determined index closest to the change trend of the predetermined index and the index most different from the change trend of the predetermined index, determine whether the influence of the adjustment of the network parameter on the change of the index is reasonable, so as to judge whether the adjustment of the network parameter is reasonable.

That is to say, assuming that the index i decreases during the observation period, network optimization personnel need to know which indicators are also declining and which indicators will rise during the analysis process, whether the KPIs involved in network adjustment meet expectations, and whether they are hidden or not. Contain ground spread to other indicators. Based on optimization experience, optimizers extract a batch of KPI data, which is defined as "experience collection", which must be mastered by adjusters.

However, for the strongly correlated indicators that do not belong to the "experience set", the network optimization personnel can use the Mahalanobis distance matrix D to find out that among all n-1 network performance indicators, the similarity with the indicator i's change trend is large or the change trend similarity For small indicators, analyze whether the influence of parameter adjustment on indicator changes is reasonable, and find out the influencing factors that may be missed outside the experience of network optimization personnel, so as to improve the accuracy of network optimization.

Embodiment 1 of the present invention provides a method for evaluating the change trend of network performance indicators. The method may include: obtaining the index values of n network performance indicators at a set geographic granularity at m consecutive time points, and obtaining n network performance indicators. The n×m-dimensional original performance index matrix of the performance index, and then by calculating the covariance matrix of the original performance index matrix, calculating the Mahalanobis distance between each other index in the n network performance indexes and the predetermined index, and then obtaining the predetermined index The indicator with the closest variation trend and the indicator with the largest difference from the variation trend of the predetermined indicator, wherein, the n is a positive integer greater than 1, and the m is a positive integer not less than 1.

That is to say, in the technical solution described in Embodiment 1 of the present invention, data mining can be used to correlate and cluster the network performance index data based on different network parameters, so as to explore and discover the similarity of unknown change trends between the indexes, It is expected to solve problems such as potential network performance failures, so as to realize the automatic judgment of the trend of network performance indicators. Compared with the existing methods, network optimization personnel do not need to draw indicator trend graphs again when analyzing the trend of indicator changes. Operations such as observation with human eyes can greatly save manpower and time, and improve analysis efficiency; in addition,

According to the method provided by the embodiment of the present invention, network optimization personnel can quickly analyze the change trends of all indicators, thereby breaking through the limitation of selecting limited indicators for analysis based on previous optimization experience, and can comprehensively analyze the status of network indicators And a comprehensive understanding of the impact of the optimization scheme on the quality of the live network, which can also improve the comprehensiveness and scalability of network optimization analysis.

Embodiment two:

Based on the same inventive concept as the first embodiment, the second embodiment of the present application provides a device for evaluating the change trend of network performance indicators. To repeat, as shown in Figure 2, the device mainly includes:

The indicator acquisition module 21 can be used to acquire the index values of n network performance indicators in a set geographic granularity at consecutive m time points, and obtain an n×m-dimensional original performance indicator matrix of n network performance indicators; wherein, the n is a positive integer greater than 1, and m is a positive integer not less than 1;

A covariance calculation module 22, which can be used to calculate the covariance matrix of the original performance index matrix obtained by the index acquisition module 21;

The indicator distance calculation module 23 can be used for any predetermined indicator in the n network performance indicators, according to the original performance indicator matrix obtained by the indicator acquisition module 21 and the covariance calculation module. The covariance matrix is used to calculate the Mahalanobis distance between each of the n network performance indicators and the predetermined indicator;

The indicator analysis module 24 can be used to determine the change trend of the n network performance indicators that is the closest to the predetermined indicator according to the Mahalanobis distance between each other indicator calculated by the indicator distance calculation module 23 and the predetermined indicator. The indicators that are close to each other and the indicators that are farthest from the change trend of the predetermined indicator.

Optionally, the covariance calculation module 22 can be specifically configured to calculate the covariance matrix Σ of the original performance indicator matrix X through the following formula:

Among them, μi is the expected value of the i-th indicator Xi, and the (i, j)th element Σij in the covariance matrix Σ is the covariance between Xi and the j-th indicator Xj, and i and j are any positive integer.

Optionally, the indicator distance calculation module 23 can be specifically configured to calculate the Mahalanobis distance Dij between any indicator Xj among the n network performance indicators and the predetermined indicator Xi through the following formula:

Wherein, Σ is the covariance matrix of the original performance index matrix X, and i and j are any positive integers not greater than n.

Optionally, the index analysis module 24 can be specifically configured to obtain a 1×(n-1)-dimensional Mahalanobis distance according to the Mahalanobis distance between each other index calculated by the index distance calculation module 23 and the predetermined index matrix; and according to the size of each matrix element in the Mahalanobis distance matrix, sort each matrix element in the Mahalanobis distance matrix in order from large to small; Among the n network performance indicators, the index that is closest to the change trend of the predetermined index is the index that is ranked last as the index that is farthest from the change trend of the predetermined index.

Optionally, the device may also include a normalization module 25:

The normalization module 25 may be configured to, before the covariance calculation module 22 calculates the covariance matrix of the original performance indicator matrix, each index value of the n network performance indicators acquired by the indicator acquisition module 21 Perform a normalization operation.

Optionally, the device may also include an indicator evaluation module 26:

The indicator evaluation module 26 can be used to determine the number of the n network performance indicators in the indicator analysis module 24 according to the Mahalanobis distance between each other indicator calculated by the indicator distance calculation module 23 and the predetermined indicator. After the index closest to the change trend of the predetermined index and the index most different from the change trend of the predetermined index, according to the determined index closest to the change trend of the predetermined index and the change of the predetermined index The index with the largest trend gap determines whether the influence of network parameter adjustment on index changes is reasonable, so as to judge whether the network parameter adjustment is reasonable.

Those skilled in the art should understand that the embodiments of the present invention may be provided as methods, devices (devices), or computer program products. Accordingly, the present invention can take the form of an entirely hardware embodiment, an entirely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) having computer-usable program code embodied therein.

The present invention is described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (devices) and computer program products according to embodiments of the invention. It should be understood that each procedure and/or block in the flowchart and/or block diagram, and a combination of procedures and/or blocks in the flowchart and/or block diagram can be realized by computer program instructions. These computer program instructions may be provided to a general purpose computer, special purpose computer, embedded processor, or processor of other programmable data processing equipment to produce a machine such that the instructions executed by the processor of the computer or other programmable data processing equipment produce a An apparatus for realizing the functions specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing apparatus to operate in a specific manner, such that the instructions stored in the computer-readable memory produce an article of manufacture comprising instruction means, the instructions The device realizes the function specified in one or more procedures of the flowchart and/or one or more blocks of the block diagram.

These computer program instructions can also be loaded onto a computer or other programmable data processing device, causing a series of operational steps to be performed on the computer or other programmable device to produce a computer-implemented process, thereby The instructions provide steps for implementing the functions specified in the flow chart or blocks of the flowchart and/or the block or blocks of the block diagrams.

While preferred embodiments of the invention have been described, additional changes and modifications to these embodiments can be made by those skilled in the art once the basic inventive concept is appreciated. Therefore, it is intended that the appended claims be construed to cover the preferred embodiment as well as all changes and modifications which fall within the scope of the invention.

Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and equivalent technologies thereof, the present invention also intends to include these modifications and variations.

Claims (12)

1. a kind of method of assessment network performance index variation tendency is it is characterised in that include：

Obtain n network performance index setting in geographical granularity in the desired value of continuous m time point, obtain Original performance index matrix to n × m dimension of n network performance index；

Calculate the covariance matrix of described original performance index matrix；

For the arbitrary desired indicator in described n network performance index, according to described original performance index square Battle array and described covariance matrix, other indexs each calculating in described n network performance index are pre- with described Determine the mahalanobis distance of index；

According to calculated other indexs each and the mahalanobis distance of described desired indicator, determine described n net The immediate index of the variation tendency with described desired indicator in network performance indications and with described predetermined finger Target variation tendency index with the biggest gap；

Wherein, described n is the positive integer more than 1, and described m is the positive integer not less than 1.

2. the method for claim 1 it is characterised in that calculated described original by below equation The covariance matrix ∑ of performance indications matrix X：

Wherein, μ i is the expected value of i-th index Xi, (i, j) the individual unit in covariance matrix ∑ Plain ∑ ij is the covariance of Xi and j-th index Xj, and i, j are respectively any positive integer being not more than n.

3. the method for claim 1 is it is characterised in that calculate described n by below equation Arbitrary index Xj in network performance index and the mahalanobis distance Dij of described desired indicator Xi：

$D_{ij}=\sqrt{\left(\right(X_i-X_j\left){\mathrm{\Σ}}^{-1}{\left(X_i-X_j\right)}^T\right)};$

Wherein, ∑ is the covariance matrix of described original performance index matrix X, and i, j are respectively and are not more than n Any positive integer.

4. the method for claim 1 is it is characterised in that according to calculated other indexs each With the mahalanobis distance of described desired indicator, determine in described n network performance index with described desired indicator The immediate index of variation tendency and the index with the biggest gap with the variation tendency of described desired indicator, bag Include：

According to calculated other indexs each and the mahalanobis distance of described desired indicator, obtain 1 × (n-1) The mahalanobis distance matrix of dimension；

According to the size of each matrix element in described mahalanobis distance matrix, according to order from big to small to institute The each matrix element stated in mahalanobis distance matrix is ranked up；

Using the index the most front that sorts as the change with described desired indicator in described n network performance index The immediate index of trend, will sort last index as the variation tendency gap with described desired indicator Big index.

5. the method for claim 1 is it is characterised in that calculating described original performance index square Before the covariance matrix of battle array, methods described also includes：

Operation is normalized to each desired value of the n network performance index getting.

6. the method for claim 1 is it is characterised in that referring to according to calculated each other Mark and the mahalanobis distance of described desired indicator, determine in described n network performance index with described predetermined finger The immediate index of target variation tendency and the index with the biggest gap with the variation tendency of described desired indicator Afterwards, methods described also includes：

According to determine the immediate index of the variation tendency with described desired indicator and with described predetermined finger Target variation tendency index with the biggest gap, determines whether the impact that the adjustment of network parameter changes to index closes Reason, to judge that network parameter regulates whether rationally.

7. a kind of equipment of assessment network performance index variation tendency is it is characterised in that include：

Index selection module, individual in continuous m for obtaining the n network performance index setting in geographical granularity The desired value of time point, obtains the original performance index matrix of n × m dimension of n network performance index；Its In, described n is the positive integer more than 1, and described m is the positive integer not less than 1；

Covariance computing module, for calculating the described original performance index obtained by described index selection module The covariance matrix of matrix；

Index distance calculation module, for for the arbitrary desired indicator in described n network performance index, Calculated according to the described original performance index matrix obtained by described index selection module and described covariance The calculated described covariance matrix of module institute, each other calculating in described n network performance index refer to Mark and the mahalanobis distance of described desired indicator；

Indicator analysis module, for according to calculated other indexs each of described index distance calculation module with The mahalanobis distance of described desired indicator, determine in described n network performance index with described desired indicator The immediate index of variation tendency and the index with the biggest gap with the variation tendency of described desired indicator.

8. equipment as claimed in claim 7 is it is characterised in that described covariance computing module is specifically used In the covariance matrix ∑ calculating described original performance index matrix X by below equation：

Wherein, μ i is the expected value of i-th index Xi, (i, j) the individual unit in covariance matrix ∑ Plain ∑ ij is the covariance of Xi and j-th index Xj, and i, j are respectively any positive integer being not more than n.

9. equipment as claimed in claim 7 is it is characterised in that described index distance calculation module is concrete For calculating arbitrary index Xj in described n network performance index and described predetermined finger by below equation The mahalanobis distance Dij of mark Xi：

$D_{ij}=\sqrt{\left(\right(X_i-X_j\left){\mathrm{\Σ}}^{-1}{\left(X_i-X_j\right)}^T\right)};$

Wherein, ∑ is the covariance matrix of described original performance index matrix X, and i, j are respectively and are not more than n Any positive integer.

10. equipment as claimed in claim 7 it is characterised in that

Described indicator analysis module, specifically for according to calculated other indexs each and described desired indicator Mahalanobis distance, obtain the mahalanobis distance matrix that 1 × (n-1) ties up；And according in described mahalanobis distance matrix Each matrix element size, according to order from big to small to each matrix element in described mahalanobis distance matrix Element is ranked up；And, using sort the most front index as in described n network performance index with described The immediate index of variation tendency of desired indicator, using sort last index as with described desired indicator Variation tendency index with the biggest gap.

11. equipment as claimed in claim 7 are it is characterised in that described equipment also includes normalization module：

Described normalization module, for calculating described original performance index matrix in described covariance computing module Covariance matrix before, each index to the n network performance index that described index selection module gets Value is normalized operation.

12. equipment as claimed in claim 7 are it is characterised in that described equipment also includes index evaluation mould Block：

Described index evaluation module, in described indicator analysis module according to calculated other indexs each With the mahalanobis distance of described desired indicator, determine in described n network performance index with described desired indicator The immediate index of variation tendency and the index with the biggest gap with the variation tendency of described desired indicator it Afterwards, according to determine the immediate index of the variation tendency with described desired indicator and with described desired indicator The with the biggest gap index of variation tendency, determine whether the adjustment of the network parameter impact to index change closes Reason, to judge that network parameter regulates whether rationally.