JP2019016958A - Quality deterioration estimation device, quality deterioration estimation method and program - Google Patents

Abstract

To provide a quality deterioration estimation device for estimating a quality deterioration value for each node by using end-to-end observation values in a communication network, which is capable of improving the estimation accuracy of the quality deterioration.SOLUTION: The quality deterioration estimation device for estimating quality deterioration on a node constituting a communication network, comprises: input means for inputting quality observation value of each observation path, the observation number of each observation route and a piece of information of observation paths; calculation means configured to calculate an estimation value of deterioration on each node on the basis of the observed quality value, the number of observations used for weighing to an unknown value of deterioration on each node and the path information; and output means for outputting the estimation value of deterioration on each node.SELECTED DRAWING: Figure 1

Description

  The present invention relates to a technique for estimating / determining how much deterioration has occurred in connection points of nodes in a communication using an end-to-end quality observation value in the communication as an input.

  In mobile communications, quality degradation occurs in various locations due to diversification of user environments and diversification of usage methods. In order to improve the quality of mobile communication services, it is important to understand the degradation location.

  However, it is possible to perform end-to-end observation (may be referred to as “measurement”) in quality observation, but the degree of deterioration at each location (server, ISP, access node, terminal, etc.) based on the observation result It is difficult to judge. As a prior art regarding this point, there are techniques disclosed in Non-Patent Documents 1 and 2. Non-Patent Document 1 discloses a technique for estimating the deterioration amount of each node with a small amount of calculation by using the relationship of nodes for estimating the deterioration degree as a graph structure.

  Non-Patent Document 2 discloses a technique for improving estimation accuracy by rejecting observation data with low reliability.

Tayuki et al., "Estimation of communication quality degradation using geographical dependence of mobile network", IEICE Technical Report, vol. 116, no. 10, CQ2016-1, pp. 1-6, April 2016 Kim Jong et al., "Selection of observation data in delayed tomography considering geographical dependence," IEICE Society Conference 2016, B-11-7, September 2016

  In the prior art, for example, when the degradation location is estimated using the end-to-end quality observed in the mobile terminal, the degradation location is estimated due to a small number of terminals to be observed or the measurement performance of the terminal is low. Therefore, there arises a problem that high-accuracy observation data necessary for this cannot be obtained. In the technique disclosed in Non-Patent Document 1, since the degradation location is estimated using the relationship between the nodes, such low-reliability observation data is estimated as the degradation amount of the node through which the observation data passes. It may affect not only the accuracy but also the multiple estimation accuracy of the neighborhood. In addition, when the observation data is small, even if the method of Non-Patent Document 2 is used, the effect is small, and the sample variance of the observation value becomes large.

  The present invention has been made in view of the above points, and improves the estimation accuracy of the quality degradation amount in the technique for estimating the value of the quality degradation amount for each node using the end-to-end observation value in the communication network. The purpose is to provide technology.

According to the disclosed technology, a quality degradation estimation apparatus that estimates quality degradation of nodes constituting a communication network,
Input means for inputting quality observation values of each observation route, the number of observations of each observation route, and route information of each observation route;
An arithmetic means for calculating an estimated value of the deterioration amount of each node based on the quality observation value, the number of observations used for weighting the unknown deterioration amount of each node, and the path information;
An output means for outputting an estimated value of the deterioration amount of each node is provided.

  According to the disclosed technique, a technique for improving the estimation accuracy of the quality degradation amount is provided in the technology for estimating the value of the quality degradation amount for each node using the end-to-end observation value in the communication network.

It is a block diagram of the quality degradation estimation apparatus 100 in embodiment of this invention. It is a hardware block diagram of the quality degradation estimation apparatus 100. It is a figure which shows the graph structure between the nodes in a group 1, and the degradation amount of a node. FIG. 10 is a diagram showing the deterioration amount of nodes belonging to group 2. It is a figure which shows dispersion | distribution of a degradation estimated amount.

  Hereinafter, an embodiment (this embodiment) of the present invention will be described with reference to the drawings. The embodiment described below is merely an example, and the embodiment to which the present invention is applied is not limited to the following embodiment. In the communication network of the present embodiment (hereinafter referred to as a target communication network), a communication device such as a server, a relay device, or a base station that is a target for estimating the amount of quality degradation is referred to as a “node”. The “node” includes not only a single network device but also a configuration including a set of network devices such as an ISP and a carrier network. The target communication network is not limited to a specific type of network, and may be, for example, a mobile network, a fixed network, or a network in which a mobile network and a fixed network are mixed. However, other nets may be used.

  In the present embodiment, each node constituting the target communication network has a value, and the larger the value, the larger the deterioration amount. Then, the quality degradation estimation apparatus 100 estimates the value of each node constituting the communication network from the quality observation value between any two nodes (end-end quality observation value). Hereinafter, the apparatus configuration and processing contents will be described.

(Device configuration)
FIG. 1 shows a configuration diagram of a quality degradation estimation apparatus 100 in the present embodiment. As shown in FIG. 1, the quality degradation estimation apparatus 100 includes an input unit 101 that inputs data used for quality degradation estimation, a computation unit 102 that performs quality degradation estimation, and an output unit 103 that outputs the result of the computation. And a data storage unit 104 that stores input data, mid-calculation data, calculation results, programs, and the like.

  The quality degradation estimation apparatus 100 in the present embodiment can be realized by causing a computer to execute a program describing the processing contents described in the present embodiment. That is, the function of the quality degradation estimation apparatus 100 can be realized by executing a program corresponding to processing executed in the apparatus using hardware resources such as a CPU and a memory built in the computer. Is possible. The above-mentioned program can be recorded on a computer-readable recording medium (portable memory or the like), stored, or distributed. It is also possible to provide the program through a network such as the Internet or electronic mail.

  FIG. 2 is a diagram illustrating a hardware configuration example of the quality degradation estimation apparatus 100 according to the present embodiment. 2 includes a drive device 150, an auxiliary storage device 152, a memory device 153, a CPU 154, an interface device 155, a display device 156, an input device 157, and the like that are mutually connected by a bus B. .

  A program that realizes processing in the quality degradation estimation apparatus 100 is provided by a recording medium 151 such as a CD-ROM or a memory card, for example. When the recording medium 151 storing the program is set in the drive device 150, the program is installed from the recording medium 151 into the auxiliary storage device 152 via the drive device 150. However, the program does not necessarily have to be installed from the recording medium 151, and may be downloaded from another computer via a network. The auxiliary storage device 152 stores the installed program and also stores necessary files and data.

  The memory device 153 reads the program from the auxiliary storage device 152 and stores it when there is an instruction to start the program. The CPU 154 realizes functions related to the quality degradation estimation apparatus 100 in accordance with a program stored in the memory device 153. The interface device 155 is used as an interface for connecting to a network. The display device 156 displays a GUI (Graphical User Interface) or the like by a program. The input device 157 includes a keyboard and mouse, buttons, a touch panel, and the like, and is used to input various operation instructions.

(Operation of the quality degradation estimation apparatus 100)
Next, as an operation of the quality degradation estimation apparatus 100, the quality degradation estimation method in the present embodiment will be described in detail.

<Definition>
As described above, a communication device such as a server, a relay device, or a base station that is a target for estimating the amount of quality degradation in the target communication network is called a node. A set of a plurality of nodes of the same type, such as a plurality of servers and a plurality of terminals, is called a group. In the present embodiment, there are group 1 and group 2, the nodes belonging to group 1 are represented as a ₁ , a ₂ ,..., A _Q1, and the nodes belonging to group 2 are represented as b ₁ , b _{2 ,.} Represent. Further, the deterioration amount of the node belonging to the group 1 is expressed as x = (x ₁ x ₂ ... X _Q1 ) ^、, and the deterioration amount of the node belonging to the group 2 is expressed as y = (y ₁ y ₂ ... Y _Q2 ) ^Τ .

  For the nodes belonging to group 1, the similarity of deterioration amounts between nodes is represented by a graph structure, and the undirected graph is represented by G. Note that the similarity of the deterioration amounts is whether the deterioration amounts are similar.

A path between the node a _i belonging to the group 1 and the node b _j belonging to the group 2 is represented as an observation path (i, j).

<Input>
At least the following data is input to the input unit 101 of the quality degradation estimation apparatus 100. The input data is stored in the data storage unit 104, and is read out and calculated by the calculation unit 102 in the subsequent processing. In addition to the following input data, values such as a threshold necessary for processing are also input as appropriate.

Input 1: Observation path identifier M, number of observation paths 1, 2,…, m,… M, number of observations per observation path N ₁ , N ₂ ,… N _m ,…, N _M , quality observation per observation path Data z _m ^(k) (m = 1, 2,…, M, k = 1, 2,…, N _m )
Input 2: Graph Laplacian for graph structure G that represents the similarity of degradation between nodes in group 1: L
Input 3: Matrix representing path information (nodes composing communication) of each observation path when input 1 is observed: A ₁ , A ₂
<Details of input 2>
A method for obtaining a graph Laplacian L constituting the graph G of the number of nodes Q ₁ of the group 1 with respect to the input 2 will be described.

A distance d _{j, k} between nodes a _j and a _k included in the graph G is defined. This distance d is an index representing the similarity of deterioration between nodes. The distance d can be a geographical distance, for example. However, the distance d is not limited to this, and may be any value as long as the similarity is a decreasing function (eg, an inversely proportional function) of the distance d. Here, the attenuation function is defined as f (x).

  The graph Laplacian L is calculated by L = D-C. C and D are as follows.

C is an N × N adjacency matrix, and the element c _{j, k} (j, k = 1,2, ..., Q ₁ ) of C is 0 when j = k, c _{j, k} = c _{k, Satisfies j} . c _{j, k} when j = k is not different between the unweighted graph and the weighted graph as follows.

In the case of a weightless graph in which no weight is given to the branch of the graph G, c _{j, k} = 1 if d _{j, k} is equal to or less than a certain threshold value e *, and c _{j, k} = 0 otherwise. In the case of a weighted graph, c _{j, k} = f (d _{j, k} ).

D is an N × N degree matrix, and the elements g _{j, k} (j, k = 1,2,..., N) of D are diagonal matrices of 0 when j ≠ k, and g _{j, j} Is given by:

＜入力３の詳細＞
入力３のA₁は、0あるいは1からなるM×Q₁の行列であり、入力３のA₂は、0あるいは1からなるM×Q₂の行列である。

<Details of input 3>
A ₁ of input 3 is an M × Q ₁ matrix consisting of 0 or 1, and A ₂ of input 3 is an M × Q ₂ matrix consisting of 0 or 1.

When the m-th (m = 1,2, ... M) observation path in the number M of observation paths is (i _m , j _m ), the (m, q ₁ ) component (m = 1,2, ... M) of A ₁ , q ₁ = 1, 2,... Q ₁ ) is ₁ if the observation path (i _m , j _m ) passes through the node a _q1, and 0 otherwise. In addition, the (m, q ₂ ) component of A2 (m = 1,2,… M, q ₂ = 1,2, ... Q ₂ ), the observation path (i _m , j _m ) passes through the node b _q2 1 if yes, 0 otherwise.

  In the present embodiment, the input 2 (graph Laplacian L) may be calculated in advance outside the quality degradation estimation device 100 and input to the quality degradation estimation device 100, or may be input to the quality degradation estimation device 100. By giving the information of the graph G and the distance d to the unit 101, the calculation unit 102 may calculate the graph Laplacian L and store it in the data storage unit 104. Also, U, which will be described later, may be calculated in advance outside the quality degradation estimation apparatus 100 and input to the quality degradation estimation apparatus 100, or may be calculated by the quality degradation estimation apparatus 100.

(Processing procedure for estimating quality degradation)
Next, processing for estimating quality degradation performed by the calculation unit 102 based on the inputs 1 to 3 and the like will be described. Basically, the results of the processing executed by the calculation unit 102 are sequentially stored in the data storage unit 104, and read and used by the calculation unit 102 in the next processing.

<Create U>
First, the calculation unit 102 obtains eigenvalues λ _i (i = 1, 2,... Q ₁ ) of the graph Laplacian L of the graph G. Also, an eigenvector u _i corresponding to the eigenvalue λ _i is obtained, and a matrix U = (u ₁ u ₂ ... U _Q1 ) (Q1 × Q1 matrix) in which eigenvectors are arranged is created. Here, U is an orthonormal matrix.

<Node quality estimation process>
The m-th (m = 1, 2,... M) observation path (i _m , j _m ) in the number M of observation paths is a path between the group a node a _im and the group 2 node b _jm . The end-to-end quality obtained by the k-th observation of the observation path (i _m , j _m ⁾ is expressed as z _m ^(k) . At this time, the observation data z _m obtained on the observation path (i _m , j _m ) is given by the following equation. That is, the observation data z _m is given as the sum of each observation value z _m ^(k) in N _m observations obtained through the same observation path.

観測データを表す集合を観測ベクトルz=(z₁ z₂ … z_M )^Τと表す。また、各観測経路における観測数N_m（m=1,2,…,M）を対角要素とする対角行列をN^*=diag(N₁ N₂ … N_M)とする。

A set representing observation data is represented as observation vector z = (z ₁ z ₂ ... Z _M ) ^Τ . In addition, a diagonal matrix whose diagonal elements are the number of observations N _m (m = 1, 2,..., M) in each observation path is N ^* = diag (N ₁ N ₂ ... N _M ).

  The relationship between z obtained as described above, the degradation amount x of the nodes belonging to group 1, and the degradation amount y of the nodes belonging to group 2 is given by the following equation.

z = N ^* A ₁ x + N ^* A ₂ y
In the above equation, multiplying N ^{* corresponds} to multiplying the deterioration amount by the number of observations (weight) for each observation path.

Further, x is expressed as x = Us using the matrix U. However, s = (s ₁ s ₂ ... S _N1 ) ^Τ . Thus, the above relational expression can be expressed as follows.

z = N ^* A ₁ Us + N ^* A ₂ y
As described above, since a relationship in which the relationship between the observation (observation value) vector and the unknown vector is expressed linearly is obtained, the calculation unit 102 solves the linear inverse problem based on the relationship as an optimization problem. Thus, the vectors s and y are estimated and further converted to estimate the vector x that is the value of each node in group 1. Specifically, it is as follows. Hereinafter, Lasso (Least absolute shrinkage and selection operator) is used as an estimation method, but estimation may be performed by a method other than Lasso. Lasso is a normalized estimation method using the sum of absolute values (L1 norm) of regression coefficients as a normalization term. As a reference for the following estimation method, for example, “Seung-Jean Kim; Stanford Univ., Stanford; Koh, K.; Lustig, M.; Boyd, S. et al.“ An Interior-Point Method for Large-Scale l1- Regularized Least Squares "".

  Specifically, the calculation unit 102 estimates a vector s by solving a Lasso estimation expression represented by the following expression using sparse approximation.

ただし、λ₁、λ₂はパラメータである。また、ベクトルw=(w₁ w₂ … w_N )^Τに対し、||w||_l（l=1,2）は次式で与えられる。

However, λ ₁ and λ ₂ are parameters. Further, with respect to the vector _{w = (w 1 w 2 ...} w N) Τ, || w || l (l = 1,2) is given by the following equation.

演算部１０２は、上記の最適化問題の解として得られたsを用いて、x=Usによりxを求める。

The calculation unit 102 obtains x by x = Us using s obtained as a solution to the optimization problem.

<Output>
Subsequently, the output unit 103 outputs the obtained result. In this embodiment, x and y are output. Note that outputting the values of x and y is an example, and another value obtained from x and y may be output.

As described above, in the present embodiment, it is possible to perform estimation in consideration of weighting according to the number of observations by applying N ^* .

(Example)
An embodiment as a specific example in which the quality degradation estimation apparatus 100 performs quality degradation estimation by the above-described method will be described.

In this embodiment, the number of nodes in group 1 is Q ₁ = 64, the number of nodes in group 2 is Q ₂ = 10, the graph structure between nodes in group 1 and the amount of degradation are given as shown in FIG. The deterioration amount of the node belonging to is given as shown in FIG.

Also, assuming that the number of observation paths is M = 2N ₁ , N _m = 50 (1-ρ) at m = 1, N _m = 50 (1 + ρ) at m = 2, m = 3,4, ..., M N _m = 50.

  With the above settings, the degradation amount estimation experiment was performed 200 times for each of ρ = 0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, and 0.9. FIG. 5 shows the variance of the estimated deterioration amount obtained by the experiment. The method of the present invention is expressed as “weighted”, the method of “Non-Patent Document 1” as “No Weight”, and the method of “Non-Patent Document 2” as “Rejected”. As shown in FIG. 5, the method of the present invention does not depend on the number of observations, and the estimation value has the smallest variance, and robust estimation is possible.

(Effect of embodiment)
For example, when the existing technology of Non-Patent Document 1 is used to estimate the amount of degradation of communication devices constituting a mobile network, stable and reliable observation data is always obtained due to a phenomenon unique to mobile communication such as movement of a terminal. It cannot be acquired. When low-reliability observation data is used, it affects not only the amount of deterioration of the communication device on the observation path from which the observation data was acquired but also the estimation accuracy of other communication devices.

  The technology according to the present embodiment is a technology that compensates for such problems of the existing technology, and actively uses the interpolation function based on the relationship between nodes and performs quality degradation by weighting according to the number of observations. By estimating the amount, it is possible to improve the estimation accuracy.

(Summary of embodiment)
As described above, according to the present embodiment, a quality deterioration estimation device that estimates the quality deterioration of nodes constituting a communication network, the quality observation value of each observation path, the number of observations of each observation path, and each Based on the input means for inputting the route information of the observation route, the quality observation value, the number of observations used for weighting the unknown deterioration amount of each node, and the route information, the deterioration amount of each node There is provided a quality deterioration estimation device comprising: an arithmetic means for calculating an estimated value; and an output means for outputting an estimated value of the deterioration amount of each node.

  The input unit 101, the calculation unit 102, and the output unit 103 described in the embodiment are examples of an input unit, a calculation unit, and an output unit, respectively.

  The computing means is a graph Laplacian matrix corresponding to a vector composed of the quality observation values, a matrix composed of the path information, the number of observations, and a graph structure representing similarity of deterioration between nodes. The estimated value of the deterioration amount of each node may be calculated by solving an estimation formula using a matrix in which eigenvectors are arranged.

The estimation formula is an estimation formula based on the relationship that the sum of the number of observations of quality observation values in a certain observation path is equal to a value obtained by weighting the deterioration amount of the nodes constituting the observation path based on the number of observations. It is good. Note that z = N ^* A ₁ x + N ^* A ₂ y is an example of the above relationship.

  Although the present embodiment has been described above, the present invention is not limited to the specific embodiment, and various modifications and changes can be made within the scope of the gist of the present invention described in the claims. Is possible.

DESCRIPTION OF SYMBOLS 100 Quality degradation estimation apparatus 101 Input part 102 Operation part 103 Output part 104 Data storage part 150 Drive apparatus 151 Recording medium 152 Auxiliary storage apparatus 153 Memory apparatus 154 CPU
155 Interface device 156 Display device 157 Input device

Claims (7)

A quality degradation estimation device for estimating quality degradation of nodes constituting a communication network,
Input means for inputting quality observation values of each observation route, the number of observations of each observation route, and route information of each observation route;
An arithmetic means for calculating an estimated value of the deterioration amount of each node based on the quality observation value, the number of observations used for weighting the unknown deterioration amount of each node, and the path information;
An output means for outputting an estimated value of the deterioration amount of each node. The computing means is a graph Laplacian matrix corresponding to a vector composed of the quality observation values, a matrix composed of the path information, the number of observations, and a graph structure representing similarity of deterioration between nodes. The quality degradation estimation apparatus according to claim 1, wherein an estimated value of the degradation amount of each node is calculated by solving an estimation formula using a matrix in which eigenvectors are arranged. The estimation formula is an estimation formula based on the relationship that the sum of the number of observations of quality observation values in a certain observation path is equal to a value obtained by weighting the deterioration amount of nodes constituting the observation path based on the number of observations. The quality degradation estimation apparatus according to claim 2, wherein: A quality degradation estimation method executed by a quality degradation estimation device that estimates quality degradation of nodes constituting a communication network,
An input step for inputting quality observation values for each observation route, the number of observations for each observation route, and route information for each observation route;
A calculation step of calculating an estimated value of the deterioration amount of each node based on the quality observation value, the number of observations used for weighting the unknown deterioration amount of each node, and the path information;
An output step of outputting an estimated value of the deterioration amount of each node. In the calculation step, the quality degradation estimation device is configured to include a vector composed of the quality observation values, a matrix composed of the path information, the number of observations, and a graph structure representing similarity between degradations between nodes. 5. The quality degradation estimation method according to claim 4, wherein an estimated value of the degradation amount of each node is calculated by solving an estimation formula using a matrix in which eigenvectors of a graph Laplacian matrix corresponding to the matrix are arranged. . The estimation formula is an estimation formula based on the relationship that the sum of the number of observations of quality observation values in a certain observation path is equal to a value obtained by weighting the deterioration amount of nodes constituting the observation path based on the number of observations. The method for estimating quality degradation according to claim 5.   The program for functioning a computer as each means in the quality degradation estimation apparatus of any one of Claims 1 thru | or 3.

Images