JP2003318985A - Method and apparatus for monitoring quality status on packet network - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide a quality status monitoring method with which the occurrence of quality reduction and a portion where a quality is lowered can be specified on a packet network. <P>SOLUTION: In the method for monitoring a quality status on the packet network with which a network quality is monitored by comparing a transmitting status of a monitoring target packet with a receiving status and a block where the quality is lowered is decided when the quality reduction is found out, a plurality of measuring instruments (11, 12 and 13) are installed on the network to be a quality measuring and monitoring target, a quality evaluation is performed in each measuring instrument by calculating a mutual correlative function based upon a packet passing time sequence obtained by probing the packet (21 and 22) and when the occurrence of the quality reduction is determined (23), a logic topology to be a tree structure is created from packet transfer route information (30). By utilizing the logic topology, the portion where the quality is lowered is decided (40). <P>COPYRIGHT: (C)2004,JPO

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a packet network quality status monitoring method for monitoring the quality status of a packet network and determining the part when quality information is found to be degraded. In particular, the present invention is a quality status monitoring method suitable for monitoring priority transfer control in a network in which the transfer priority is classified by using the transfer priority of packets as a transfer class and QoS control is performed to perform priority transfer control for each class. It is about. Furthermore, the present invention relates to a quality status monitoring method suitable for monitoring the transfer quality of packets of real-time applications that are easily affected by delay fluctuations and band fluctuations such as video and audio transfer.

[0002]

2. Description of the Related Art In order to realize quality control monitoring by class-based priority control, it is considered that it can be realized by comparatively monitoring the state of delay variation between classes. Further, in order to transfer a real-time application with high quality, it is required that the delay variation is small. Therefore, in order to realize quality monitoring, a method of calculating the one-way delay time as described below can be considered. -Frequency synchronization and time synchronization are performed in the packet probe device. -Probe a packet between packet probe devices. -Receives probe data strings with probe time from the packet probe devices on the transmitting and receiving sides. -Calculate the one-way delay time from the probe time of the detected packet and obtain the distribution.

Next, as another method of monitoring the quality, a method of calculating the jitter indicating the delay variation can be considered. The following processing must be performed to obtain the jitter. -Perform frequency synchronization in the packet probe device. -Probe a packet between packet probe devices. -Receives a probe data string with a probe time from the sending and receiving packet probe devices. -Search for the same packet from these data strings. -Calculate the one-way delay time from the probe time between transmission and reception. -Calculate the jitter from the one-way delay time to obtain the distribution.

As described above, in order to measure the one-way delay time, it is necessary to calculate the one-way delay time by obtaining the difference between the packet transmission time and the packet reception time. For this reason, in the measurement of the one-way delay time, it is required that all the time points of the packet probe device match (time synchronization), and then the clock progresses be synchronized (frequency synchronization).

A method of measuring jitter and performing quality monitoring is as follows.
Since the jitter is calculated as the difference in one-way delay time, accurate time synchronization is not required. Therefore, in order to obtain the jitter distribution and perform quality control monitoring, it is sufficient that frequency synchronization of the packet probe device can be realized. In order to obtain the one-way delay time and the jitter, it is necessary to perform the same packet search as many times as the number of probe packets from the set of data strings on the transmitting side and the receiving side. Further, the method of obtaining the distribution of one-way delay and jitter does not consider the packet loss resulting from the quality variation.

[0006] Next, quality control monitoring is performed by a method for obtaining one-way delay time and jitter, and in order to point out the location when quality is deteriorated, the packet probe devices are comprehensively arranged inside the monitored network. A method of grasping the quality deterioration section from the measurement result between the probe devices can be considered. According to this method, a distance between two pairs of probe devices is defined as a measurement section, a one-way delay time and a jitter are calculated within the measurement section, and whether or not the quality is deteriorated is determined based on the result. If quality degradation has occurred, the measurement section is reported as the quality degradation location. By performing this process on all the packet probe devices, it is possible to determine the quality deterioration point.

[0007]

As described above, accurate time synchronization and frequency synchronization are required to measure the one-way delay time. However, in recent years, when the bandwidth of the network has been widened and there is no queuing delay, a minimum delay time of several μsec / km is provided. For this reason, the time synchronization for obtaining the one-way delay time is required to have an accuracy on the order of microseconds or less. A method of using GPS is conceivable to achieve such accuracy, but problems such as limitation of installation locations occur, and it is difficult to comprehensively dispose packet probes. A clock synchronization method using ISDN has been proposed, but this method cannot synchronize the time with an accuracy of ± 10 μs or less.

However, frequency synchronization can be realized if there is an accurate frequency source such as frequency synchronization using ISDN. Frequency synchronization can achieve an accuracy of several microseconds or less.
For example, 1500 by on a 1 Gbit / s network
Since the time required for the te packet to pass is 12 μsec, it can be measured with an accuracy of 10 ^−-6 sec. Therefore, it is conceivable to perform the quality monitoring by obtaining the jitter, which is a measurement method that can be realized only by frequency synchronization.

Here, it is assumed that in a network in which priority control is performed as a quality control method, a failure occurs in which a packet having a high priority and a packet having a low priority have the same priority. In this situation, if the traffic volume of the non-measurement packets is large, the test packet may always be positioned at the end of the queue. In this case, the delay time becomes long, but the delay variation may become small. For this reason,
In such a state, if the method using the distribution of jitter is applied, quality deterioration cannot be found. In order to avoid such a problem, in consideration of the fact that the packet loss probability is extremely high when the packet is always located behind the queue, the packet loss factor is included in the quality monitoring method. You need to choose the method that is included.

Next, a problem will be described regarding a method of ascertaining a location where the quality is deteriorated. It is necessary to comprehensively arrange the packet probe devices in order to grasp the location where the quality is deteriorated after the failure is detected. At this time, it is necessary to receive the probe data strings from all the packet probe devices and perform the same packet search from the data strings at both ends in the set monitoring section. If the packet probe devices are arranged exhaustively, the increase of one packet probe device will increase the monitoring interval of more than one interval, and the number of tables for the same packet search required for jitter calculation will increase and the processing time will increase. Is an issue. Furthermore, when the delay variation is large, it is necessary to enlarge the table for searching the same packet, and the time required for searching the same packet may be very long.

[0011]

It is an object of the present invention to use only an accurate frequency source without using high precision time synchronization, to include packet loss information in addition to delay variation information, and to perform the same packet search. It is to provide a quality monitoring method for performing quality monitoring without performing it. Further, another object of the present invention is to provide a method for determining a quality deterioration point where the quality deterioration occurs by using end-end measurement without exhaustively arranging the packet probe device in the network. That is.

[0012]

FIG. 1 shows a basic configuration of a quality status monitoring system including a quality status monitoring device in a packet network according to the present invention. The quality time monitoring apparatus according to the present invention is composed of a quality monitoring apparatus 20 for monitoring the quality state of the packet network 1, a topology grasping apparatus 30, and a quality degradation point determination apparatus 40. As shown in FIG. 1, a plurality of packet probe devices 11 to 13 driven by a high-accuracy frequency source and having a time series generation function are installed at the end of the network 1 to be measured and monitored. A packet probe device probes a packet to perform end-to-end measurement. At this time, the time sequence generation function of each packet probe device obtains the packet transmission time sequence in the transmission side packet probe device 11, and the packet reception time sequence in each of the reception side packet probe devices 12 and 13.

A transmission time sequence and a reception side packet probe device 12 obtained by the transmission side packet probe device 11.
The packet reception time series obtained by 13 and 13 are supplied to the quality monitoring device 20. In the quality monitoring device 20, E
From the packet transmission time sequence and the packet reception time sequence obtained as a result of the nd-to-end measurement, a cross-correlation function is obtained using the synthesis function creating function 21 and the cross-correlation calculation function 22, and the obtained cross-correlation function is obtained. Based on this, the quality degradation determination function 23 determines whether or not there is quality degradation, and outputs the quality degradation determination result. As a result, quality monitoring is realized without performing the same packet search as in high precision clock synchronization. Further, when the difference in priority is observed as packet loss, the correlation is low, so that it is possible to measure quality deterioration, for example, in which the packet is always located behind the queue.

In the topology grasping device 30, the branch / merge short-circuit removing function 31 is used to remove the branch or merge of the route information from the packet transfer route information collected by the network management information or the like. next,
The logical topology creating function 32 removes redundant structures and creates a logical topology having a tree structure. here,
The root of the tree becomes the packet probe device on the transmitting side, and the leaves become the packet probe devices on all receiving sides. The direction indicates the packet transfer direction. Using this logical topology, when quality degradation has occurred, the location where quality degradation has occurred is determined.

In the quality degradation point determination device 40, the quality degradation determination result and the data series are received from the quality monitoring device 20, and the ethical topology is received from the topology grasping device 30. Figure out if you are. Next, the degraded portion estimation function 42 estimates the degraded portion more finely. Then, the outputs from the degradation point determination function 41 and the degradation point estimation function 42 can be visually displayed on the display device 50.

[0016]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS FIG. 2 shows, as an embodiment of the present invention, a processing flow of a packet network quality condition monitoring apparatus according to the present invention. The same components as those used in FIG. 1 will be described with the same reference numerals. In FIG. 2, a portion indicated by a thick solid line corresponds to the present invention. Below, these functions are demonstrated in detail. (Packet probe device) In the packet probe device, a clock is operated by using a highly accurate frequency source, and the progress of the clock is accurately synchronized. This device has a time sequence generation function. This function will be described below. "Time Sequence Generation Function" This function converts the passage time of a packet into a data series sampled at the sampling frequency f. At this time, the time at which the packet is probed is set to 1, and the other times are set to 0 to generate a data string. For example, when f = 10 Hz and the test packet is probed at 0.0 seconds and 0.3 seconds, 0.0 seconds = 1,0.
A data sequence is generated with 1 second = 0, 0.2 = 0, 0.3 second = 1. This data string is divided at time intervals t and continuously transmitted to the quality monitoring device 20. Using the above example, when t = 1 second, 10 samples are transmitted to the quality monitoring device 20 every other second.

The data series created from the transmission side time series is represented as x _s ^c _n, and the data series created from the reception side time series is represented as x ^cp _n . Here, c indicates the quality control class to which this data string belongs, and from 0 to 1, 2,
3, ..., C, and the priority increases from 1 to c, for example. Numbers 1, 2, 3, ..., N are sequentially given to the receiving side packet probe device from 1, and p indicates the receiving side packet probe device number. In the case of a network that is not classified, C has only one fixed value.

When the quality is dynamically evaluated using the test packet, the test packet is sent from one packet probe device to all other packet probe devices by using multicast or plural unicasts. To be The test packet is formed as a continuous constant pattern having a periodic characteristic with a constant interval ΔT. The packet probe device focuses on the test packet generated in this way and collects the transmission or reception time of the test packet.
In the case of passively evaluating the quality of traffic flowing in end-to-end, the packet probe device at the end point between transmission and reception probes only the traffic and collects the transmission time and the reception time. In addition, the periodicity of the packet transferred in advance is measured, and the value is set to ΔT. For example, when packets are generated at regular intervals, the interval is ΔT.

(Quality Monitoring Device) The quality monitoring device 20 for monitoring the quality of the network is realized by the synthesis function creating function 21, the cross-correlation calculating function 22, and the quality deterioration determining function 23. The operation contents will be described below according to the processing flows of these devices. In "Synthesis function creating function" present apparatus, the transmitting side and the receiving side data sequence described above x _s ^c _n and _{^{x cp n (0 ≦ c ≦}} C, 1 ≦
p ≦ N), a prepared function g,
For example,

[Outer 1] A new sequence h _s ^c _n , which is a composite function with a function such as
Get h ^cp _n . In fact, creating a series g _n obtained by sampling the function g at the frequency ^{_{f, x s c n, x}} cp n is a composite function of the g _n h _s ^c _n, the h ^cp _n, Fast Fourier Transform (FFT ) Is used to calculate.

"Cross-correlation calculation function"
Sequence h synthesized by the number creation function 21_s ^c _n , h^cp _n 
(0 ≤ c ≤ C, 1 ≤ p ≤ N) is received
Series h_s ^c _n And one receiving side data sequence h^cp _n Pair
Cross-correlation is obtained by FFT, and a new sequence r^cp _n To
obtain. Here, C × N calculations are performed. In addition, goods
If required by the quality deterioration determination function 23, h_s ^c _n of
The autocorrelation is obtained by FFT, and the new series r _s ^c _n Seeking
Meru. Note that traffic is being probed passively
At this time, if ΔT is not required, please use this function.
H_s ^c _n Is calculated to obtain ΔT. Where r_s ^c 
_n , r^cp _n Against

[Outside 2] The result of multiplying the exponential weight, such as a new r _s ^c
_n, it is also conceivable to r ^cp _n.

"Quality degradation determination function" In the present apparatus, the cross-correlation calculation function 22 causes all the data strings r _s ^c _n , r ^cp _n
(0 ≦ c ≦ C, 1 ≦ p ≦ N) is received, and the data in the range of 0 to ΔT or the lag at which the first peak appears is t ^cp, and the data in the range of t ^cp to t ^cp + ΔT. withdrawn, this new r _s ^c _n, and r ^cp _n. The number of data points is m = ΔT / f. In addition, using these, quality evaluation is performed by any of the following methods.・ The maximum value and the minimum value of r ^cp _n are ^calculated, and the absolute value of the difference is E
^cp . ^-The maximum value of r ^cp _n is obtained, and the value is set as E ^cp .・

[Outside 3] For E,

[Outside 4] F ^cp is obtained. Then normalize E ^cp by dividing E ^cp by F ^cp . In the range of 0 to ΔT, prepare a series w _{m in} which a function that becomes a bathtub curve is sampled at a frequency f,

[Outside 5] For E,

[Outside 6] F ^cp is obtained. Then normalize E ^cp by dividing E ^cp by F ^cp .

Alternatively, E is obtained as follows without extracting the range of the data series. &Μiδδoτ; r
The power spectrum of ^cp _n is obtained, and the spectral density of the frequency component having the period of ΔT is E ^cp . Note obtains the power spectrum of r _s ^c _n, normalizing the E ^cp spectral density of the frequency component as a period of △ T. Next, E ^cp is 1
When the threshold value close to is not exceeded, 1 ^cp = 0 is set, and when the threshold value is exceeded, 1 ^cp = 1 is set. When monitoring the priority error, it is confirmed that E ^0P <E ^1P <E ^2P <... <E ^NP in the packet probe number on the receiving side. If this relationship is maintained, 1 ^P = 1 for the quality deterioration determination result 1. This indicates that no quality deterioration has occurred between classes. When the above relationship is not maintained, 1 ^P = 0. This indicates that a priority error has occurred between classes. In the following, 1 ^cp in a particular class will also be referred to as 1 ^P.

Next, this operation is performed for all E's, and 1
_Let ^P and E ^cp or h ^cp _n or r ^cp _n be output values. Continuous monitoring is realized by continuously performing the above operations.

(Topology grasping device) The topology grasping device 30 includes a route information estimating function 33 and a route information collecting function 3.
4, a branch / merge short-circuit removal function 31 and a logical topology creation function 32. The route information estimating function 33 and the route information collecting function 34 can be realized by conventional techniques. For example, the route information estimation function 33 can be performed by observing the packet loss in the receiving side packet rope, or using the management information when managing the packet propagation route, or using the SNMP ( Si
This can be achieved by using a network management protocol such as mple Network Management Protocol. Therefore, the packet loss information is supplied from the packet probe device to the route information estimation function 33, and the route information estimated by the function 33 is used in the logical topology creation function 32.
Is supplied to. Here, the description will be continued assuming that the packet transfer route is obtained by these techniques. The packet transfer path is a loop-free path from one transmitter to one receiver.

The route information collecting function 34 collects the route information by using the management information or the network management protocol when managing the packet propagation route, and collects the collected route information from the branch / merge short-circuit removing function 31 and It is supplied to the logical topology creation function 32.

In the branch / merge short-circuit removal function 31 and the logical topology creation function 32, the packet transfer route obtained by the route information estimation function 33 or the route information collection function 34 is converted into a logical topology that can be used in the quality degradation point determination device 40. It has the function of converting. The logical topology has a tree structure and is composed of a node V, a link L, a root Vr, and a leaf Vl. A leaf Vl indicates a set of lines and routers or switches, and indicates a branch point of a route normally composed of routers or switches. The link L indicates a packet transfer path which is composed of a line and a router and has no branch. Route V
r indicates a packet probe on the transmitting side, and leaf Vl indicates a packet probe on the receiving side. These are connected according to the following rules. In Vr, one or more nodes V, one or more leaves V1, or one or more nodes V and one or more leaves V1 are connected by a link L. -Leaf V1 has two or more nodes V, or 2
More than one leaf V1 or more than one node V
And one or more leaves V1 are connected by a link L. The leaf V1 is connected by a link L to one node V or one root Vr.

The actual packet transfer path is set so that all routers and switches are unidirectionally connected, assuming that the path is a directed graph when viewed from one packet transmitter. There are no strongly connected components. However, there is a possibility that the structure will have a path that merges after branching. Therefore, in the present invention, in order to obtain a logical topology having a tree structure according to the above rules, as a first step, a branch / merge short-circuit removal function is used to find a path that branches and joins, and Short-cut and remove all links and nodes in the closed area to obtain a tree structure topology. When performing measurement using multicast, it is possible to directly output the multicast tree from the route information estimation function 33 or the route information collection function 34 to the logical topology creation function without using the branch / merge short-circuit removal function. Is. The presence of a strongly connected component (loop) is a problem of route control, and the present invention presupposes that the route control is normally performed and there is no strongly connected component.

"Branch / Merge Short-Circuit Removal Function" This apparatus short-circuits and removes all the links in the area closed by branching and merging. As one method of removing a short circuit, an embodiment of a branch / merging short circuit removal function by a method of color-coding links into two colors will be described below. The propagation token has one or more data tokens. The data token has data including a number of branches, a branch node number, a color, and a starting leaf number. The data token is the number of branches, branch node number,
It has a data stack area consisting of a set of color and initial leaf number. Also, the nodes on the packet transfer path have node numbers that do not overlap. At this time, tokens are played according to the following rules. -On one link, only one propagation token can pass in the direction opposite to the direction of the directed tree (route table). -When the propagation token arrives at the node, all data tokens are extracted from the propagation token. -When output from a node, all data tokens are wrapped in a propagation token. The initial attributes of the data token are branch number = none, branch node number = none, color = white, the stack is empty, and the leaf number at which the token was generated is given as the initial leaf number. The initial attribute of the link is white. -When the propagation token branches, that is, when it reaches a node having two or more inputs, the data token is extracted from the propagation token, and all data tokens are
The current attributes are stored in the stack, and the number of branches = the number of branch links, the color = red, and the branch node number = node number are given as new attributes. The initial leaf number is not changed. Next, all data tokens are wrapped in one propagation token, a copy of the propagation token is made, and the same propagation token is sent on all links. -If there are multiple branches from the leaf, create a data token with the same attributes as above, wrap the data token in one propagation token, make a copy of the propagation token, and create the same propagation token. Stream on all links. -In a link in which a propagation token including a data token having a red attribute has flowed, if the link attribute is white, change it to red. Link attributes never go from red to white. -When the tokens merge, that is, when they reach a node having two or more outputs, wait for the propagation token to arrive from all the links and perform the following operations. For the data tokens that have the same branch node number and the same initial leaf number and branch node number, if the number of data tokens is equal to the number of branches that is the attribute of the data token, set the data token attribute. The attributes retrieved from the stack are replaced and combined into a single token. Repeat this work as much as possible. If there is no data token to combine, all data tokens are wrapped in one propagation token and output.

Links can be color coded by passing tokens from all leaves according to the above rules. A logical topology having a tree structure can be obtained by removing the short circuit in the red part of the color-coded links to form one node.

FIG. 3 shows an operation example. In FIG. 3, a thick solid line indicates a link having a red attribute, a thin solid line indicates a link having a white attribute, and the arrow direction indicates the direction indicated by the route table. Circles are nodes, and the node numbers are shown in the circles. The squares are data tokens, (a,
b, c, d) indicates the attribute, and [(a, b, c, d)] indicates the attribute stored in the stack. Here, a indicates the number of branches, b indicates the branch node number, c indicates the color attribute, and d indicates the starting leaf number. In the stack, the attributes are stored so that they are packed in the top, and the attributes are fetched in order from the top.
The balloon indicates a propagation token.

In FIG. 3, the token is illustrated as issued from the starting leaf number 1. Initially, the attributes of all links are white, there is one data token in the propagation token, and the attributes are (-,-, white, 1).
This propagation token flows backward on the directed graph. Since it is branched into two at node 6, data token has an attribute of (2, 6, red, 1), and (-,-,
White, 1) attributes are stored in the stack. The attribute of the link turns red because the propagation token that has the data token, which is a red attribute, has passed. Next, in the node 5, since the branch is performed again, the data token is
It becomes an attribute of (2, 5, red, 1), and becomes (2, 6, red, 1)
Attributes are stored on the stack. Although merging is performed in the node 2, since there is no data token that can be combined, all data tokens are combined into one propagation token and passed upstream. Merging is performed at node 1. First, since two data tokens having the attributes (2, 5, red, 1) appear, they are combined as one data token and the attribute (2, 6, red) taken out from the stack is given. Furthermore, in node 1, (2, 6,
Since there are two data tokens that have the attribute of red, 1), these are also combined into one data token.
As a result, as at the end of the figure, the merging and branching points are expressed as red links, and by short-circuiting and removing this link, it is possible to change to two links and one node.

"Logical Topology Creation Function" In this device, based on the tree structure obtained by the branch / junction short-circuit removal function, modification is made so that the structure complies with the above-mentioned rules. In this device, a part having a single link input and a single link output and a node and a link is changed to a single link. FIG. 4 illustrates an actual change. In FIG. 4, circles indicate nodes, arrows indicate links, and squares indicate leaves. In addition, the short-circuited nodes are indicated by diagonal lines. A logical topology is created by the above-mentioned device and is output from the topology grasping device.

(Deterioration Point Determining Device) The quality degradation point determining device 40 receives I ^P from the quality monitoring device 20,
This is a function of estimating a quality deterioration point by the deterioration point determination function 41 and the deterioration point estimation function 42. In the following, description will be given according to the processing order of these devices. “Degraded part determination function” When the quality monitoring device 20 determines that the quality has deteriorated, the degraded part determination function 41 receives the quality monitoring result from the quality monitoring device 20 and receives the logical topology from the topology grasping device 30. The degradation point determination function performs processing using these results.

A basic determination method will be described with reference to FIG. The degradation point determination function receives I ^P from the quality monitoring device and determines the degradation point. When the topology information and the quality monitoring result shown on the left side of FIG. 5 are received, the route Vr
It can be said that there is little difference between the transmission and reception patterns on the path from the to the leaves V12 and V13, that is, there is no packet loss and there is little delay variation. However, the quality is degraded on the route to the leaf Vl1. Here, in the route to the leaf V11, a quality deterioration occurs at a position other than the route to the leaves 2 and 3, or at the output interface portion at the branch point of the route to the leaves 2 and 3 and the route to the leaf 1. it seems to do. Therefore, in the route to the leaf 1, a link not included in the route to the leaves 2 and 3 is marked, and further, to a node that becomes a branch point of the route from the route to the leaves 2 and 3 to the leaf 1. Mark it. The part marked with this mark is output as a part where there is a high possibility that quality deterioration has occurred. In FIG. 5, the link in which the quality deterioration has occurred is marked to make it clear. In this way, the degraded portion determination function 41 identifies the link portion in which the quality degradation occurs based on the quality degradation determination result supplied from the quality monitoring device and the logical topology supplied from the topology grasping device 30.

In the following, a detailed description will be given according to the structure of the logical topology. FIG. 6 exemplifies the topology information received from the topology grasping apparatus, the information regarding one class received from the quality monitoring apparatus, or the result of the priority error determination. In FIG. 6, the number inside the leaf indicates the leaf number, and the contents enclosed in parentheses indicate the attributes of the leaf. (O) indicates that the quality deterioration has not occurred, that is, I ^P = 1 and (X) indicates that the quality deterioration has occurred, that is, I ^P = 0. Circles indicate nodes, and numbers inside the nodes indicate node numbers.

First, at a node to which only leaves are connected, attributes are given according to the following rules. Figure 6
Then, nodes 7, 9, 10, 11, 13, 14, 15,
16, 17 are targeted.・ If all the leaf attributes are (○). Give the node a (○, ○) attribute.・ If all leaf attributes are (×). Give the node an attribute of (×, ×).・ If the leaf attributes are (○) and (×). Give the node the (×, ○) attribute.

The result is illustrated in FIG. In addition,
Regarding the notation of (A, B), A indicates the quality state at the node, which is called the own node state. B indicates the possibility that quality deterioration has occurred in the upper part from that node, and this is called the propagation state.

Next, attributes are given to the nodes having attributes and the nodes having only leaves as children as follows. In the figure, 6, 8 and 12 are targeted.・ When the propagation status of all child nodes is ○ and the child has leaves, and the attributes of all child leaves are ○,
(○, 〇). -When the propagation state of all the child nodes is x and the child has leaves, and the attributes of all the child leaves are x.
(X, x). -If the child has a leaf, including the leaf attribute, and if the propagation state of the child node is mixed with ○ and ×,
(X, o).

The attribute of the root Vr is the same as above. The result of repeating the above operation is illustrated in FIG.

When the own node state is x, it indicates that the quality deterioration can be observed in the lower node or leaf. This may be due to the possibility that the quality degradation has already occurred from the upper node and the possibility that it has occurred from the own node to the lower nodes, including the own node. However, if the propagation status of the own node is ○, it means that there is a part where the quality deterioration has not occurred in some of the lower nodes, and therefore the quality deterioration occurs in the section from the upper node to the own node. It is unlikely that
Therefore, it is considered highly likely that the quality degradation has occurred in the links from the self-node including the self-node to the child nodes whose propagation state is x or the leaf having the (x) attribute.

Based on the assumptions in the previous paragraph, marks are added according to the following rules, and the set of switches and routers indicated by the marked nodes and links is considered to have a possibility of quality deterioration. This is the output of the degradation point determination function. -Add a mark to the node that has a status of ×, including the root. -A mark is attached to a link from a marked node to a child node whose propagation status is x or to a leaf having an attribute of (x).

The results are shown in FIG. When the root attribute is (x, x), the child node or leaf viewed from the root is marked, and the link to the node or leaf is marked. In FIG. 9, node numbers 1, 3, 6, 7, 15,
Within the 14 subtrees, it cannot be determined in which section the quality degradation has occurred. Therefore, if it is necessary to analyze such a subtree, a location where quality degradation may occur is estimated by the degraded location estimation function 41.

When a node or leaf having the "degraded part estimation function" (x, x) attribute constitutes a subtree, all the subtrees are taken out. Hereinafter, the estimation method will be described for one of the extracted subtrees, but the degraded portion estimation function 42 performs the following processing for all the extracted subtrees. As a method for estimating the quality deterioration point, historical data of the data series E ^CP belonging to all leaves inside the subtree (time series data such as the past 24 hours and 164 hours),
A method of obtaining the correlation by using h ^CP _n or r ^CP _n and estimating the lowered portion can be considered. A brief description will be given with reference to FIG.

In FIG. 10, it can be seen that the same packet time series arrives at the leaves Vl2 and Vl3 from the upper and lower correlations. However, packets arrive at the leaf Vl1 in a pattern different from that of the leaves Vl2 and Vl3. It is considered that the correlation between the data series such as E ^CP , h ^CP _n , and r ^CP _n becomes high when there is no quality deterioration point in the route after passing the quality deterioration point.
Therefore, in the upper part of the figure, from the root Vr to the leaf 1
It is conceivable that the quality may have deteriorated on the route to the node 2 and on the route from the route Vr to the node 2. For the bottom of the figure, from root Vr to leaf Vl1
It is considered that the quality may have deteriorated on the route leading to. In FIG. 10, a mark is added to the estimated position thus obtained. The degraded portion estimation function outputs the portion to which the mark is added as the estimated portion. In the following, a detailed description will be given using a new tree structure shown in FIG.

In the tree, a subtree composed of leaves and nodes having only leaves as children is a group. FIG. 12 shows the result of grouping. In FIG. 12, a subtree surrounded by a dotted line is a group. In each group, the history data of the data series E ^CP corresponding to the leaf (time series data such as the past 24 hours and 164 hours) or h ^CP _n or r ^CP _n is used to determine all the leaves in the group. The strength of the correlation is obtained by the combination. The result of the calculation. Combinations with a high correlation are marked with a circle, and combinations with a low correlation are marked with a cross. Using this result, attributes are given to the nodes and links in the group according to the following rules. -When all combinations in the group are "○", the attribute of (○) is given to all the nodes and links in the group. -When all combinations in the group are "x", assign the attribute (x) to all nodes and links in the group.-When the combination attributes include "○" and "x". , A node is given an attribute of (x), two links leading to two leaves corresponding to a combination of "○" are given an attribute of (○), and other links are given an attribute of (x). Assign attributes.

The results are shown in FIG. As a method for reducing the number of correlation calculations, a method may be considered in which a polygon having leaves to be calculated as vertices is created, a full-mesh shape is drawn, and triangles are focused to determine the correlation.
For example, if there are five leaves consisting of a, b, c, d, and e, the vertices of the pentagon are a, b, c, d, and e, respectively.
In the festival that focuses on the triangles a, b, and c, if a-b is calculated as ◯ and bc is calculated as x, then a-c is calculated.
Becomes x. If cd is calculated as x, ac
Needs to be calculated, and when ae is calculated as o, be becomes o. These are summarized as follows.

When the second side is ◯, the third side is ◯.・ When one side is ○ and the other side is ×. The third side is x.・ If the 2nd side is ×, calculate the 3rd side.

Next, a representative data series group is selected and given to the node as follows. -If all combinations in the group are "○", the data corresponding to one leaf is arbitrarily used as the representative data series group. -If all the combinations in the group are "x", the data corresponding to all the leaves is set as the representative data series group. -When the combination of "○" and "x" is mixed, pay attention to the combination that becomes "○", select the combination with the highest correlation from the related (closed) combinations, and arbitrarily select one data. Select and exclude others. For example, in FIG. 13, the closed combination in the group to which the node 8 belongs is leaves V15, 6, 7, and, for example, the data of leaf V15 is included in the representative data series group. In the node 10, the leaf Vl10
And 11, 12 and 22 are independently closed combinations, so that, for example, the data of leaves 10 and 12 are included in the representative data series group. The data corresponding to all the leaves of the remaining combinations of “x” are also included in the representative data series group. The result is shown in FIG.

Next, a node having an attribute added to a child and a leaf that is not an operation target in the previous stage, and a node not having an attribute attached to itself is a processing target node. A subtree composed of these nodes and leaves is set as a new group. In the example of FIG. 14, the nodes 4 and 11 are the processing target nodes, and the group is surrounded by the dotted line.

In each group, the correlation is obtained for the representative data series group held by the child nodes of the processing target node. At this time, the representative data series group is collected in units belonging to the child nodes viewed from the processing target node, and this is called a partial representative data series group. If you have a child,
A data series corresponding to a leaf is one partial representative data series group.

The processing target node calculates the cross-correlation between the partial representative data series groups. Do not combine between data series in the partial representative data series group. For example, if the processing target node has child nodes A and B, the representative data series group of the node A is 1 and 2, and the representative data series group of the node B is 3 and 4, It is grouped into two sets of partial representative data series called 3 and 4. The combination is 1-
3, 1-4, 2-3, and 2-4.

As a result of the calculation of the correlation between the partial representative data series groups, if there is at least one combination "○", the correlation between the partial representative data series groups is (○). When all combinations are “x”, the correlation between the partial representative data series groups is (x). Attributes are given to nodes and links as follows using the obtained correlation results. -When all the correlations between the partial representative data series groups are (O), the attribute of "O" is given to the processing target node and link in the group. -When all the correlations between the partial representative data series groups are (x), the attribute of "x" is given to the nodes and links in the group. -When (○) and (x) are mixed in the correlation between the partial representative data series groups, the child node corresponding to the partial representative data series that becomes (○) and the link to the child node are marked with "○". An attribute is given, and an attribute of "x" is given to the remaining child nodes and links. Also, an attribute of "x" is given to the processing target node.

Next, a representative data series group is selected. The result is shown in FIG. -When the combination in the group is "○", one data from the combination with the highest correlation is arbitrarily used as the representative data series group. -If all combinations in the group are "x", all data are set as the representative data series group.・ When "○" and "×" are mixed in the combination,
Pay attention to the combination of "○", select the combination with the highest correlation from the related (closed) combinations, select one data arbitrarily, and exclude the others. For example, in FIG. 15, there are two types of closed combinations 2-5 and 4-9 between the partial representative data series 7 and 8. In this case, for example, 2 and 4 are set as representative data, and 5 and 9 are excluded. Data that is not excluded from the remaining combinations of "x" are also included in the representative data series group.

This result is illustrated in FIG. After that, a new processing target node is determined and the same calculation is performed. The above operation is repeated until there are no more nodes to be processed. Finally, mark the nodes and links with the (X) attribute.
An actual display example is shown in FIG.

[0055]

As described above, according to the present invention, it is possible to monitor the quality control state of the network in which the priority control is performed, and to grasp the control abnormal state such as the priority error. Further, when a quality abnormality occurs, it is possible to grasp the abnormal portion. It is possible to eliminate the need for time quality measurement such as one-way delay time at the time of quality measurement. As a result, the quality can be achieved by the measuring devices distributed at a plurality of locations realizing only frequency synchronization.

[Brief description of drawings]

FIG. 1 is a configuration diagram of a quality status monitoring device in a packet network.

FIG. 2 is an embodiment of a quality status monitoring device in a packet network.

FIG. 3 is an example of an implementation method for removing a branching / merging structure from the packet route information at a short end.

FIG. 4 is a modification example for creating a logical topology from a topology that has been shortened and removed.

FIG. 5 is an illustration of a basic grasping method for grasping a location where quality is deteriorated.

FIG. 6 is an example of output results of the quality monitoring device and the topology grasping device.

FIG. 7 is an illustration of a result of creating a node attribute from a leaf attribute.

FIG. 8 is an example of a result of creating attributes of a parent node from attributes of a child node.

FIG. 9 is an example of a portion where it is determined that quality deterioration has occurred.

FIG. 10 is an illustration of a simple estimation method based on correlation.

FIG. 11 is an example of a directed tree processed by a degraded portion estimation function.

FIG. 12 is an example of grouping nodes and leaves.

FIG. 13 is a view showing an example of the internal processing result of grouping.

FIG. 14 is an illustration of a selection result of a representative data series group.

FIG. 15 is an example of a processing result of the degraded portion estimation function.

FIG. 16 is an illustration of a selection result of a representative data series group.

FIG. 17 is a view showing an example of a final result of estimation of a quality deterioration portion.

[Explanation of symbols]

1 packet network 11,12 Packet probe device 20 Quality monitoring device 21 Composite Function Creation Function 22 Mutual function calculation function 23 Quality degradation judgment function 30 Topology grasping function 31 Split / merge short-circuit removal function 32 Logical topology creation function 33 Route information estimation function 34 Route information collection function 40 Quality degradation point determination device 41 Degradation point determination function 50 display

Claims (19)

[Claims] 1. A method for monitoring network quality by comparing a transmission state and a reception state of a monitored packet, and determining a quality degradation occurrence section when a quality degradation is detected, which is a quality measurement monitoring target. Multiple measuring devices are installed in the same network, and the quality is evaluated by calculating the cross-correlation function based on the packet transit time series obtained by probing the packets in each measuring device and evaluating the quality. If it is determined that the quality of the packet network, a logical topology having a tree structure is created in advance from the packet transfer route information, and the location where the quality deterioration occurs is determined using the logical topology. Monitoring method. 2. A transmission side data sequence representing a packet transmission time sequence and a reception side data representing a packet reception time sequence output from a plurality of packet probe devices arranged in a packet network to be monitored and having frequency sources synchronized with each other. A method of monitoring the quality of a packet network using a sequence, and when it is found that the packet network has deteriorated in quality, a method of monitoring the quality state in a packet network for determining a quality degradation part, wherein the transmission side data sequence and Receives the data sequence on the receiving side, obtains the cross-correlation function using the data sequence on the transmitting side and the data sequence on the receiving side, and determines from the obtained cross-correlation function whether or not quality degradation has occurred, and determines the quality of the packet network. The process of outputting the evaluation result and the packet transfer route information of the packet network are received, and a tree structure is formed from the received packet transfer route information. A packet comprising a step of creating a logical topology, and a step of determining a quality deterioration part based on the quality evaluation result output from the quality monitoring device and the logical topology output from the topology grasping device. Network quality condition monitoring device. 3. The quality monitoring method for a packet network according to claim 1, wherein a test packet is generated to dynamically monitor the network quality. 4. A method of using a multicast as a method of generating a test packet in the quality status monitoring method in a packet network according to claim 3. 5. The test state generation method in the packet network quality control method according to claim 3, wherein:
How to use multiple unicast. 6. The quality monitoring method for a packet network according to claim 1, wherein after the cross-correlation function is obtained, the quality is evaluated by using the maximum correlation value. 7. The quality monitoring method for a packet network according to claim 1 or 2, wherein, after obtaining a cross-correlation function, a generation cycle of a packet or a group of packets is determined, and a time series in one cycle range is extracted, A method of performing quality evaluation using the maximum value of correlation values within the time series. 8. The quality monitoring method for a packet network according to claim 1 or 2, wherein, after obtaining a cross-correlation function, a generation cycle of a packet or a group of packets is determined, and a time series in one cycle range is extracted, A method of obtaining the maximum value and the minimum value of the correlation value in the time series, and using the difference between them to evaluate the quality. 9. The quality monitoring method for a packet network according to claim 1 or 2, wherein an autocorrelation function of a packet transmission time series is obtained after obtaining a cross-correlation function, and an occurrence cycle of a packet or a packet group is determined. The time series in one period range is taken out, the integrated value of the cross-correlation function and the integrated value of the autocorrelation function in the time series are obtained, and the integrated value of the cross-correlation function is divided by the integrated value of the autocorrelation function. How to evaluate quality. 10. The quality monitoring method for a packet network according to claim 1, wherein a generation cycle of a packet or a group of packets is determined after obtaining a cross-correlation function, and a time series within one cycle range is extracted and extracted. Exponentially weighting the time series by the vertical axis method,
A method of obtaining the maximum value and the minimum value of the correlation value in the time series, and using the difference between them to evaluate the quality. 11. The quality monitoring method for a packet network according to claim 1, wherein after obtaining a cross-correlation function, an auto-correlation function of a packet transmission time series is obtained.
The generation cycle of a packet or a group of packets is determined, a time series within one cycle range is extracted, and the extracted time series is exponentially weighted by the vertical axis method, and the integrated value of the cross-correlation function within the time series is calculated. A method of evaluating the quality by obtaining the integral value of the autocorrelation function and using the value obtained by dividing the integral value of the cross-correlation function by the integral value of the autocorrelation function. 12. The method for monitoring the quality state in a packet network according to claim 1 or 2, when determining the location where the quality deterioration occurs, by propagating a token with an attribute on a logical topology, branching or merging occurs. A method of obtaining a tree structure by removing short-circuiting a closed area. 13. The quality monitoring method for a packet network according to claim 1 or 2, wherein a location where quality degradation has occurred is determined from a measurement result of end / end, and when a location where the quality degradation is observed from the measurement result. , A method of obtaining a link and a node showing quality deterioration by removing a node and a link which are considered not to have quality deterioration from a path judged to have a quality deterioration by using a structure of a logical topology. 14. The quality monitoring method for a packet network according to claim 13, wherein when a quality degradation point is determined as a subtree, end-end is performed between leaves.
A method for further finely estimating the location where the quality is degraded by obtaining the correlation between the measurement results and using the result and the subtree structure. 15. The quality monitoring method for a packet network according to any one of claims 1 to 14, wherein the packet network is classified into transfer priorities with a packet transfer priority as a transfer class, and classified by class. A method characterized in that the network is a network in which QoS control is performed to perform priority transfer control. 16. The method according to any one of claims 1 to 15.
A method of monitoring a network, which comprises monitoring a priority error of a network transferred by priority using the method of monitoring the quality state in a packet network according to the item. 17. A transmission side data sequence representing a packet transmission time sequence and a reception side data representing a packet reception time sequence output from a plurality of packet probe devices arranged in a packet network to be monitored and having frequency sources synchronized with each other. A packet network quality state monitoring device that monitors the quality of a packet network using a sequence and determines that a quality degradation has occurred in the packet network, and is a packet network quality state monitoring device that determines a quality degradation part, Receives the data sequence on the receiving side, obtains the cross-correlation function using the data sequence on the transmitting side and the data sequence on the receiving side, and determines from the obtained cross-correlation function whether or not quality degradation has occurred, and determines the quality of the packet network. A quality monitoring device that outputs the evaluation result and packet transfer route information of the packet network are received, and a tree structure is received from the received packet transfer route information. A topology grasping device that creates a certain logical topology, and a quality deteriorated portion determination device that determines a quality deteriorated portion based on the quality evaluation result output from the quality monitoring device and the logical topology output from the topology grasping device. A packet network quality condition monitoring device. 18. The packet network quality condition monitoring apparatus according to claim 17, wherein the quality monitoring apparatus provides a transmission-side combining function and a reception-side combining function for each of the transmission-side data series and the reception-side data series. A composite function generating means for generating, a cross-correlation calculating means for calculating the cross-correlation between the transmitting-side combining function and the receiving-side combining function, and determining whether or not the quality deterioration has occurred from the obtained cross-correlation And a quality deterioration determining means for controlling the quality. 19. A packet network quality condition monitoring apparatus according to claim 17, wherein said topology grasping apparatus receives packet loss information from said packet probe apparatus and estimates packet route information. A route information collecting unit that collects packet route information from the network management information; a branch merge short-circuit removing unit that removes a branch and / or a merge from the route information output from the route information collecting unit; An apparatus for creating a logical topology based on the output from the information estimating means and the output from the branch / junction short-circuit removing means.