JP2009194628A - Link band extension amount calculating method and system, and program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To efficiently calculate a facility extension amount which is not apt to be neither too much nor too small without getting puzzled with an abrupt rise in link use rate. <P>SOLUTION: Use rates of links among nodes 3a to 3c are collected from the nodes every day at predetermined time intervals and stored in a storage device S601, link use rates of a link to be processed within a predetermined number of days are read out of the storage device to generate statistic information (probability density distribution) of respective link use rates at each collecting time point from the number of link use rates collected at the collecting time S602. The generated statistic information is used to find the probability that a link use rate larger than a threshold predetermined for congestion occurrence detection is obtained S603, it is decided S604 whether the found probability value exceeds a predetermined congestion occurrence rate threshold S604, and when it is decided that the predetermined congestion occurrence rate threshold is exceeded, a minimum link band increase amount with which the probability value found at S603 is less than the congestion occurrence rate threshold is calculated S605. <P>COPYRIGHT: (C)2009,JPO&INPIT

Description

  The present invention relates to a facility management technique for a network composed of a plurality of nodes and a plurality of links, and more particularly to a technique suitable for avoiding an excessive increase in facilities.

  As a technique for efficiently managing the facilities of a network composed of a plurality of nodes and a plurality of links, for example, the techniques described in Non-Patent Documents 1 and 2, or as shown in FIGS. There is a technique for calculating the amount of additional bandwidth of a link based on the usage rate of a certain link.

  FIG. 7 is a block diagram showing the configuration of a conventional system for calculating the amount of expansion of the link bandwidth, and FIG. 8 is a block diagram showing an example of the internal configuration of the facility expansion amount calculation function in FIG.

  The technologies shown in FIGS. 7 and 8 observe the usage rate of the link, compare the peak value with a predetermined threshold value, and if the peak value of the link usage rate exceeds the threshold value, future demand will be increased. The amount of equipment based on the forecast will be increased.

  FIG. 7 shows the configuration of a system that calculates the amount of additional equipment for a network formed by connecting nodes 73a to 73c such as computer terminals and routers. The link usage rate observation / collection function 71 The traffic information including the link usage rate is collected using MIB (Management Information Base) implemented in 73a to 73c, and the collected link usage rate peak value information is input to the facility expansion amount calculation function 72 To do.

  As information input to the facility expansion amount calculation function 72, in addition to the peak value of the link usage rate from the link usage rate observation / collection function 71, information on the link bandwidth input from the network operator, link usage rate, There is a threshold for the bandwidth used. Information regarding the link bandwidth may be input from the link usage rate observation / collection function 71.

  As shown in FIG. 8, the facility expansion amount calculation function 72 includes an insufficient bandwidth calculation unit 72a, a link usage rate comparison unit 72b, and an additional bandwidth determination unit 72c as programmed computer processing functions.

  The link usage rate comparison unit 72b compares the peak value of the link usage rate input from the link usage rate observation / collection function 71 with a threshold value determined by the operator.

  If the peak value of the link usage rate is higher than the threshold value as a result of the comparison, the link usage rate comparison unit 72b calculates the missing ratio and outputs it to the insufficient bandwidth calculation unit 72a. For example, when the peak value of the link usage rate is 70% and the threshold is 60%, 10%, which is the insufficient ratio (70-60), is input to the insufficient bandwidth calculation unit 72a.

  The insufficient bandwidth calculation unit 72a integrates the insufficient ratio input from the link usage rate comparison unit 72b and the link bandwidth previously input from the operator or the link usage rate observation / collection function 71 and stored in the storage device. Then, the insufficient bandwidth is calculated.

  For example, when the link bandwidth is 100 Mbps and the ratio of lack is 10%, the lack bandwidth is 10 Mbps (100 × 10 ÷ 100). Then, the insufficient bandwidth calculation unit 72a inputs the insufficient bandwidth calculated in this way to the additional bandwidth determination unit 72c.

  Based on the insufficient bandwidth input from the insufficient bandwidth calculation unit 72a, the expansion bandwidth determination unit 72c calculates the expansion amount according to the policy conditions determined in advance by the operator stored in the storage device. For example, when a policy condition for adding 1.5 times the insufficient bandwidth is defined, if the insufficient bandwidth is 10 Mbps, the additional bandwidth is output as 15 Mbps.

  The output of the extension band determining unit 72c becomes the output of the facility extension amount calculation function 72, and based on the extension amount calculated by the facility extension amount calculation function 72, the extension of the facility in the network is examined.

  However, in such a conventional technique, when a sudden traffic inflow occurs in the network composed of the nodes 73a to 73c, the link usage rate observation / collection function 71 collects and inputs to the facility expansion amount calculation function 72. The peak value of the link usage rate becomes excessively high, and the expansion amount calculated by the facility expansion amount calculation function 72 is excessively high.

  As a result, there is a possibility that the number of facilities will be excessively increased, and the network operation cost will increase unnecessarily.

  As for the link usage rate, a technique for comparing the measured link usage rate with a threshold value to determine whether or not the link is a bottleneck is disclosed in Patent Document 1, and a link at a predetermined time interval. Patent Document 2 discloses a technique for calculating an average usage rate and determining a link congestion state if the calculated link average usage rate is equal to or greater than a predetermined value.

JP 2006-191219 A JP 2002-252649 A Z. -L, Zhang et al. "On scalable design of bandwidth brokers," IEICE Trans. Commun. vol. E84-B, no. 8, Aug. 2001. Katsushi Yamaoka, "Tokyo Institute of Technology Challenge 2006 Award for Efficient Internet Video Communication-Study on Coexistence and Accommodation of Streams on the Internet", [online], Tokyo Tech Chronicle, July. 2007, [Search January 16, 2008], Internet <URL: http: www. titech. ac. jp / publiccatons / j / channel / 423 / 423-1. html>

  The problem to be solved is that the conventional technology calculates the amount of link bandwidth expansion based on the comparison between the peak value of the link usage rate and the threshold, and there is a sudden inflow of traffic on the link. When this occurs, the peak value of the link usage rate used for comparison with the threshold value becomes excessively high, and the amount of expansion to be calculated becomes excessively high.

  An object of the present invention is to solve these problems of the prior art and to avoid an unnecessary increase in the operation cost of the network by avoiding an excessive installation of equipment to the network.

  In order to achieve the above object, the present invention is characterized in that, based on the statistical information of the traffic volume in the link, the probability of occurrence of congestion in the link is calculated, and an appropriate amount of additional network equipment is obtained.

  According to the present invention, statistical information is generated from the collected link usage rate and bandwidth, and the probability of occurrence of congestion is calculated using this statistical information, which is confused by an unexpected increase in link usage rate. Since the amount of equipment necessary to keep the congestion occurrence rate below a certain value is also clarified, it is possible to calculate an efficient equipment addition amount that is unlikely to cause excess or deficiency.

  The best mode for carrying out the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a configuration example of a link bandwidth extension amount calculation system according to the present invention, FIG. 2 is a block diagram showing a configuration example of a facility extension amount calculation function in FIG. 1, and FIG. FIG. 4 is an explanatory diagram showing an example of the configuration of information stored in the link usage rate database in FIG. 2 and an example of the configuration of information generated by the statistical processing unit in FIG. 2, and FIG. 4 is an explanation showing an example of the operation of the congestion occurrence rate calculating unit in FIG. FIG. 5 and FIG. 5 are explanatory diagrams showing an operation example of the extension amount calculation unit in FIG.

  In FIG. 1, the link bandwidth extension amount calculation system of this example is composed of a link usage rate observation / collection function 1 comprising the conventional technology and a facility extension amount calculation function 2 that characterizes the present invention. Function 1 and facility expansion amount calculation function 2 are such that a CPU (Central Processing Unit) accesses a program or data stored in an external storage device such as an HDD via the main memory and executes a computer process using the program. It is a device to do.

  As described above, the link bandwidth extension amount calculation system of the present example including the link usage rate observation / collection function 1 and the facility extension amount calculation function 2 is a facility for a network formed by connecting nodes 3a to 3c such as computer terminals and routers. The amount of expansion is calculated, and the link usage rate observation / collection function 1 collects traffic information including the link usage rate for each link from the MIB installed in each of the nodes 3a to 3c. The rate information is input to the facility expansion amount calculation function 2. The link usage rate is a value obtained by dividing the bandwidth actually used in the link by the link bandwidth in the link.

  As information input to the facility expansion amount calculation function 2, in addition to the link usage rate information from the link usage rate observation / collection function 1, information related to the link bandwidth respectively input from the network operator input device according to an operation instruction or the like And a congestion occurrence rate threshold and a link usage rate threshold. Information on the link bandwidth may be input from the link usage rate observation / collection function 1.

  As shown in FIG. 2, the facility expansion amount calculation function 2 includes, as programmed computer processing functions, a link usage rate DB 2a, a statistical processing unit 2b, a congestion occurrence rate calculation unit 2c, a congestion occurrence rate comparison unit 2d, and an addition amount calculation. Part 2e.

  In the link usage rate DB 2a, the link usage rate for each link input from the link usage rate observation / collection function 1 is determined for each link, as shown in FIG. Table information having contents divided for each time zone is generated and stored in the storage device.

  In the example of the table shown in FIG. 3A, the link usage rate in a certain link is registered every day from January 1, every 0:00, 6:00, 12:00, and 18:00. .

  For example, for January 1, the link usage rate at 0 o'clock is 57%, the link usage rate at 6 o'clock is 20%, the link usage rate at 12 o'clock is 46%, and the link usage rate at 18 o'clock Is 52%.

  The statistical processing unit 2b uses the link usage rate information held in the link usage rate DB 2a to create a probability density distribution of each link usage rate in a certain link as shown in FIG. 3B. .

  That is, the statistical processing unit 2b obtains information on the recent several days from the link usage rate DB 2a, generates probability density distribution data for each link usage rate for each time, and inputs the data to the congestion rate calculation unit 2c. To do.

  In the example shown in FIG. 3B, the probability density generated at each time is shown for all the link usage rates in a certain link shown in FIG. For example, regarding the link usage rate “57%” at 1 o'clock on January 1 in FIG. 3A, as shown in FIG. 3B, 7% at 0 o'clock, 0% at 6 o'clock, 12:00 Then, the probability density is 5% and 9% at 18:00.

  In FIG. 3B, the sum of the values (%) in each time column is 100%. For example, at 6 o'clock, the probability density of each link usage rate of 54% to 57% is 1%. And the total of 3% of 0% is only 3%, the probability density of the link usage rate not shown is 20% and 30% is as high as 10% and 15%, and the usage rate of the link is higher than 0:00. Low.

  The congestion occurrence rate calculation unit 2c generates a probability distribution function based on the probability density distribution of the link usage rate generated and input by the statistical processing unit 2b, and calculates the congestion occurrence rate for each time. The calculation procedure will be described with reference to FIG.

  In FIG. 4, on the horizontal axis, the link usage rate values (..., 54%, 55%, 56% at arbitrary times in the “probability density distribution of link usage rates for each time” shown in FIG. , 57%,...), And the result of plotting the probability density at the link usage rate is shown as a graph on the vertical axis.

  In the graph shown in FIG. 4, for example, if it corresponds to the data at 0 o'clock in the table in FIG. 3B, the distribution is a normal distribution centering on a probability density of 12% with a link usage rate of 55%.

  Thus, the graph of FIG. 4 shows an example in which a certain link usage rate in the network has a normal distribution. In this case, for example, when a certain link band in a certain network is given, the congestion occurrence rate in the link can be calculated by calculating the distribution function. On the contrary, it is also possible to obtain the link bandwidth necessary for keeping the congestion occurrence rate below a certain value.

  In other words, the congestion occurrence rate calculation unit 2c exceeds the link usage rate threshold value (link usage rate value at which congestion is considered to be exceeded) determined and set in advance by an operator or the like for each of all links. The link usage rate is specified, the occurrence probability of each specified link usage rate (area of the hatched portion of the graph shown in FIG. 4) is obtained, and this is calculated as the congestion occurrence rate.

  Then, in the congestion occurrence rate comparison unit 2d, the calculation result in the congestion occurrence rate calculation unit 2c is compared with a threshold value (congestion occurrence rate threshold value) determined in advance by an operator or the like, and the congestion occurrence rate calculation unit 2c When the calculation result (probability value / congestion rate) exceeds the congestion rate threshold, the extension amount calculation unit 2e is instructed to calculate the extension amount.

  The expansion amount calculation unit 2e calculates the minimum link bandwidth value and the amount to be increased so as not to exceed the congestion occurrence rate threshold.

  As a specific example, for example, when the “congestion occurrence rate threshold” is 65% and the congestion occurrence rate calculated by the congestion occurrence rate calculation unit 2c is 70% exceeding this 65 (%), the processing is as follows: It becomes as follows.

  The “link usage rate threshold” shown in FIG. 4 is “p% of link capacity (0 ≦ p ≦ 100)”, and the link usage rate value (“congestion”) that is considered to cause congestion as a network operation policy. If the current threshold capacity (link bandwidth) is 200 Mbps, the upper limit of the link bandwidth that can be used at present is 130 Mbps (= 200 Mbps × 65%). It becomes. When the current congestion occurrence rate calculated by the congestion occurrence rate calculation unit 2c is 70%, a link bandwidth of 140 Mbps (= 200 Mbps × 70%) is used.

  The link bandwidth value x for satisfying the 65% congestion threshold value for the use of the 140 Mbps link bandwidth can be calculated as approximately “216 Mbps” by the expression “x = 140 × 100 ÷ 65%”.

  Then, the expansion amount calculation unit 2e subtracts the current link bandwidth value (200 Mbps) from the link bandwidth value (216 Mbps) thus obtained, and outputs the value (16 Mbps) as the bandwidth value to be added. To do. This output becomes the output of the facility expansion amount calculation function 2. In this case, in FIG. 4, the position of the threshold of the link usage rate is the same, and the shape of the probability density distribution is reduced to the left side in the figure.

  Note that, based on the output from the facility expansion amount calculation function 2 (“band value to be expanded”), the expansion band determination processing unit (not shown) determines the expansion amount according to a policy (condition) predetermined by the operator. At that time, the expansion band determination processing unit outputs the largest “additional expansion” which is output as the result of each processing executed by the facility expansion amount calculation function 2 for each time (0, 6, 6 and 12 o'clock). Based on the “bandwidth value”, the expansion amount is determined.

  FIG. 5 shows the relationship between the congestion occurrence rate of a certain link in the network and the average value and standard deviation of the link usage rate. The larger the average value of the link usage rate, the more the link usage rate. It is shown that the greater the standard deviation, the higher the congestion occurrence rate.

  Then, by periodically updating the statistical information, that is, the generation of the data (probability density distribution of the link usage rate for each time) shown in FIG. 3B by the statistical processing unit 2b, and the congestion occurrence rate calculating unit 2c By periodically calculating the congestion occurrence rate described with reference to FIG. 4, it is possible to establish an extension guideline for the future.

  For example, in FIG. 5, as shown as “statistical information of one month ago”, “current statistical information”, and “predicted value of next month”, “statistical information of one month ago” to “current From the amount of change to "statistical information", it is possible to predict that "predicted value for next month" will exceed "congestion rate is constant", so that "predicted value for next month" is "congestion rate is constant" An extension guideline can be set in advance so as not to exceed.

  FIG. 6 is a flowchart showing a processing operation example of the link bandwidth extension amount calculation method according to the present invention by the facility extension amount calculation function in FIG.

  First, the link usage rate DB 2a predetermines the link usage rate between each node from the nodes 3a to 3c such as the network computer terminal and router via the link usage rate observation / collection function 1 in FIG. A first procedure of collecting and storing in the storage device at a specified time interval is executed (step S601).

  Next, the statistical processing unit 2b reads the link usage rate for a predetermined number of days of the link to be processed from the storage device, and calculates the link usage rate at the collection time from the number of collected link usage rates at each collection time. The second procedure for generating the statistical information (probability density distribution) is executed (step S602).

  Furthermore, a link usage rate larger than a threshold (link usage rate threshold) predetermined for congestion occurrence detection by using the statistical information generated by the statistical processing unit 2b in the second procedure by the congestion occurrence rate calculation unit 2c. A third procedure for obtaining the probability of obtaining is performed (step S603).

  Then, the congestion occurrence rate comparison unit 2d executes a fourth procedure for determining whether or not the probability value obtained by the congestion occurrence rate calculation unit 2c in the third procedure exceeds a predetermined congestion occurrence rate threshold. (Step S604) When the congestion occurrence rate comparison unit 2d determines that the probability value has exceeded the congestion occurrence rate threshold in the fourth procedure, the extension value calculation unit 2e obtains the probability value obtained in the third procedure as the congestion occurrence rate. A fifth procedure for obtaining the minimum link bandwidth expansion amount lower than the threshold is executed (step S605).

  In this way, in this example, statistical information is generated from the collected link usage rate, and based on this statistical information, the probability of occurrence of congestion in the link is calculated, and an appropriate amount of network equipment is obtained. The amount of equipment necessary to keep the congestion occurrence rate below a certain value will not be confused by the sudden rise in link usage rate, and the amount of equipment necessary to keep the excess and deficiency will be calculated. It becomes possible to do.

  That is, according to this example, the problem of the prior art (the amount of link bandwidth expansion is calculated based on the comparison between the peak value of the link utilization rate and the threshold value, and a sudden inflow of traffic on the link) If this occurs, the peak value of the link usage rate used for comparison with the threshold value will be excessively high, and the amount of expansion to be calculated will be excessively high). Thus, it is possible to prevent an unnecessary increase in the operation cost of the network.

  In addition, this invention is not limited to the example demonstrated using FIGS. 1-6, In the range which does not deviate from the summary, various changes are possible. For example, in this example, “link usage rate” is collected and statistical information is generated, congestion occurrence probability is calculated, additional bandwidth is calculated, etc., but instead of “link usage rate”, The “link bandwidth” may be used.

  In this example, “probability density distribution” is used as the statistical information, but “average value, standard deviation or variance”, “frequency”, and the like may be used. In other words, if it is known that “normal distribution is shown” as the trend of link usage rate or link usage bandwidth, the shape of probability density distribution can be specified from the average value and variance (or standard deviation). The amount of link bandwidth expansion can be calculated by the procedure described above. Further, even if it is not a normal distribution, if it is known that some distribution is shown, there may be a parameter that determines the shape of the distribution.

  Further, in this example, the data in FIG. 3A is collected at each time of day (every 6 hours at 0:00, 6 o'clock, 12 o'clock, and 18 o'clock), but it is 2 hours or 4 hours. It may be collected every time, or an average value of data collected in a certain time zone may be used instead of every time.

  Further, regarding the computer configuration of this example, a computer configuration without a keyboard or an optical disk drive device may be used, and an optical disk, an FD (Flexible Disk), or the like may be used as a recording medium for recording a program. As for the program installation, the program may be downloaded and installed via a network via a communication device.

It is a block diagram which shows the structural example of the link bandwidth extension amount calculation system which concerns on this invention. It is a block diagram which shows the structural example of the installation amount calculation function in FIG. FIG. 3 is an explanatory diagram illustrating a configuration example of information stored in a link usage rate database in FIG. 2 and a configuration example of information generated by a statistical processing unit in FIG. 2. It is explanatory drawing which shows the operation example of the congestion incidence rate calculation part in FIG. It is explanatory drawing which shows the operation example of the additional amount calculation part in FIG. It is a flowchart which shows the processing operation example of the link bandwidth extension amount calculation method based on this invention by the facility extension amount calculation function in FIG. It is a block diagram which shows the structure of the system which calculates the addition amount of the conventional link band. It is a block diagram which shows the internal structural example of the installation amount calculation function in FIG.

Explanation of symbols

  1: Link usage rate observation / collection function, 2: Facility expansion amount calculation function, 2a: Link usage rate DB, 2b: Statistical processing unit, 2c: Congestion rate calculation unit, 2d: Congestion rate comparison unit, 2e: Expansion Amount calculation unit, 71: link usage rate observation / collection function, 72: facility expansion amount calculation function, 72a: insufficient bandwidth calculation unit, 72b: link usage rate comparison unit, 72c: expansion bandwidth determination unit.

Claims (5)

A first procedure for collecting a link usage rate between nodes from a node constituting a network every day and at a predetermined time interval and storing the same in a storage device;
A link usage rate for a predetermined number of days of the link to be processed is read from the storage device, and statistical information of each link usage rate at the collection time is generated from the number of collected link usage rates at each collection time. And the steps
Using the statistical information generated in the second procedure, a third procedure for obtaining a probability that a link usage rate larger than a threshold value of link usage rate predetermined for detecting congestion occurrence is obtained;
A fourth procedure for determining whether or not the probability value obtained in the third procedure exceeds a predetermined congestion occurrence rate threshold;
If it is determined in the fourth procedure that the probability value has exceeded the congestion occurrence rate threshold,
And a fifth procedure for calculating a minimum link bandwidth increase amount at which the probability value obtained in the third procedure is lower than the congestion occurrence rate threshold. A first procedure for collecting and storing in a storage device the bandwidth used for links between nodes from nodes constituting the network every day and at predetermined time intervals;
A link usage bandwidth for a predetermined number of days of the link to be processed is read from the storage device, and statistical information of the link usage bandwidth at the collection time is generated from the number of collections of the link usage bandwidth at each collection time. And the steps
Using the statistical information generated in the second procedure, a third procedure for obtaining a probability that a link use band larger than a threshold of the link use band set in advance for detecting the occurrence of congestion is obtained;
A fourth procedure for determining whether or not the probability value obtained in the third procedure exceeds a predetermined congestion occurrence rate threshold;
If it is determined in the fourth procedure that the probability value has exceeded the congestion occurrence rate threshold,
And a fifth procedure for calculating a minimum link bandwidth increase amount at which the probability value obtained in the third procedure is lower than the congestion occurrence rate threshold. A link bandwidth extension amount calculation method according to claim 1 or claim 2,
The statistical information generated in the second procedure is
A link bandwidth extension amount calculation method, characterized in that the average value and standard deviation or variance, or probability density distribution or frequency are used.   The program for making a computer perform each procedure in the link bandwidth extension amount calculation method in any one of Claims 1-3. A system for calculating an increase in the bandwidth of a link between nodes constituting a network,
A link bandwidth extension amount calculation system comprising means for executing each procedure in the link bandwidth extension amount calculation method according to any one of claims 1 to 3 as programmed computer processing execution means.

Images