JP4530285B2 - Network accommodation design system, method for determining concentration ratio thereof, program, and storage medium therefor - Google Patents

Description

  The present invention relates to a network accommodation design system that determines the number of users that can be accommodated in a network in consideration of the communication quality of users, or designs a network in accordance with the number of users, a method for determining the concentration ratio thereof, a program, and a storage medium thereof .

  When designing the bandwidth of a concentrator line or router that aggregates a plurality of user lines into one, if the required bandwidth is obtained by simply adding the contracted bandwidth of each user line, the network resources are wasted due to excessive quality. Therefore, the accommodation design needs to be performed in consideration of the actual traffic of each user. Patent Document 1 discloses a technique for determining the number of user lines accommodated in a router using the average traffic value of each user and the concentrated line speed as input parameters.

On the other hand, as a technique for designing the bandwidth of a line concentrator or a router, a technique is known in which a design band is determined by multiplying a band obtained by simply adding contract bandwidths of each user by a predetermined concentration ratio ρ. Here, in order to satisfy the communication quality of the user, it is necessary to set the concentration rate so that the peak traffic value does not exceed the design bandwidth. However, since the peak traffic value of the concentrated line depends not only on the contracted bandwidth of each user but also on the average value, peak value, and variance of each user's traffic value, the concentrated line rate cannot be determined quantitatively. For this reason, in the prior art, the concentration rate is set based on an empirical rule, and the accommodation design is performed based on this concentration rate.
JP 2003-318972 A

  The technique of Patent Document 1 described above is not suitable for the accommodation design of a network that guarantees QoS because the communication quality cannot be specifically specified in the accommodation design.

  Conventionally, the concentration rate is determined based on an empirical rule, and the empirical rule cannot set the concentration rate accurately, resulting in excessive quality and waste of network resources or lack of network resources. Thus, there is a technical problem that the predetermined communication quality cannot be satisfied.

  Furthermore, although the concentration rate has regional unevenness and temporal or temporal unevenness, these cannot be accurately set based on empirical rules.

  The object of the present invention is to solve the above-mentioned problems of the prior art, and by analyzing the actual traffic of each user and statistically obtaining the concentration rate, enabling an accurate accommodation design regardless of the uneven distribution rate of the concentration rate. Another object of the present invention is to provide a network accommodation design system, a method for determining the concentration ratio thereof, a program, and a storage medium thereof.

  In order to achieve the above object, the present invention is characterized in that a network accommodation design system for designing accommodation of a concentrated line that aggregates a plurality of user lines and a method for determining the concentrated ratio thereof include the following means: There is.

  (1) The network accommodation design system of the present invention includes means for inputting a contracted bandwidth (ci) and an average traffic value (mi) for each user line, means for designating communication quality required for the concentrator line, and Concentration rate calculating means for calculating the concentration rate (ρ) of the concentrated line based on the contracted bandwidth, average traffic value and communication quality, and bandwidth determining means for determining the design bandwidth of the concentrated line based on the concentration rate ρ. It is characterized by including.

  (2) The concentration ratio determining method of the present invention includes a procedure for inputting a contracted bandwidth (ci) and an average traffic value (mi) for each user line, a procedure for specifying communication quality required for the concentration line, A procedure for estimating the standard deviation (σ) of the traffic on the concentrated line based on the contracted bandwidth (Ci) and average traffic value (mi) of the user line, and the sum of the standard deviation (σ) and the average traffic value of each user line And a procedure for estimating the peak traffic value (Z) of the concentrated line based on the quality parameter representative of the communication quality, and the sum of the peak traffic value (Z) of the concentrated line and the contracted bandwidth of each user line (ΣCi) And a procedure for obtaining the concentration ratio ρ based on the above.

  According to the above feature (1), it is possible to input a contracted bandwidth (ci) and an average traffic value (mi) of a plurality of user lines, and further specify a desired communication quality to satisfy the desired communication quality. Since the line concentration rate ρ is obtained, an appropriate accommodation design reflecting the actual traffic value of each user is possible.

  In addition, since the concentration rate of the concentration line can be determined based on the actual traffic value, it is possible to design an appropriate accommodation even if each user's actual traffic value is unevenly distributed locally or temporally or temporally. become.

  According to the above feature (2), the concentrator line that satisfies the desired communication quality can be obtained simply by inputting the contracted bandwidth (ci) and the average traffic value (mi) of each user line and specifying the desired communication quality. The concentration ratio ρ of can be statistically determined.

  DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the best embodiment of the present invention will be described in detail with reference to the drawings. FIG. 1 is a block diagram of a network accommodation design system according to the present invention, in which user lines ULi (UL1, UL2,... ULn) assigned to each user are aggregated into a concentrated line AL by an access switch 1 and are distributed over a wide area such as the Internet. Connected to network 2. The traffic monitoring unit 3 periodically monitors the traffic of each user line ULi and registers the monitoring result in the traffic database (DB) 4. In this embodiment, the average traffic value mi (m1, m2,... Mn) of each user line ULi is registered as a monitoring result.

  User information for managing the contract bandwidth ci of each user line ULi with a user ID is registered in the user DB 5. In the facility DB 6, information related to network facilities such as the user line ULi and the concentrator line AL is registered. In the accommodation status DB 7, the accommodation status of users in the network is registered.

  The network accommodation design unit 8 includes a concentration rate calculation unit 81 that calculates the concentration rate ρ of the concentration line AL, and an accommodation determination unit 82 that performs an accommodation determination based on the calculation result of the concentration rate ρ. The display unit 9 and the input operation unit 10 is connected. The network accommodation design unit 8 can be realized by mounting a program for executing arithmetic processing to be described later on a general-purpose computer having an MPU, ROM, RAM, interface, and an internal bus for interconnecting them. From the input operation unit 10, a line capacity excess rate α allowed for the concentrated line AL or a standard deviation coefficient k described later is designated as the communication quality required for the concentrated line AL.

  FIG. 2 is a diagram schematically showing the accommodation design method by the network accommodation design unit 8, and the thickness of each user line ULi and the concentration line AL represents the bandwidth capacity.

  In the present embodiment, the concentrator calculation unit 81 registers the contracted bandwidth ci (c1, c2... Cn) of each user line ULi (UL1, UL2... ULn) registered in the user DB 5 and the traffic DB 4. The average of the concentrated line AL using the average traffic value mi (m1, m2... Mn) of each user line ULi and the communication quality (line capacity excess rate α or standard deviation coefficient k) input from the input operation unit 10 as input parameters. The traffic value M, the peak traffic value Z, the variance V and the standard deviation σ are calculated. Based on these calculation results, the concentration rate ρ of the concentration line AL is finally obtained and output to the display unit 9.

  Further, in the accommodation determination unit 82, based on the calculation result of the capacity (allocated bandwidth) of the concentrated line AL registered in the facility DB 6, the current accommodated state registered in the accommodated state DB 7, and the concentration ratio ρ. In addition, the bandwidth allocated to the concentrator line AL is reviewed, and whether or not the new user can be additionally accommodated in the concentrator line AL based on the information regarding the contract bandwidth of the new user input from the input operation unit 10 or the like. Is determined. This determination result is also output to the display unit 9.

  Next, the calculation method and the accommodation design method of the concentration ratio ρ by the network accommodation design unit 8 will be described more specifically.

  The concentration rate ρ of the concentration line AL is considered to be the ratio of the actual traffic usage to the bandwidth C obtained by simply adding the contracted bandwidth ci of each user. If the peak traffic value is Z and the number of users is n, the concentration rate ρ can be obtained by the following equation (1).

  The peak traffic value Z of the concentrated line AL is unknown, but if the average traffic value of the concentrated line AL is M, the standard deviation of traffic in the concentrated line AL is σ, and the standard deviation coefficient is k, the peak traffic value Z is It is obtained by the following equation (2).

  Although the standard deviation σ of traffic on the concentrated line AL is unknown, this standard deviation σ is obtained as the square root of the variance V of the concentrated line. However, the variance V of the traffic value in the concentrated line AL dynamically changes depending on the usage form of the user, the application used, and the time zone. Therefore, in this embodiment, assuming the worst value in which the traffic value of each user line ULi fluctuates to the maximum within the contracted bandwidth ci, the variance vi of each user line ULi is obtained, and the distribution V of the concentration line AL is calculated as the sum Σvi. Is calculated.

  That is, the traffic value variance vi of each user line Ui is expressed by the following equation (3), where mi is the average traffic value of each user line Ui, xp is the traffic value at a predetermined sampling timing p, and P is the number of samplings. Is done.

  Here, since the traffic value xp is in the range of 0 ≦ xp ≦ ci, assuming the worst state in which the traffic value xp changes maximum for each sampling within the range, (xp−mi ) Repeats “mi” when xp is the minimum value “0” and “(ci−mi)” when xp is the maximum value “ci”, and mi, (ci− mi), mi, (ci-mi) ... Therefore, if the number of samplings P is infinite, the above equation (3) can be rewritten into the following equations (4) and (5).

  Assuming that the traffic behavior of each user line Ui is independent of each other, the variance V of the concentrated line AL is obtained as the sum of the variance vi of each user line Ui as in the following equation (6).

  The average traffic value M of the concentrated line AL is expressed by the following equation (7) as the sum of the average traffic values of each user line Ui, and the above equation (2) is expressed by the following equation (8): (1) is rewritten as the following equation (9). That is, the concentration rate ρ of the concentration line is obtained using only the contracted bandwidth ci and the average traffic value mi of each user line Ui as input parameters.

  In this embodiment, the concentration rate ρ of the concentration line AL is obtained over a period of 24 hours, for example, every 5 minutes, and the concentration value ρ of the concentration line AL is represented by the worst value among them.

  In the above formulas (2), (8), and (9), the standard deviation coefficient k is a parameter that serves as an index of network quality. Increasing this value decreases the line capacity excess rate α and improves network quality. However, since the line collection rate ρ increases, the communication resource saving rate decreases. On the other hand, if the standard deviation coefficient k is reduced, the line capacity excess rate α is increased and the network quality is lowered, but the concentrating rate ρ is reduced, so that the communication resource saving rate is increased.

  Thus, since there is a quantitative relationship between the line capacity excess rate α and the standard deviation coefficient k, for example, if the Chebyshev inequality is applied to the traffic value of the concentrated line AL, an example is shown in FIG. As described above, the probability α of exceeding the line capacity (Z> M + kσ) can be less than 1 / k 2. Therefore, by setting the standard deviation coefficient k to “10”, the probability of excess line capacity α can be suppressed to less than 1%.

  The contracted bandwidths ci (c1, c2 ... cn) of each user line ULi (UL1, UL2 ... ULn) are the same, the traffic behavior of each user line Ui is independent of each other, and the number of accommodated users is If there are 10 or more people, it can be assumed that the traffic is normally distributed. Therefore, assuming that the traffic distribution of the concentrated line AL can be assumed to be a normal distribution as shown in FIG. 4, the standard deviation coefficient k is set to “2” by applying the standard normal distribution function. The probability of excess capacity α can be suppressed to less than 2.3%, and the probability of excess line capacity α can be suppressed to less than 0.13% by setting the standard deviation coefficient k to “3”.

  In this embodiment, as shown in the above equation (9), the concentration rate ρ is obtained as a function of the standard deviation coefficient k, so that the concentration rate corresponding to the desired line capacity excess rate α, in other words, the concentration concentration corresponding to the desired network quality. The rate ρ can be easily obtained.

  In the above description, the standard deviation coefficient k is specified as an index of network quality. However, since the relationship between the standard deviation coefficient k and the channel capacity excess rate α is known as described above, the network accommodation design unit 8 For example, the excess line capacity ratio α may be designated, and the network capacity design unit 8 may convert the excess capacity ratio α into the standard deviation coefficient k and apply it to the above equation (9).

  FIG. 5 is a diagram showing a list of contract bandwidth c, average traffic value m, estimated maximum variance v, and estimated standard deviation σ for 32 users. FIG. 6 shows a standard deviation coefficient k based on FIG. Is a list of the average traffic value M, the estimated maximum variance V, the estimated standard deviation σ, the peak traffic value, the line usage rate, and the concentration rate ρ of the concentrated line AL calculated as “3”.

  In this embodiment, the average traffic value M of the concentrated line AL is “1005”, the estimated standard deviation σ is “249.249”, and the standard deviation coefficient k is “3”. For example, “1752.747” is obtained as the peak traffic value Z of the concentrated line AL. Since the sum of the contracted bandwidths of the user lines is 3200 [Mbps], if these are substituted into the above equation (1), 0.548 (= 1752.747 / 3200) is obtained as the concentration ratio ρ (= Z / Σci). .

  Therefore, if 54.8% of the total 3200 [Mbps] of the contracted bandwidth ci of each user is allocated as the design bandwidth C of the line concentration line AL, high-quality communication is provided to all users without wasting communication resources. Can be guaranteed. Then, if the standard normal distribution function is applied assuming that the traffic distribution of the concentrated line AL can be assumed to be a normal distribution, it can be predicted that the probability of exceeding the line capacity is suppressed to less than 0.13% (k = 3).

  Furthermore, even when there is a request for registration of a new user, if the usage status of each user is regarded as being equivalent to that of an existing user, the peak traffic value Z can be calculated based on the contract bandwidth of each new user. It is possible to quantitatively determine whether or not it can be accommodated, and if it cannot be accommodated, an increase in the line capacity can be easily obtained.

  Furthermore, according to the present embodiment, it is possible to optimize the bandwidth allocated to the concentrator line AL that aggregates the user line ULi simply by monitoring the traffic of the user line ULi. In other words, even if the traffic value is unevenly distributed locally or temporally, it is possible to easily optimize the bandwidth allocated to each concentrated line AL.

It is a block diagram of a network accommodation design system according to the present invention. It is the figure which showed typically the accommodation design method by a network accommodation design part. It is the figure which showed the method of applying the Chebyshev inequality to the traffic distribution and estimating the probability of excess line capacity. It is the figure which showed the method of applying the standard normal distribution function to traffic distribution, and estimating the probability of channel capacity excess. FIG. 6 is a diagram showing a list of examples of contract bandwidth c, average traffic value m, estimated maximum variance v, and estimated standard deviation σ. Based on FIG. 5, the average traffic value M, the estimated maximum variance V, the estimated standard deviation σ, the peak traffic value Z, the line usage rate, and the concentration rate ρ of the concentrated line AL calculated with the standard deviation coefficient k as “3” are obtained. It is the figure displayed as a list.

Claims (9)

In a network accommodation design system that performs accommodation design of concentrator lines that aggregate multiple user lines,
Means for entering the contracted bandwidth (ci) and average traffic value (mi) of each user line;
Means for designating communication quality required for the line concentrator;
Concentration rate calculating means for calculating the concentration rate (ρ) of the concentration line based on the contracted bandwidth, average traffic value and communication quality;
And a bandwidth determination means for determining a bandwidth of the concentration line based on the concentration rate ρ. The concentration rate calculating means includes:
Means for estimating a standard deviation (σ) of traffic on a concentrated line based on a contracted bandwidth and an average traffic value of each user line;
Means for estimating a peak traffic value (Z) of a concentrated line based on the standard deviation (σ), a sum of average traffic values of each user line and a quality parameter representative of the communication quality;
2. The network accommodation design system according to claim 1, further comprising means for obtaining a concentration rate based on a peak traffic value of the concentration line and a sum (ΣCi) of contract bandwidths of each user line.   3. The network accommodation design system according to claim 2, further comprising means for setting the quality parameter based on a Chebyshev inequality, wherein the communication quality is a line capacity excess rate of a concentrator line.   3. The network accommodation design system according to claim 2, further comprising means for setting the quality parameter based on a normal distribution approximation of a traffic distribution, wherein the communication quality is a line capacity excess rate of a concentrated line. In a concentration ratio determination method for determining a concentration ratio of a network accommodation design system that performs accommodation design of a concentration line that aggregates a plurality of user lines,
Enter the contracted bandwidth (ci) and average traffic value (mi) for each user line;
A procedure for specifying the communication quality required for the line concentrator;
A procedure for estimating the standard deviation (σ) of the traffic on the concentrated line based on the contracted bandwidth (Ci) and the average traffic value (mi) of each user line,
A step of estimating a peak traffic value (Z) of a concentrated line based on the standard deviation (σ), a sum of average traffic values of each user line and a quality parameter representative of the communication quality;
A method for determining a concentration ratio of a network accommodation design system, comprising: a step of obtaining a concentration ratio ρ based on a peak traffic value (Z) of the concentration line and a sum of contracted bandwidths (ΣCi) of each user line.   6. The method according to claim 5, further comprising a step of setting the quality parameter based on Chebyshev's inequality, wherein the communication quality is a line capacity excess rate of the line concentrator. .   6. The network concentration design system according to claim 5, further comprising a step of setting the quality parameter based on a normal distribution approximation of a traffic distribution, wherein the communication quality is a line capacity excess rate of the concentration line. Rate determination method. In the concentration rate determination program for determining the concentration rate of the network accommodation design system that performs the accommodation design of the concentration line that aggregates multiple user lines,
On the computer,
Enter the contracted bandwidth (ci) and average traffic value (mi) for each user line;
A procedure for specifying the communication quality required for the line concentrator;
A procedure for estimating the standard deviation (σ) of the traffic on the concentrated line based on the contracted bandwidth (Ci) and the average traffic value (mi) of each user line,
A step of estimating a peak traffic value (Z) of a concentrated line based on the standard deviation (σ), a sum of average traffic values of each user line and a quality parameter representative of the communication quality;
A procedure for obtaining the concentration rate ρ based on the peak traffic value (Z) of the concentration line and the sum of the contracted bandwidth of each user line (ΣCi),
A program for running In the concentration rate determination program for determining the concentration rate of the network accommodation design system that performs the accommodation design of the concentration line that aggregates multiple user lines,
On the computer,
Enter the contracted bandwidth (ci) and average traffic value (mi) for each user line;
A procedure for specifying the communication quality required for the line concentrator;
A procedure for estimating the standard deviation (σ) of the traffic on the concentrated line based on the contracted bandwidth (Ci) and the average traffic value (mi) of each user line,
A step of estimating a peak traffic value (Z) of a concentrated line based on the standard deviation (σ), a sum of average traffic values of each user line and a quality parameter representative of the communication quality;
A procedure for obtaining the concentration rate ρ based on the peak traffic value (Z) of the concentration line and the sum of the contracted bandwidth of each user line (ΣCi),
The computer-readable recording medium which recorded the program for performing this.

Images