Control device and program

Abstract

A control device 1 comprises: a monitoring data processing unit 16 that stores traffic of a plurality of user terminals 3 and calculates an average value and a standard deviation of each of (1) a bandwidth used by high-priority traffic, (2) the number of simultaneously connected users of best-effort communication, and (3) a bandwidth used by users of best-effort communication in a predetermined period in a physical interface that performs quality of service (QoS) control using a two rate three color marker (trTCM) in units of traffic of the user terminal 3; and a setting value calculation unit 17 that calculates a CIR and a PIR used for the trTCM in the physical interface. The setting value calculation unit 17 calculates a weight of the standard deviation of (1) the bandwidth used by high-priority traffic from the degree of influence of each standard deviation in the physical interface, and calculates the CIR and the PIR, which have a negative correlation with the average value of (1) the bandwidth used by high-priority traffic and have a negative correlation with the product of the standard deviation of (1) the bandwidth used by high-priority traffic and the weight of the standard deviation.

Description

Control device and program

[0001] The present disclosure relates to a control device and a program.

[0002] Conventionally, there is a technology that achieves fairness in communication among users in best-effort traffic at the physical interface (IF) of an edge router. Here, fairness in communication among users means providing equal communication capacity to each user.

[0003] For example, according to the bandwidth used by each user's traffic, packets of the traffic are labeled with Green, Yellow, and Red using the two-rate Three Color Marker (trTCM). There is a technology for controlling Quality of Service (QoS) based on the priority corresponding to the color of the packet (see Non-Patent Documents 1 and 2).

[0004] In the trTCM, the traffic packets of each user are labeled as Green, Yellow, or Red using the Committed Information Rate (CIR) and Peak Information Rate (PIR) of the traffic of each user. The values ​​of CIR and PIR are assumed to be set by an operations manager or the like.

[0005] When controlling the communication bandwidth of a best-effort communication line so as to be fair among users, there is a technology that avoids limiting the communication bandwidth for users with low communication volume as much as possible and prevents a decrease in the utilization efficiency of the communication line (see Patent Document 1). Patent Document 1 applies a communication volume limit to flows that consume a large amount of bandwidth, thereby avoiding limiting the communication bandwidth for users with low bandwidth consumption.

[0006] Japanese Patent Publication No. 2020-28087

[0007] RFC2859 "A Time Sliding Window Three Color Marker (TSWTCM)," [online], June 2000, [Retrieved June 20, 2024], Internet <URL: https: / / datatracker.ietf.org / doc / html / rfc2859> Hongjun Su et al., "ItswTCM: A New Aggregate Marker to Improve Fairness in DiffServ," [online], November 2001, GLOBECOM'01. IEEE Global Telecommunications Conference (Cat. No.01CH37270), [Retrieved June 20, 2024], Internet <URL: https: / / ieeexplore.ieee.org / stamp / stamp.jsp?tp=&arnumber=965893>

[0008] However, the method described in Patent Document 1 increases the free space of the physical IF, which may result in ineffective utilization of resources.

[0009] For example, if the utilization rate of a physical IF in an edge router is close to 100%, the flow bandwidth observed at a certain device may be an observed value resulting from pressure caused by other flows accommodated on that physical IF, or it may be an observed value resulting from the conditions of the upstream network leading up to the device in question.

[0010] The method described in Patent Document 1 is effective when pressure is placed on a flow by other flows that share a single physical IF. On the other hand, when the method described in Patent Document 1 is affected by an upstream network, the bandwidth used does not change even if the capacity of the physical IF is freed up, which may result in an increase in the free space on the physical IF.

[0011] The present disclosure has been made in consideration of the above circumstances, and an object of the present disclosure is to provide a technology that can achieve both fairness among users and effective use of resources in best-effort communication.

[0012] A control device according to one embodiment of the present disclosure includes a monitoring data processing unit that calculates the average value and standard deviation of (1) the bandwidth used by high-priority traffic, (2) the number of simultaneously connected users for best-effort communication, and (3) the bandwidth used by users for best-effort communication over a predetermined period in a physical interface that accommodates traffic from multiple user terminals and performs QoS (Quality of Service) control using trTCM (two rate three color marker) on a traffic unit basis for the user terminals, and a setting value calculation unit that calculates the CIR (Committed Information Rate) and PIR (Peak Information Rate) used in the trTCM in the physical interface. The setting value calculation unit calculates, based on the influence of each standard deviation in the physical interface, (1) a weight for the standard deviation of the bandwidth used by high-priority traffic, and calculates a CIR and PIR that are negatively correlated with the average value of the bandwidth used by high-priority traffic and (1) negatively correlated with the multiplication of the standard deviation of the bandwidth used by high-priority traffic and the weight of the standard deviation.

[0013] A program according to one embodiment of the present disclosure causes a computer to function as a monitoring data processing unit that calculates the average value and standard deviation of (1) the bandwidth used by high-priority traffic, (2) the number of simultaneously connected users for best-effort communication, and (3) the bandwidth used by users for best-effort communication over a predetermined period in a physical interface that accommodates traffic from multiple users and performs QoS control using trTCM (two rate three color marker) on a per-user traffic basis, and as a setting value calculation unit that calculates the CIR (Committed Information Rate) and PIR (Peak Information Rate) used in the trTCM in the physical interface, and the setting value calculation unit (1) calculates a weight for the standard deviation of the bandwidth used by high-priority traffic based on the influence of each standard deviation in the physical interface, and calculates a CIR and PIR that (1) have a negative correlation with the average value of the bandwidth used by high-priority traffic and (1) have a negative correlation with the multiplication of the standard deviation of the bandwidth used by high-priority traffic and the weight of the standard deviation.

[0014] According to the present disclosure, it is possible to provide a technology that can achieve both fairness among users and effective use of resources in best-effort communication.

[0015] FIG. 1 is a diagram illustrating the system configuration of a communication system according to the present disclosure. FIG. 2 is a diagram illustrating the physical IF of an edge router. FIG. 3 is a diagram illustrating a bandwidth available for BE traffic. FIG. 4 is a diagram illustrating parameters assigned to one color in a trTCM. FIG. 5 is a diagram illustrating monitoring data. FIG. 6 is a diagram illustrating functional blocks of a control device. FIG. 7 is a diagram illustrating the effect of a calculation method according to the present disclosure. FIG. 8 is a flowchart illustrating processing in the control device. FIG. 9 is a sequence diagram illustrating processing in a communication system. FIG. 10 is a diagram illustrating experimental results of setting values ​​calculated using the calculation method according to the present disclosure. FIG. 11 is a diagram illustrating the hardware configuration of a computer used in the control device.

[0016] Hereinafter, embodiments of the present disclosure will be described with reference to the drawings. In the description of the drawings, the same parts are designated by the same reference numerals and the description thereof will be omitted.

[0017] 1 includes a control device 1, an edge router 2, a user terminal 3, and a monitoring device 4. The communication system 9 only needs to include a plurality of user terminals 3, and the number of user terminals 3 is not important.

[0018] The user terminal 3 is the end point of communication in the core network 5. In this disclosure, the user terminal may be simply referred to as a user.

[0019] The edge router 2 accommodates traffic from multiple user terminals 3. The edge router 2 forwards communications in the core network 5 to each of the multiple user terminals 3. The edge router 2 has multiple physical IFs. One physical IF has multiple VLANs (Virtual LANs). One physical IF accommodates multiple users by assigning them to each VLAN.

[0020] The physical interface of the edge router 2 performs QoS control using trTCM (two rate three color marker) for each traffic unit of the user terminal.

[0021] The monitoring device 4 records, for each hour, traffic information passing through each physical interface of the edge router 2. The monitoring device 4 records the traffic information for each hour in the monitoring data M.

[0022] The control device 1 refers to the monitoring data M and calculates the committed information rate (CIR) and peak information rate (PIR) used in the trTCM as the setting data T. The control device 1 calculates the CIR and PIR that can achieve both fairness among users and effective use of resources in best effort (BE) communication. The control device 1 outputs the calculated CIR and PIR to the edge router 2. The edge router 2 refers to the setting data T and sets the parameters of the physical IF.

[0023] The control device 1 calculates the CIR or PIR periodically or when a predetermined condition is met, allowing the edge router 2 to achieve both fairness among users and effective use of resources in best-effort communication, according to the current communication situation.

[0024] One physical IF of the edge router 2 will be described with reference to Figure 2. Focusing on the BE traffic of one physical IF, each packet of the BE traffic is labeled Green, Yellow, or Red by the trTCM on a VLAN basis. One queue for BE traffic is prepared for one physical IF. QoS (Quality of Service) control is performed in the BE traffic queue using Weighted Random Early Detection (WRED). When traffic becomes congested, WRED controls the discarding of packets in the order R → Y → G.

[0025] The network operator that operates the core network 5 wants to provide equal communication capacity to user terminals for BE traffic. Therefore, in order to perform QoS control in the queue for BE traffic in the subsequent stage in each VLAN, the edge router 2 assigns a label to each packet of traffic using the trTCM based on the value of the traffic bandwidth used by each VLAN.

[0026] The labels have three colors, for example, Red (R), Yellow (Y), and Green (G). Green (G) is assigned to packets whose bandwidth usage is less than the CIR, Yellow (Y) is assigned to packets whose bandwidth usage is equal to or greater than the CIR but less than the PIR, and Red (R) is assigned to packets whose bandwidth usage is equal to or greater than the PIR. Here, CIR and PIR are preset values.

[0027] For example, if CIR = 8Mbps and PIR = 16Mbps, when downloading the same 100MB file, the packets in the queue will be as follows:

[0028] If the download takes more than 100 seconds and is less than 8Mbps, the packets are all Green. If the download takes more than 8Mbps but less than 16Mbps, the packets are labeled Green or Yellow. If the download takes more than 16Mbps but less than 50 seconds, the packets are labeled Green, Yellow, or Red.

[0029] When traffic on a physical interface becomes congested, the edge router 2 discards packets in the order of Red (R) → Yellow (Y) → Green (G) according to the length of the queue.

[0030] In this disclosure, one physical IF is shared by high-priority QoS class communications and best-effort communications. As shown in Fig. 3, the bandwidth available for BE traffic is the bandwidth provided by one physical IF minus the bandwidth used by high-priority traffic.

[0031] As shown in Figure 4, three parameters are set for each color in the trTCM. Min Threshold is the threshold at which random dropping begins. Max Threshold is the threshold at which tail dropping is performed. Mark Probability (MP) is the probability of discarding a packet. These three parameters are set for each color.

[0032] When Red is set for a physical interface with the settings shown in Figure 4, packets begin to be dropped when the buffer fills up in the queue and the accumulation amount reaches 30% of the maximum queue length. Until the accumulation amount reaches 40% of the maximum queue length, the discard rate increases to 50% in proportion to the accumulation amount in the queue. When the accumulation amount exceeds 40% of the maximum queue length, 100% of the traffic is discarded.

[0033] The monitoring data M will be described with reference to FIG. 5 . The monitoring data M is time-series data regarding the utilization rate of the physical IF. In the present disclosure, the monitoring data M includes at least the following data for each time period: (1) the bandwidth used by high-priority traffic, (2) the number of simultaneously connected users of best-effort communication, and (3) the bandwidth used by users of best-effort communication. The monitoring data M may also include other data such as the number of simultaneously connected users, the number of simultaneously connected sessions, the ratio of high-priority traffic, and the ratio of BE traffic for each time period.

[0034] In the present disclosure, the control device 1 refers to the monitoring data M to calculate the CIR and PIR to avoid congestion. The monitoring data M may include only data when bandwidth is constrained. When bandwidth is constrained, the control device 1 may stop operation if the utilization rate of the physical IF is equal to or greater than a predetermined threshold, such as "physical IF utilization rate 70% or greater."

[0035] The setting data T includes the CIR and PIR calculated by the control device 1 with reference to the monitoring data M. The setting data T is output to the edge router 2. The edge router 2 updates the CIR and PIR of the physical IF with reference to the setting data T.

[0036] 6, the control device 1 includes various data such as statistical data 11, and the functions of a monitoring data processing unit 16 and a setting value calculation unit 17. Each piece of data is stored in a storage device such as a memory 902 or a storage 903. Each function is implemented in a CPU 901.

[0037] The statistical data 11 is data including statistical values ​​of the monitoring data M calculated by the monitoring data processing unit 16. The statistical data 11 includes at least the average value and standard deviation of (1) the bandwidth used by high-priority traffic, (2) the number of simultaneously connected users of best-effort communication, and (3) the bandwidth used by users of best-effort communication, for a predetermined period of time.

[0038] In the present disclosure, the "bandwidth used by high-priority traffic" may be replaced with (1) or "(1) bandwidth used by high-priority traffic." Similarly, the "number of simultaneously connected users of best-effort communication" may be replaced with (2) or "(2) number of simultaneously connected users of best-effort communication." The "bandwidth used by users of best-effort communication" may be replaced with (3) or "(3) bandwidth used by users of best-effort communication."

[0039] The monitoring data processing unit 16 refers to the monitoring data M and calculates statistical values ​​of the monitoring data M. The monitoring data processing unit 16 calculates (1) the average value and standard deviation of the bandwidth used by high-priority traffic, (2) the average value and standard deviation of the number of simultaneously connected users of best-effort communication, and (3) the average value and standard deviation of the bandwidth used by users of best-effort communication for the target period. The monitoring data processing unit 16 outputs statistical data 11 including the calculated average values ​​and standard deviations of each of (1) to (3).

[0040] The monitoring data processing unit 16 may calculate each value by referring only to data during bandwidth constraints in the monitoring data M. This allows the control device 1 to calculate CIR and PIR that can respond to changes in the number of simultaneously connected users and the occupancy rate of high-priority traffic during bandwidth constraints.

[0041] The set value calculation unit 17 refers to the statistical data 11 to calculate the CIR and PIR, and outputs the set data T.

[0042] The setting value calculation unit 17 first calculates the weight of the standard deviation of the bandwidth used by high-priority traffic (1) from the value of each standard deviation in the statistical data 11 and the influence of each standard deviation of (1)-(3) in the physical interface.

[0043] Specifically, the setting value calculation unit 17 calculates (1) the influence of the standard deviation of the bandwidth used by high-priority traffic, (2) the influence of the standard deviation of the number of simultaneously connected users of best-effort communication, and (3) the influence of the standard deviation of the bandwidth used by users of best-effort communication, for the target period in the monitoring data M. The setting value calculation unit 17 calculates the weight of (1) the standard deviation of the bandwidth used by high-priority traffic so that the sum of the influences of the standard deviations of (1) to (3) satisfies 1.

[0044] Here, we will explain how to calculate the weight of each standard deviation. Since the units of the standard deviations (1) to (3) are different, we will calculate the weight of the standard deviation in bps (bits per second) as the degree of influence on the physical IF capacity.

[0045] The method for calculating the influence of the standard deviation of (1) will be explained. The standard deviation of the bandwidth used by (1) high-priority traffic is a fluctuation range that directly affects the capacity of the physical IF. The influence of the standard deviation of the bandwidth used by (1) high-priority traffic is set to (1) the standard deviation of the bandwidth used by high-priority traffic.

[0046] A method for calculating the influence of the standard deviation of (2) will be described. Each user terminal uses, on average, a bandwidth equal to the average bandwidth used by users of (3) best-effort communication. The influence of the standard deviation of the number of users simultaneously connected to (2) best-effort communication is calculated by multiplying the standard deviation of the number of users simultaneously connected to (2) best-effort communication by the average bandwidth used by users of (3) best-effort communication.

[0047] A method for calculating the influence of the standard deviation of (3) will be explained. The number of simultaneously connected users is, on average, the average number of simultaneously connected users of (2) best-effort communication. The influence of the standard deviation of the bandwidth used by users of (3) best-effort communication is calculated by multiplying the standard deviation of the bandwidth used by users of (3) best-effort communication by the average number of simultaneously connected users of (2) best-effort communication.

[0048] The setting value calculation unit 17 calculates the influence of the standard deviation on the physical IF for each of (1)-(3), and then calculates the weight of the standard deviation of the bandwidth used by the high-priority traffic (1) so that the sum of the influences of each standard deviation satisfies 1.

[0049] (1) After calculating the weight of the standard deviation of the bandwidth used by the high-priority traffic, the set value calculation unit 17 calculates the CIR and PIR. The CIR and PIR each have a negative correlation with (1) the average value of the bandwidth used by the high-priority traffic, and also a negative correlation with (1) the product of the standard deviation of the bandwidth used by the high-priority traffic and the weight of the standard deviation.

[0050] More specifically, the set value calculation unit 17 calculates a first value by subtracting (1) the average bandwidth of the high-priority traffic and (1) the standard deviation of the bandwidth of the high-priority traffic multiplied by the weight from the capacity of the physical interface. The set value calculation unit 17 divides the first value by the total number of users accommodated in the physical interface to calculate the CIR.

[0051] The setting value calculation unit 17 calculates a second value by adding the average value of (2) the number of simultaneously connected users of best-effort communication to the value obtained by multiplying the standard deviation of (2) the number of simultaneously connected users of best-effort communication by the weight. The setting value calculation unit 17 calculates the PIR by dividing the first value by the second value.

[0052] In the present disclosure, the set value calculation unit 17 calculates the CIR and the PIR as shown in equation (1).

[0053]

[0054] Here, as a method for calculating the CIR and PIR of the trTCM, there is a method using only the average value, as shown in equation (2).

[0055]

[0056] However, if calculations are performed using only the average values, as in equation (2), the values ​​of (1)-(3) when based on CIR and PIR will actually fluctuate in a direction that compresses the bandwidth with a probability of approximately 50%.

[0057] Therefore, there is a method of calculating the CIR and PIR by taking into account the buffer and the standard deviation of the average value, as in equation (3).

[0058]

[0059] However, if the standard deviation is simply taken into account in the denominator as in equation (3), the labeling standard may be too conservative, resulting in a risk of not being able to effectively utilize the resources of the physics IF.

[0060] The effect of the calculation method according to the present disclosure will be described with reference to FIG. 7. In FIG. 7, regions R1-R3 respectively indicate the number of simultaneously connected users in best-effort communication and the range of occupied bandwidth for high-priority communication when CIR and PIR are calculated according to predetermined conditions and only packets labeled Red are discarded. Region R1 is calculated using only the average value as shown in formula (2). Region R2 is calculated using the method according to the present disclosure, using the average value, standard deviation, and its weight as shown in formula (1). Region R3 is calculated using the average value and standard deviation as shown in formula (3).

[0061] Generally, fairness is considered to be high when the total number of packets labeled Red is kept within a predetermined number and only packets labeled Red are discarded.

[0062] When using the CIR and PIR calculated by formula (2), if the number of simultaneously connected users for best-effort communication and the occupied bandwidth for high-priority communication for actual traffic are within the range of region 1, the physical interface will likely not discard packets, or if it does discard packets, it will only need to discard Red packets. However, if the CIR is low, Yellow packets may also be discarded, which may reduce fairness.

[0063] Region R3, when using the CIR and PIR calculated using Equation (3), is wider than region R1, when using the CIR and PIR calculated using Equation (2). Region R3 indicates that there is a high probability that no packet discards will occur, or that even if discards do occur, it is sufficient to discard Red. However, when the PIR is low, communication packets from low-traffic users tend to be marked Red and discarded. Therefore, if a small "standard deviation of traffic volume between users experiencing packet discards" is used as an indicator of fairness, the standard deviation will be calculated to be large, which tends to reduce fairness.

[0064] In contrast, region R2, which is intermediate between regions R1 and R3, is expected to be able to achieve both fairness among users and effective utilization of resources in best-effort communication compared to regions R1 and R3 by taking into account the weight of the standard deviation.

[0065] (Control Method) A control method in the control device 1 according to the present disclosure will be described with reference to FIG.

[0066] In step S11, the control device 1 waits for the timing to calculate the CIR and PIR. When the timing arrives, the control device 1 acquires the monitoring data M in step S12.

[0067] In step S13, the control device 1 calculates statistical data 11 from the monitoring data M. The statistical data 11 includes the average values ​​and standard deviations of (1)-(3).

[0068] In step S14, the control device 1 calculates a weight for the standard deviation for each of (1) to (3) from the statistical data 11. In step S15, the control device 1 calculates the CIR and PIR using the weight for the standard deviation calculated in step S14.

[0069] In step S16, the control device 1 generates setting data T including the CIR and PIR calculated in step S15. The control device 1 outputs the setting data T to the edge router 2.

[0070] A control method in a communication system 9 according to the present disclosure will be described with reference to FIG.

[0071] In step S101, the monitoring device 4 monitors information on the edge router 2. In step S102, the monitoring device 4 collects information on the edge router 2. In step S103, the monitoring device 4 stores the information collected in step S102 in monitoring data M. Here, the monitoring device 4 collects information (1) to (3) at least for each hour and stores it in the monitoring data M. The processing of steps S101 to S103 is performed periodically or when a predetermined condition is met.

[0072] In step S104, the monitoring data processing unit 16 of the control device 1 refers to the monitoring data M. In step S105, the monitoring data processing unit 16 collects information from the monitoring data M. The processing of steps S104-S105 is performed periodically or when a predetermined condition is satisfied. The processing of steps S104-S105 may also be performed asynchronously with the processing of steps S101-S103.

[0073] In step S106, the monitoring data processing unit 16 processes the monitoring data M collected in step S105 to generate statistical data 11. In step S107, the setting value calculation unit 17 of the control device 1 refers to the statistical data 11 generated in step S106, and in step S108 calculates setting values. The setting values ​​include CIR and PIR. The setting value calculation unit 17 generates setting data T including the calculated setting values.

[0074] In step S109, the setting value calculation unit 17 transmits the setting data T to the edge router 2. The edge router 2 uses the transmitted setting data T to reset the CIR and PIR of the trTCM.

[0075] 9 is repeatedly executed periodically or when a predetermined condition is satisfied. The edge router 2 can be operated using the CIR and PIR calculated according to the latest user usage status.

[0076] FIG. 10 shows experimental results for the CIR and PIR calculated by the control device according to the present disclosure. In FIG. 10, the calculation criteria for the CIR and PIR are indicated in the row direction. "Mean only" indicates calculation using only the mean value, as shown in the above formula (2). "Mean and unweighted standard deviation" indicates calculation using the mean value and standard deviation, as shown in the above formula (3). "Proposed method" indicates calculation using the mean value, standard deviation, and their weights, as shown in the above formula (1).

[0077] To achieve fairness, it is preferable that only Red packets are discarded. However, the "average value only" case is undesirable because Yellow packets are also discarded.

[0078] In contrast, in the "mean value and unweighted standard deviation" case and the "proposed method" case, only Red's packets are discarded, so fairness appears to be ensured. The standard deviation of the number of packets transmitted by users whose packets were discarded is smaller in the "proposed method" case than in the "mean value and unweighted standard deviation" case, which is thought to improve fairness.

[0079] The control device 1 according to the present disclosure can output a setting value that can achieve both fairness between users and effective utilization of resources in best-effort communication by considering the influence of the standard deviation on the physical IF as a weight for the standard deviation.

[0080] The control device 1 according to the present disclosure described above uses, for example, a general-purpose computer system including a CPU (Central Processing Unit, processor) 901, a memory 902, a storage 903 (HDD: Hard Disk Drive, SSD: Solid State Drive), a communication device 904, an input device 905, and an output device 906. In this computer system, the CPU 901 executes a program loaded on the memory 902, thereby realizing each function of the control device 1.

[0081] The control device 1 may be implemented by one computer or by multiple computers, and may also be a virtual machine implemented on a computer.

[0082] The program of the control device 1 can be stored in a computer-readable recording medium such as a HDD, SSD, USB (Universal Serial Bus) memory, CD (Compact Disc), DVD (Digital Versatile Disc), or can be distributed via a network. The computer-readable recording medium is, for example, a non-transitory recording medium.

[0083] The present disclosure is not limited to the above-described embodiments, and various modifications are possible within the scope of the present disclosure.

[0084] REFERENCE SIGNS LIST 1 control device 2 edge router 3 user terminal 4 monitoring device 5 core network 9 communication system 11 statistical data 16 monitoring data processing unit 17 setting value calculation unit 901 CPU 902 memory 903 storage 904 communication device 905 input device 906 output device M monitoring data T setting data

Claims

1. A control device comprising: a monitoring data processing unit that calculates the average value and standard deviation of (1) the bandwidth used by high-priority traffic, (2) the number of simultaneously connected users for best-effort communication, and (3) the bandwidth used by users for best-effort communication over a predetermined period of time in a physical interface that accommodates traffic from multiple user terminals and performs QoS (Quality of Service) control using trTCM (two rate three color marker) for each traffic unit of the user terminal; and a setting value calculation unit that calculates the CIR (Committed Information Rate) and PIR (Peak Information Rate) used in the trTCM in the physical interface, wherein the setting value calculation unit (1) calculates a weight for the standard deviation of the bandwidth used by high-priority traffic based on the influence of each standard deviation in the physical interface, and calculates a CIR and PIR that (1) have a negative correlation with the average value of the bandwidth used by high-priority traffic and (1) have a negative correlation with the multiplication of the standard deviation of the bandwidth used by high-priority traffic and the weight of the standard deviation.

2. The control device according to claim 1, wherein (1) the influence of the standard deviation of the bandwidth used by high-priority traffic is set to (1) the standard deviation of the bandwidth used by high-priority traffic, (2) the influence of the standard deviation of the number of simultaneously connected users of best-effort communication is set to (2) the standard deviation of the number of simultaneously connected users of best-effort communication multiplied by (3) the average bandwidth used by users of best-effort communication, and (3) the influence of the standard deviation of the bandwidth used by users of best-effort communication is set to (3) the standard deviation of the bandwidth used by users of best-effort communication multiplied by (2) the average number of simultaneously connected users of best-effort communication, and a weight is calculated for (1) the standard deviation of the bandwidth used by high-priority traffic so that the sum of the influences of each standard deviation satisfies 1.

3. The control device according to claim 1, wherein the set value calculation unit calculates the CIR and PIR from equation (1).

4. A program that causes a computer to function as a monitoring data processing unit that calculates the average value and standard deviation of (1) the bandwidth used by high-priority traffic, (2) the number of simultaneously connected users for best-effort communication, and (3) the bandwidth used by users for best-effort communication for a predetermined period in a physical interface that accommodates traffic from multiple users and performs QoS control using trTCM (two rate three color marker) on a traffic unit basis for each of the users; and a setting value calculation unit that calculates the CIR (Committed Information Rate) and PIR (Peak Information Rate) used in the trTCM in the physical interface, wherein the setting value calculation unit (1) calculates a weight for the standard deviation of the bandwidth used by high-priority traffic from the influence of each standard deviation in the physical interface, and calculates a CIR and PIR that (1) have a negative correlation with the average value of the bandwidth used by high-priority traffic and (1) have a negative correlation with the multiplication of the standard deviation of the bandwidth used by high-priority traffic and the weight of the standard deviation.

Images