KR100793633B1 - Device and method of providing traffic conditioning - Google Patents

Abstract

A traffic conditioner and a method thereof are provided to minimize a loss of packets in a burst situation by predicting traffic according to the size of a queue or the type of traffic and executing traffic conditioning on the basis of it. A traffic conditioner comprises a packet classifier(210), a traffic prediction meter(220), a packet marker(230), and a packet shaper(240). The packet classifier(210) classifies inputted packets. The traffic prediction meter(220) predicts the traffic flow of the classified packets and measures whether the traffic flow conforms to a predefined traffic profile. The traffic prediction meter(220) comprises an average prediction measurement part. The average prediction measurement part, based on the average of traffic information collected during a designated period, predicts traffic flow speed and judges whether the predicted traffic flow speed conforms to the traffic profile. The packet marker(230) executes packet marking according to the prediction and measurement results of the traffic prediction meter(220). The packet shaper(240) shapes the packets according to a processing result of the packet marker(230).

Description

Traffic conditioning device and how to provide traffic conditioning {DEVICE AND METHOD OF PROVIDING TRAFFIC CONDITIONING}

1 is an exemplary block diagram of an apparatus for providing traffic conditioning according to the prior art.

2 is a view illustrating an operation method of a packet marking unit implemented in the form of sr-TCM in the apparatus for providing traffic conditioning according to the prior art.

3 is an exemplary block diagram of an apparatus for providing traffic conditioning according to the present invention;

Figure 4 is an exemplary block diagram of a traffic prediction measurement unit of the traffic conditioning providing apparatus according to the present invention.

5 is an exemplary flow chart of a method for providing traffic conditioning in accordance with the present invention.

<Description of the symbols for the main parts of the drawings>

110: packet classifier 120: traffic measurement unit

130: packet marking unit 140: packet shaping unit

210: packet classification unit 220: traffic prediction measurement unit

223: average value prediction measurement unit 225: smoothing prediction measurement unit

228: prediction controller 230: packet marking unit

240: packet shaping unit

The present invention relates to a traffic conditioning apparatus and a method for providing traffic conditioning. More specifically, the present invention relates to a traffic conditioning apparatus and a method of providing a traffic conditioning. A traffic conditioning apparatus and a method for providing traffic conditioning that minimize packet loss in a burst situation.

According to the development of communication networks such as the Internet, there are various network service requests by users. At the same time, there is a growing demand for quality of service (QoS) guarantees.

Traffic conditioning is used to implement this QoS on the edge router of the Internet network.

Traffic Conditioning checks whether the traffic input from the edge router stage conforms to a predefined user's traffic profile, and performs basic QoS processing or other additional depending on the compliance. It provides a process that provides the basis for QoS processing.

Such basic QoS processing may be, for example, processing such as marking a packet that changes the priority of a packet or packet drop through rate limiting, and additional QoS processing may be the result of traffic conditioning. For example, a process may be used to influence traffic flow control or packet scheduling.

A functional module that implements such traffic conditioning is referred to herein as a "traffic condition provision device".

The configuration of the traffic conditioning providing apparatus depends on the range in which the traffic conditioning function is defined.

When defining the traffic conditioning in a broad aspect, the traffic conditioning providing apparatus includes a packet classifier, a traffic meter, a packet marker, a packet dropper, and a packet shaping unit ( shaper).

However, the conventional traffic condition providing apparatus generally includes only a packet measuring unit and a packet marking unit, and includes only a function of measuring traffic and indicating compliance in a packet according to the result.

1 is an exemplary block diagram of an apparatus for providing traffic conditioning according to the prior art.

As illustrated, a conventional traffic condition providing apparatus includes a packet classification unit 110, a traffic measurement unit 120, a packet marking unit 130, and a packet shaping unit 140.

The packet classifier 110 classifies packets.

The traffic measuring unit 120 is input to the traffic conditioning providing apparatus and measures the traffic flow of packets classified by the packet classifying unit 110.

Typically, bandwidth is measured through the input rate of traffic flow or the degree of burst is measured.

The traffic measuring unit 120 compares a predetermined or predefined traffic profile with a traffic profile of the input flow to determine compliance, and accordingly, the packet marking unit 130 performs necessary marking processing.

In general, it is classified into three cases of satisfying a predetermined or predetermined traffic profile, exceeding a predetermined range, or exceeding a predetermined range.

For example, when the packet marking unit 130 is implemented in DiffServ format, the packet marking unit 130 marks the packets as "Green", "Yellow", and "Red" for each case. Alternatively, such marking may be indicated in each case using the terms "Committed", "Excess", and "Violated" for, for example, Cisco equipment.

DiffServ is used in the form of a single-rate three-color marker (sr-TCM) and a double-rate three-color marker (tr-TCM).

For example, the sr-TCM and tr-TCM described in the RFC 2697 and RFC 2698 specifications, respectively, basically use a double token bucket structure.

That is, while sr-TCM uses only the committed information rate (CIR) as the token update rate, tr-TCM uses both CIR and peak information rate (PIR) as the token update rate.

2 is a view showing an operation method of a packet marking unit implemented in the form of sr-TCM in the apparatus for providing traffic conditioning according to the prior art.

As shown, the input packet is marked as a "Green" packet if its size B is smaller than the first token counter Tc, and if it is larger than the first token counter Tc but smaller than the second token counter Te, Yellow "packet, if larger than the second token counter Te, it is marked as" Red "packet.

The packet shaping unit 140 shapes the traffic based on the results of the traffic measuring unit 120 and the packet marking unit 130.

As described above, in the conventional traffic condition providing apparatus, the sr-TCM or tr-TCM used as the packet marking unit implements packet marking in such a manner that the token bucket counter is incremented according to a predetermined or predefined rate. This method has a merit that it can be easily implemented as a simple structure. However, even when a packet flows into a burst, a packet may be discarded when the packet reaches a certain speed regardless of the size of the queue.

Therefore, there is an increasing need for a method for packet marking in consideration of the queue size.

An object of the present invention is not a method of increasing the token bucket counter, but a traffic conditioning apparatus and traffic for minimizing packet loss in a burst situation by predicting the traffic according to the size of the queue buffer or the type of the traffic and performing the traffic conditioning based on the traffic. The present invention provides a method of providing conditioning.

In order to achieve the above technical problem, the present invention provides a packet classification unit for classifying an input packet, and a traffic prediction measurement unit for predicting the traffic flow of the classified packet and measuring whether the traffic flow conforms to a predefined traffic profile. And a packet marking unit for marking a packet according to the prediction and the measurement result of the traffic prediction measuring unit, and a packet shaping unit for shaping the packet according to the processing result of the traffic prediction measuring unit and the packet marking unit. The predictive measurer includes a mean value predictive measurer that predicts the traffic flow rate based on a value obtained by averaging the collected traffic information within a predetermined period and determines whether the traffic flow conforms to the traffic profile. .

In another aspect, the present invention provides the steps of: (a) classifying incoming packets; (b) predicting traffic flows of the classified packets and measuring whether the traffic flow conforms to a predefined traffic profile based on this; c) marking a packet according to the prediction and measurement result, and (d) a packet shaping unit for shaping the packet according to the results of steps (b) and (c), wherein the step (b) (B-1) extracting a value obtained by averaging the traffic information collected within a predetermined period; (b-2) predicting the traffic flow rate based on the averaged value; and (b-3 Determining whether the traffic flow conforms to the traffic profile.

Hereinafter, a traffic conditioning apparatus and a method for providing traffic conditioning according to the present invention will be described in more detail with reference to the accompanying drawings.

3 is an exemplary block diagram of an apparatus for providing traffic conditioning according to the present invention.

As illustrated, the apparatus for providing traffic conditioning according to the present invention includes a packet classification unit 210, a traffic prediction measurement unit 220, a packet marking unit 230, and a packet shaping unit 240.

The packet classification unit 210 classifies a packet input to the traffic condition providing apparatus according to the present invention. This packet classification is the same as the conventional traffic condition providing apparatus, so a detailed description thereof will be omitted.

The traffic prediction measuring unit 220 estimates the traffic flow of the packets classified by the packet classification unit 210 and measures whether the prediction result conforms to a predefined traffic profile.

The packet marking unit 230 marks the packet according to the prediction and the measurement result of the traffic prediction measuring unit 220.

The packet shaping unit 240 shapes the packets according to the processing results of the traffic prediction measuring unit 220 and the packet marking unit 220.

The present invention is characterized in that the traffic prediction measuring unit 220 predicts traffic and measures compliance based on the traffic prediction. The other packet marking unit 230 or the packet shaping unit 240 is similar to the conventional case, so A description thereof will be omitted and the traffic prediction measuring unit 220 will be mainly described.

4 is an exemplary block diagram of a traffic prediction measuring unit of the apparatus for providing traffic conditioning according to the present invention.

As shown, the traffic prediction measurement unit 220 of the present invention includes an average value prediction measurement unit 223, a smoothing prediction measurement unit 225, and a prediction control unit 228. Although illustrated, the traffic prediction measurement unit 220 is configured to include both the average value prediction measurement unit 223, the smoothing prediction measurement unit 225, and the prediction control unit 228. The configuration may include only the mean value prediction measurement unit 223 or only the smoothing prediction measurement unit 225.

Hereinafter, each configuration will be described in more detail.

The average value prediction measurement unit 223 predicts the traffic flow rate based on the averaged value of the traffic information collected within a predetermined period and determines whether the traffic flow conforms to the predefined traffic profile.

In other words, it is determined through the prediction whether the traffic flow conforms to the traffic profile.

In this case, if the predicted traffic flow rate is MA, the number of traffic information collected within a certain period is N, and the traffic flow rate is collected within a certain period, D, the following equation 1 can be configured.

MA _{t +1} = (D _t + D _{t -1} + D _{t -2} + ... + D _{t -N + 1} ) / N

(Where t represents each measurement point)

That is, after collecting the traffic flow rate through a certain period, it is assumed that the future traffic flow rate is an average traffic flow rate within this period, and determines whether to comply with the predefined traffic profile based on this.

Here, N value may be the most important parameter in traffic flow rate prediction. That is, it is determined whether or not the traffic flow rate prediction based on the present invention is effectively applied according to how many average traffic flow rate data are obtained.

This N value is preferably set based on the size of the queue buffer corresponding to the traffic conditioning apparatus according to the present invention. That is, it is set based on the traffic flow that can be stored in the queue buffer.

For this reason, when the traffic conditioning apparatus according to the present invention includes the mean value predictive measurer 223, an average value parameter setter (not shown) may be further included for the variable setting of the N value.

In addition, the smoothing prediction measuring unit 225 determines whether the traffic flow conforms to the traffic profile based on the current traffic flow rate and the predicted traffic flow rate.

In this case, when the smoothing prediction measurement unit 225 determines the traffic flow rate as the basis of F _{t + 1} , the current traffic flow rate as D _t , and the predicted traffic flow rate as MA _t . , F _{t + 1} may be configured to satisfy the following equation (2).

F _{t + 1} = αD _t + (1-α) MA _t (α is a parameter representing the ratio of the current traffic flow rate value to the past rate value)

Likewise, the α value may be the most important parameter in the traffic flow rate prediction. In other words, depending on how much the rate is adjusted, it is determined whether the traffic flow rate prediction based on the present invention is effectively applied. In other words, if the α value is properly adjusted, the degree of smoothing the rapidly increasing traffic flow rate value in the burst situation can be controlled.

This α value is preferably set based on the traffic conditions of the traffic conditioning apparatus according to the present invention. To this end, when the traffic conditioning apparatus according to the present invention includes the smoothing prediction measuring unit 225, the traffic conditioning apparatus may further include a smoothing parameter setting unit (not shown) that variably sets the α value.

Meanwhile, as shown in FIG. 4, the traffic prediction measurement unit 220 of the traffic conditioning apparatus according to the present invention may be configured to perform prediction and measurement by using both the average value prediction measurement unit 223 and the smoothing prediction measurement unit 225. have. In this case, traffic conditioning can be performed more efficiently by mixing the average value prediction measuring unit 223 and the smoothing prediction measuring unit 225 or selecting them according to traffic conditions.

In this case, the traffic prediction measuring unit 220 of the traffic conditioning apparatus according to the present invention further includes a prediction controller 228.

The prediction controller 228 controls the operations of the mean value prediction measuring unit 223 and the smoothing prediction measuring unit 225.

For example, in the case of the average predictive measurer 223, control is possible by setting the N value in accordance with the traffic situation, or in the case of the smoothed predictive measurer 225, the α value is controlled in accordance with the traffic condition. Is possible.

Alternatively, if the prediction controller 228 determines that only one of the average value prediction measuring unit 223 and the smoothing prediction measuring unit 225 is suitable for the traffic situation, the prediction control unit 228 selects such a suitable configuration to predict the traffic flow, and as a result, It is possible to control to determine whether or not to comply with the traffic profile, and through this configuration it is possible to respond to various traffic conditions.

According to the configuration of the traffic prediction measurement unit 220 according to the present invention, unlike the conventional traffic measurement method such as sr-TCM or tr-TCM, it is smoothed so that the queue buffer corresponding to the traffic conditioning apparatus can be used properly. Depending on the traffic situation, for example, bursty high-speed traffic can minimize losses.

In addition, for example, the average value prediction measurement unit 223 and the smoothing prediction measurement unit 225 are appropriately selected according to the queue buffer size and the type of the traffic of the edge router in which the traffic condition providing apparatus according to the present invention is arranged, and the parameters are set by the network administrator. Can be controlled as desired.

5 is an exemplary flowchart of a method for providing traffic conditioning according to the present invention.

First, the input packet is classified (S110).

Thereafter, the traffic flow of the packet classified in step S110 is predicted, and based on this, it is measured whether the traffic flow conforms to a predefined traffic profile (S120).

Thereafter, the packet is marked according to the result of the prediction and measurement in step S120 (S130).

Thereafter, the packets are shaped according to the results of steps S120 and S130 (S140).

In the method for providing traffic conditioning according to the present invention, steps S110, S130, and S140 are the same as in the conventional case, and thus detailed description thereof will be omitted.

Step S120 is characterized by estimating the traffic flow rate through the above-described average value prediction measurement or smoothing prediction measurement and measuring whether the traffic flow conforms to a predefined traffic profile based on this.

For this purpose, each prediction is performed by referring to Equation 1 or Equation 2 above. Also, the parameter setting step of Equation 1 or Equation 2 above includes an efficient traffic conditioning according to the size of the queue buffer or the traffic situation. to provide.

The details of the method for providing traffic conditioning according to the present invention are duplicated with the description of the apparatus for providing traffic conditioning according to the present invention with reference to FIGS.

Although the configuration of the present invention has been described in detail, it is only for illustrating the present invention, and the protection scope of the present invention is not limited thereto, and the protection scope of the present invention is defined through the description of the claims.

As described above, according to the present invention, it is possible to minimize the packet loss in the burst situation by estimating the traffic according to the size of the queue buffer or the type of the traffic, rather than increasing the token bucket counter. For example, QoS can be guaranteed when applied to an edge router or edge multilayer switch portion of an internet network.

Claims (19)

A packet classification unit for classifying an input packet, A traffic prediction measurement unit for predicting a traffic flow of the classified packet and measuring whether the traffic flow conforms to a predefined traffic profile; A packet marking unit for marking a packet according to the prediction and the measurement result of the traffic prediction measuring unit; Including a packet shaping unit for shaping the packet according to the processing results of the traffic prediction measurement unit and the packet marking unit,  The traffic prediction measurement unit, An average value prediction measurement unit for predicting a traffic flow rate based on a value obtained by averaging traffic information collected within a predetermined period and determining whether the traffic flow conforms to the traffic profile Traffic conditioning apparatus comprising a. delete The method of claim 1, When the predicted traffic flow rate is MA, the number of traffic information collected within a certain period is N, and the traffic flow rate collected within the predetermined period is D, MA _{t + 1} = (D _t + D _t-1 + D _t-2 + ... + D _{t-N + 1} ) / N, where t represents each measurement point The traffic conditioning device that satisfies. The method of claim 3, wherein the traffic prediction measurement unit, An average value parameter setting unit configured to set the N value according to a size of a queue buffer corresponding to the traffic conditioning apparatus Traffic conditioning apparatus further comprising. The method of claim 1, wherein the traffic prediction measurement unit, Smoothing prediction measurement unit for determining whether the traffic flow conforms to the traffic profile based on the current traffic flow rate and the predicted traffic flow rate Traffic conditioning apparatus comprising a. The method of claim 5, When the traffic flow rate based on determining whether the smoothing prediction measurement unit conforms to the traffic profile is F _{t + 1} , and the current traffic flow rate is D _t , and the predicted traffic flow rate is MA _t , the F _{t +1} is F _{t + 1} = αD _t + (1-α) MA _t (α is a parameter representing the ratio of the current traffic flow rate value to the past rate value) The traffic conditioning device that satisfies. The method of claim 6, wherein the traffic prediction measurement unit, Smoothing parameter setting unit for variably setting the α value Traffic conditioning apparatus further comprising. The method of claim 1, wherein the traffic prediction measurement unit, A smoothing prediction measurement unit for determining whether the traffic flow conforms to the traffic profile based on a current traffic flow rate and a predicted traffic flow rate; Prediction controller for controlling the operation of the average prediction measurement unit and the smoothing prediction measurement unit Traffic conditioning apparatus comprising a. The method of claim 8, When the predicted traffic flow rate of the mean value predictive measurer is MA, the number of traffic information collected within a predetermined period is N, and the traffic flow rate collected within the predetermined period is D. MA _{t + 1} = (D _t + D _t-1 + D _t-2 + ... + D _{t-N + 1} ) / N, where t represents each measurement point The traffic conditioning device that satisfies. The method of claim 8, When the traffic flow rate that is the basis for determining compliance with the traffic profile of the smoothing prediction measurement unit is F _{t + 1} , the current traffic flow rate is D _t , and the predicted traffic flow rate is MA _t . F _{t + 1} is F _{t + 1} = αD _t + (1-α) MA _t (α is a parameter representing the ratio of the current traffic flow rate value to the past rate value) The traffic conditioning device that satisfies. The method of claim 10, And the prediction controller sets the N value or the α value to perform the control. The method of claim 8, And the prediction controller selects and controls only one of the average value prediction measuring unit and the smoothing prediction measuring unit to determine whether the traffic flow conforms to the traffic profile. (a) classifying incoming packets; (b) predicting the traffic flow of the classified packet and determining whether the traffic flow conforms to a predefined traffic profile based on the traffic flow; (c) marking a packet according to the prediction and measurement result; (d) a packet shaping unit for shaping the packet according to the result of step (b) and step (c) Including but not limited to: Step (b) is, (b-1) extracting a value obtained by averaging traffic information collected within a predetermined period of time; (b-2) predicting the traffic flow rate based on the averaged value; (b-3) determining whether the traffic flow conforms to the traffic profile Traffic conditioning method comprising a. delete The method of claim 13, When the predicted traffic flow rate is MA, the number of traffic information collected within a certain period is N, and the traffic flow rate collected within the predetermined period is D, MA _{t + 1} = (D _t + D _t-1 + D _t-2 + ... + D _{t-N + 1} ) / N, where t represents each measurement point Traffic conditioning method that satisfies. The method of claim 15, wherein step (b) comprises: (b-4) further comprising setting the N value. The method of claim 15, wherein step (b) comprises: (b-5) extracting an elementary traffic flow rate that determines whether the traffic flow conforms to the traffic profile based on the current traffic flow rate and the predicted traffic flow rate; (b-6) determining whether the elementary traffic flow rate conforms to the traffic profile Traffic conditioning method comprising a. The method of claim 17, When the basic traffic flow rate is F _{t + 1} , the current traffic flow rate is D _t , and the predicted traffic flow rate is MA _t , the F _{t + 1} is represented by the following equation. F _{t + 1} = αD _t + (1-α) MA _t (α is a parameter representing the ratio of the current traffic flow rate value to the past rate value) Traffic conditioning method that satisfies. The method of claim 18, wherein step (b) (b-7) variably setting the N value or the α value.

Images