This packet classifier module classifies packets using two token buckets. Each packet is classified depending on which token bucket is the first one that contains the required number of tokens for the packet.

The first token bucket is filled with tokens using the committed information rate. The maximum number of tokens is determined by the committed burst size. The second token bucket is filled with tokens using the excess information rate. The maximum number of tokens is determined by the excess burst size. Excess tokens from the first token bucket overflow into the second token bucket.

By default, packets consume 1 token per byte.

See also: SingleRateTwoColorMeter, SingleRateThreeColorMeter(1,2), DualRateThreeColorMeter

See also: SingleRateTwoColorClassifier, SingleRateThreeColorClassifier

Inheritance diagram

The following diagram shows inheritance relationships for this type. Unresolved types are missing from the diagram.

Extends

Name	Type	Description
MultiTokenBucketClassifier	simple module	This module classifies packets into the first bucket that contains the required number of tokens. The packet is forwarded on the output gate that corresponds to the selected bucket.

Parameters

Name	Type	Default value	Description
displayStringTextFormat	string	"contains %n tk\nmetered %p pk (%l)"	determines the text that is written on top of the submodule
reverseOrder	bool	false
buckets	object	[{initialNumTokens: dropUnit(committedBurstSize) / bitsPerToken, maxNumTokens: dropUnit(committedBurstSize) / bitsPerToken, tokenProductionRate: dropUnit(committedInformationRate) / bitsPerToken}, {initialNumTokens: dropUnit(excessBurstSize) / bitsPerToken, maxNumTokens: dropUnit(excessBurstSize) / bitsPerToken, tokenProductionRate: dropUnit(excessInformationRate) / bitsPerToken}]	array of objects specifying the token buckets example: [{initialNumTokens: 100, maxNumTokens: 1000, tokenProductionRate: 10, excessTokenModule: ".other"}]
tokenConsumptionPerPacket	double	0	how many tokens are consumed per packet
tokenConsumptionPerBit	double	1 / bitsPerToken	how many tokens are consumed per bit
committedInformationRate	double		committed information rate
excessInformationRate	double		excess information rate
committedBurstSize	int		committed burst size
excessBurstSize	int		excess burst size
bitsPerToken	int	8	how many bits are represented by 1 token

Properties

Name	Value	Description
display	i=block/classifier
class	MultiTokenBucketClassifier

Gates

Name	Direction	Size	Description
in	input
out [ ]	output

Signals

Name	Type	Unit
tokensChanged	double

Statistics

Name	Title	Source	Record	Unit	Interpolation Mode
numTokens	number of tokens	tokensChanged	vector	tk	linear

Source code

//
// This packet classifier module classifies packets using two token buckets.
// Each packet is classified depending on which token bucket is the first one
// that contains the required number of tokens for the packet.
//
// The first token bucket is filled with tokens using the committed information
// rate. The maximum number of tokens is determined by the committed burst size.
// The second token bucket is filled with tokens using the excess information
// rate. The maximum number of tokens is determined by the excess burst size.
// Excess tokens from the first token bucket overflow into the second token bucket.
//
// By default, packets consume 1 token per byte.
//
// @see ~SingleRateTwoColorMeter, ~SingleRateThreeColorMeter, ~DualRateThreeColorMeter
// @see ~SingleRateTwoColorClassifier, ~SingleRateThreeColorClassifier
//
simple DualRateThreeColorClassifier extends MultiTokenBucketClassifier
{
    parameters:
        double committedInformationRate @unit(bps); // committed information rate
        double excessInformationRate @unit(bps); // excess information rate
        int committedBurstSize @unit(b); // committed burst size
        int excessBurstSize @unit(b); // excess burst size
        int bitsPerToken = default(8); // how many bits are represented by 1 token
        buckets = [{initialNumTokens: dropUnit(committedBurstSize) / bitsPerToken, maxNumTokens: dropUnit(committedBurstSize) / bitsPerToken, tokenProductionRate: dropUnit(committedInformationRate) / bitsPerToken},
                   {initialNumTokens: dropUnit(excessBurstSize) / bitsPerToken, maxNumTokens: dropUnit(excessBurstSize) / bitsPerToken, tokenProductionRate: dropUnit(excessInformationRate) / bitsPerToken}];
        tokenConsumptionPerBit = 1 / bitsPerToken;
}

File: src/inet/queueing/classifier/DualRateThreeColorClassifier.ned

---

This documentation is released under the Creative Commons license