JPH11252097A - Device and method for sending packet - Google Patents

Abstract

PROBLEM TO BE SOLVED: To facilitate the selection of a packet and to suppress a burst to a minimum by sending packets from respective groups until a cumulative value considering the weight of a transmission packet amount for the each group exceeds a threshold value on a relevant round and setting a maximum value among the cumulative values of the respective groups at the time point of the preceding round end to a threshold value on the following round. SOLUTION: A packet sending device is provided with a packet queue (VQ) 2 for every group for storing packets for every group and a list managing part 8 for managing a list (SQ) 4 of identifiers of the QV in which an output waiting the packets exist. Further, a selector 6 is provided for repeatedly outputting the packet from the relevant VQ continuously as much as an amount corresponding to conditions while referring to the identifier of the VQ existing in the head of the SQ at the list managing part 8. Since the packet of the VQ for every group is fairly outputted based on the weight, the variable of F (e.g. transmission packet amount/weight, for example,) is calculated for every the VQ.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a packet transmitting apparatus and a packet transmitting method for fairly transmitting packets which are classified into a plurality of groups and awaiting output in the components of a packet communication network based on the weight of each group. About.

[0002]

2. Description of the Related Art In recent years, the importance of information communication has been increasing. In particular, the packet communication network called the Internet
With the advent of attractive applications represented by the World Wide Web, it has attracted many people around the world, both in the business world and as personal entertainment. The Internet has gained increasing awareness of its importance and its use has increased. And its importance is further increased by the increase in users. Thus, today, many companies and individuals are greatly expecting the development of packet communication network technology represented by the Internet.

One of the advantages of a packet communication network is that a plurality of users can flexibly share a physical line. In order to take advantage of this advantage, there has been proposed a packet transmission device that guarantees that the physical line capacity is fairly shared among users.

[0004] M. Shredhar and G. V
Argese: “Efficient Fair Q
ueueing Using Definit Rou
ndRobin ", pp231-242, SIGCOM
M '95, 1995. At DRR (Deficit
A packet sending algorithm called Round Robin) was shown. The DRR manages an identifier of a flow in which a packet waiting to be transmitted exists in a queue called an active list. A flow is a group of packets identified by a combination of a sender address, a recipient address, and the like. The flow identifier is extracted from the head of the active list, and the packets of the flow are collectively transmitted based on the weight determined for each flow. Specifically, the packets of the flow are continuously transmitted within a range in which the total length of the packets to be transmitted does not exceed the weight value set for the flow. If there is still a packet waiting to be output in the flow after transmitting the weight value to the full value, the flow identifier is re-entered at the end of the active list.

In the DRR, by setting an appropriate weight value for each flow, it is possible to set a throughput to be transmitted to a physical line to a desired ratio for each flow. The higher the weight, the higher the percentage of assigned throughput.

In order to determine which packet of a flow should be output next, it is only necessary to refer to the flow identifier at the head of the active list. There is an advantage that the required processing amount is constant and small.

However, in order to make use of this feature, it is necessary to set the weight of the DRR to a value equal to or larger than the maximum packet length assumed for the flow. Due to this constraint, a larger weight value must be set as the flow to be set to a smaller throughput has a larger maximum packet length.
Setting a large weight means that a temporarily large amount of packets are output in bursts from one flow.

In the end, the conventional packet transmitting apparatus has an advantage that packet selection is easy. However, even when a short packet is transmitted, a weight set based on the expected maximum packet length is used. There is a drawback that the burst property of each packet becomes high, and as a result, the buffer overflow of the packet easily occurs in the network.

[0009]

As described above, the conventional packet transmitting apparatus uses a weight set on the basis of the expected maximum packet length even when transmitting a short packet. There is a problem that the burst property is increased, and as a result, the buffer overflow of the packet in the network easily occurs.

The present invention has been made in view of the above circumstances, and has an advantage that a packet to be transmitted is easy to select, and a packet which can suppress packet transmission from each flow to a necessary minimum burst. An object of the present invention is to provide a transmission device and a packet transmission method.

[0011]

Means for Solving the Problems The present invention (claim 1) provides:
A packet transmitting apparatus for transmitting packets based on a list of identifiers of groups to which packets to be transmitted exist, wherein in each round of packet transmission based on the list, a weight of a transmitted packet amount for each group is set. Until the considered cumulative value (F in the embodiment) exceeds the threshold value (Fmax in the embodiment) in the round, means for transmitting packets from each group, and means for sending out packets from each group at the end of the preceding round. Means for setting a maximum value of the accumulated values (F) as a threshold value (Fmax) in a subsequent round.

[0012] Preferably, each time one packet is transmitted, the transmitting means divides a value obtained by dividing a packet length of the packet by a predetermined weight for a group to which the packet belongs to a value of the group up to that time. A new cumulative value may be obtained by adding to the cumulative value.

Preferably, the transmitting means calculates the cumulative value of the group to which the packet belongs each time one packet is transmitted, and the calculated cumulative value is still less than or equal to the threshold value in the round. If there is still a packet waiting to be transmitted in the group, one packet may be continuously transmitted from the group.

Preferably, a cumulative value of a group in which a packet to be newly transmitted is generated and the identifier is input to the list is set to a threshold value in the round or a predetermined value smaller than the threshold value. Means may be further provided.

Preferably, the apparatus may further comprise means for converting the accumulated value and the threshold value into smaller values at a predetermined timing while maintaining their relationship.
Preferably, in the case where one packet is composed of a plurality of cells having an order, when the last cell constituting the packet has arrived, the apparatus further comprises means for treating the packet as having arrived. You may.

Preferably, when one packet is composed of a plurality of cells having an order, when the first cell constituting a packet belonging to a predetermined group arrives, the packet is regarded as having arrived. Means for handling, and means for controlling so as not to transmit packets belonging to another predetermined group until the first cell transmits the last cell constituting the packet that has arrived. You may.

The present invention (claim 8) provides a means for storing input packets for each group, a means for managing a list of identifiers of groups to which packets to be sent exist, and a method for storing packets based on the list. Means for repeatedly selecting a group to be transmitted one by one; and, for each packet transmitted for the selected group, a value obtained by dividing a packet length of the packet by a predetermined weight for the group until then. Is added to the cumulative value of the group (F in the embodiment) to obtain a new cumulative value (F), and the obtained new cumulative value (F) is still the threshold value at the time (Fmax in the embodiment). If the following is true, means for continuously repeating the processing of sending one packet from the group and processing by this means are present in the list. Means for setting the maximum value among the accumulated values (F) of the respective groups at the time point as a new threshold value (Fmax) when all the identifiers have been executed. And

According to a ninth aspect of the present invention, there is provided a packet transmitting method in a packet transmitting apparatus for transmitting a packet based on a list of identifiers of groups in which a packet to be transmitted exists, wherein each round of packet transmission based on the list is provided. (Circulation), the cumulative value (F in the embodiment) in consideration of the weight of the transmission packet amount for each group is:
The threshold value in the round (Fma in the embodiment)
Until x), packets are transmitted from each group, and the maximum value of the accumulated value (F) of each group at the end of each round is set to the threshold value (F) in the next round.
max).

The present invention (claim 10) relates to a packet transmitting method in a packet transmitting apparatus for transmitting packets based on a list of identifiers of groups in which packets to be transmitted exist in each round, wherein the packet is transmitted based on the list. Each time one packet is transmitted for one selected group, the value obtained by dividing the packet length of the packet by a predetermined weight for the group is calculated as the cumulative value of the group up to that point (F in the embodiment). ) Is added to obtain a new cumulative value (F), and the obtained new cumulative value (F) is obtained.
Is still equal to or smaller than the threshold value in the round (Fmax in the embodiment), one packet is continuously transmitted from the group, and the obtained cumulative value (F) is determined as the threshold value in the round (Fmax). Fmax), a process of selecting the next group based on the list is performed.
The process is repeated until the process is performed for all identifiers existing in the list, and the maximum value among the cumulative values (F) of the respective groups at the end of this round is set as a threshold value (Fmax) in the next round. It is characterized by the following.

The present invention can be applied to, for example, a packet transmission scheduler in a router of the Internet network or a cell transmission scheduler of an ATM switch in an ATM communication network. Note that the present invention relating to the apparatus is also realized as an invention relating to a method, and the present invention relating to a method is also realized as an invention relating to an apparatus.

Further, the present invention relating to an apparatus or a method is provided for causing a computer to execute a procedure corresponding to the present invention (or for causing a computer to function as means corresponding to the present invention, or for causing a computer to correspond to the present invention). The present invention is also realized as a computer-readable recording medium in which a program for realizing the function of performing the above is recorded.

According to the present invention, in each round of packet transmission, packets are transmitted from each group until the cumulative value in consideration of the weight of the transmission packet amount for each group exceeds the threshold value in the round. At the same time as sending, the maximum value of the accumulated value of each group at the end of each round is set as a threshold value in the next round, so that the selection of packets to be sent depends on the number of groups, for example, the number of flows. Using the amount of packets and weights that are actually being sent out from time to time,
Since the amount of packets that each group can transmit in one round can be dynamically determined, the packet transmission of each group can be suppressed to the minimum necessary burst.

[0023]

Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 shows an example of the configuration of a packet transmission device to which the present invention is applied. This packet transmission device
For example, in the case of an ATM switch or an IP router, it corresponds to an output processing portion of each port. If each port performs priority control or band division control of a plurality of classes, it corresponds to an output processing portion in each of one or a plurality of classes.

First, a basic configuration of the present packet transmitting apparatus will be described. As shown in FIG. 1, the present packet transmitting apparatus includes a group-based packet queue (VQ) 2 for storing packets for each group, and a list (SQ) 4 for managing a list (SQ) 4 of VQ identifiers in which packets to be output exist. Management unit 8 and VQ at the head of SQ of list management unit 8
, And repeatedly outputting packets (one or more packets) from the corresponding VQ by an amount corresponding to a condition to be described later.
And a selector 6. Here, the meaning of “continuously” means that it is continuous in packets managed by the same SQ, and when a plurality of SQs are combined by priority control or the like, another SQ May be output.

When a packet waiting to be transmitted is generated in a VQ corresponding to a certain group, the SQ of the list
An identifier indicating Q is entered. The VQ identifier at the head of the SQ is extracted from the SQ after being referred to by the selector 6, and if there is still a packet waiting to be transmitted in the VQ indicated by the identifier, the identifier is added to the end of the SQ. It is put in again.

Next, the VQ shown in FIG. 1 will be described. As described above, the VQ is a queue for storing packets to be transmitted, which is divided into groups. Hereinafter, the identifier of VQ will be referred to as VCI. The classification of the group referred to here is, for example, by virtual connection, by IP packet flow, by application, or by quality class.

By the way, as the packet queue configuration of the present packet transmitting apparatus, various configurations can be considered according to the application. That is, (1) In addition to the configuration illustrated in FIG. 1, all “VQs” are simple FIFO queues, and (2) some “VQs” are simple FIFO queues. "VQ", a plurality of VQs scheduled by priority and weight are further included in the "VQ".
And (3) a configuration in which all “VQs” have a hierarchical structure.

In the above (2) and (3), for each VQ having a hierarchical structure, the queue may have two or more hierarchical structures (for example, scheduling is performed by priority or weight in "VQ"). There may be a configuration in which there are a plurality of “VQs”, and among the “VQs”, there are a plurality of VQs scheduled with priority and weight.
Of course, VQ having a hierarchical structure in (2) and (3) above
When there are a plurality of VQs, all the VQs having a hierarchical structure may have the same hierarchical structure, or may have a unique hierarchical structure for each VQ.

The configuration adopting a hierarchical structure for the packet queue is effective when sharing a physical link between groups having different characteristics and requirements, and the present invention (for example, its characteristic scheduling algorithm) The present invention can be applied to each of the portions that perform weighted scheduling in each layer. For example, it is possible to adopt a configuration in which the configuration of FIG. 1 exists in a certain VQ of FIG.

Next, the SQ of FIG. 1 will be described. As described above, from which VQ the packet should be output is S
Specified by Q. The SQ manages a list of VCIs of a VQ in which a packet to be transmitted exists, and indicates from which VQ the packet should be output by picking up the VCIs in order. Although the present invention works effectively regardless of the specific implementation method of the SQ, the SQ may simply be a VCI FIFO like an active list of a DRR. When a packet arrives at the idle VQ where there is no packet waiting to be transmitted and becomes active, the VC is added to the end of the SQ.
Enqueue I. The packet is sent at the beginning V of the SQ.
It starts from VQ indicated by CI. When interrupting the packet transmission of the current VCI and proceeding to the next VCI, the first V
CI is dequeued from the FIFO, and the head of SQ is changed to the new V
Set to CI. If there is still a packet to be transmitted to the VQ of the extracted VCI, the VCI is enqueued again at the end of the SQ.

Next, a unit of a series of processing and a round used in the following will be described. All VCIs managed by SQ
Is referred to herein as a “round”. Since the rounds are repeated, the end point of one round is the start point of the next round. Since the present invention works regardless of the specific definition of a round, its precise definition (eg, the definition of a break between adjacent rounds) is not necessarily important in the present invention. If the number of VCIs stored in the SQ is set in the counter at the start of the process, the counter is decremented each time one VCI is processed, and the time when the value reaches 0 is the end of the round. The end point of the round can be determined.

Next, variables F and Fmax used below will be described. In the present invention, a variable F is calculated for each VQ in order to fairly output the packets of each VQ (that is, each group) based on the weight. F is a value in consideration of the weight of the transmission packet amount of each VQ (for example, transmission packet amount / weight). For example, every time a packet is output from a certain VQ, a value obtained by dividing the packet length of the packet by the weight is added to F. The weight is determined in advance for each VQ.

In the present invention, a packet is output so that F becomes as equal as possible in each VQ. Therefore, one of the features of the present invention is that one threshold value Fmax is determined in each round, and packets are transmitted until the F of each VQ exceeds Fmax.

Another feature of the present invention is that at the start of each round (ie, at the end of the previous round), the largest of the Fs in each VQ is replaced by the F in the next round.
max.

Thus, the number of packets transmitted from each VQ increases only when a long packet is transmitted from one of the VQs by chance, but the output of packets from each VQ becomes small when only short packets are transmitted. That is, the value of Fmax is dynamically adjusted so that the burstiness of packet transmission from each VQ is suppressed as low as possible.

It should be noted that a certain VQ outputting a packet
If the FQ is still equal to or smaller than Fmax, but the VQ becomes empty, the identifier of the VQ is dequeued from the SQ, and the process proceeds to the next process.

Further, at the time when the last output waiting packet is output in a certain VQ, its F exceeds Fmax,
In addition, it is conceivable that the F becomes the maximum value at the end of the round. In such a case, the F may be set as the Fmax of the next round, but the VCI is included in the SQ in the next round. The maximum value of F of the VQ in which is present may be set to Fmax.

In order to avoid an increase in bursts due to transmission of a very long packet (that is, a packet having a very large packet length / weight) (especially when a very long packet continues at a specific VQ), it is necessary to use an adjacent packet. Determine the maximum value dmax of the increment of Fmax in the round to be performed,
If the value of F exceeds Fmax + dmax, processing such as changing the value of F from Fmax to Fmax + dmax of the round may be performed. In this case, the fairness may be slightly reduced, but there is an advantage that the burstiness is further reduced.

Here, one mode of the operation of the packet transmitting apparatus to which the present invention is applied will be described with reference to a specific example of the transmission order of the packets transmitted from the packet transmitting apparatus illustrated in FIG. FIG. 2 shows VQ [0], VQ
This is an example in which there are packets to be transmitted to three VQs [1] and VQ [2]. The packets are transmitted in order from the right packet to the left packet in FIG.

The Fmax of the round r is represented by Fmax (r). F of each of VQ [0] to VQ [2] is F
(0) to F (2). The weights of VQ [0] to VQ [2] are represented by W (0) to W (2).

The packet length of the packet p is represented by L (p). In this example, Fmax (0) at the start of round 0
And the initial values of F (0) to F (2) are set to 0.

Also, it is assumed that the weights W (0) to W (2) of each VQ are appropriately determined. That is, the width of the rectangle indicating the packet in FIG. 2 corresponds to the packet length / weight.

Here, for convenience of explanation, VQ
It is assumed that packet output is performed in the order of [0], VQ [1], and VQ [2]. By the way, in round 0, Fma
x (0) = F (0) = F (1) = F (2) = 0, and F of any VQ does not exceed Fmax (0), but when packets are output one by one, F is also Fmax
(0) will be exceeded. That is, one packet p00 is output from VQ [0], one packet p10 is output from VQ [1], and one packet p20 is output from VQ [2]. Ends, and the process proceeds to the next round 1.

At this point, the value of F of each VQ is F (0) = L (p00) / W (0) and F (1) = L
(P10) / W (1), F (1) = L (p20) / W
(2). Here, F (0) has the largest value among F (1) to F (3). Therefore, in Round 1, the value of Fmax (1) is the value of Fmax at the end of Round 0. Is the maximum value of F (0).

In round 1, first, packet p01 is output from VQ [0]. Then, F (0) = L (p
00) / W (0) + L (p01) / W (0) is Fmax
Since it exceeds (1), VQ [0] in round 1 is now
The transmission of the packet from is terminated. Similarly, VQ [1]
When one packet p11 is output from VQ and one packet p21 is output from VQ [2], the Fs of all VQs exceed Fmax (1) and shift to the next round 2.

Fmax (2) of round 2 becomes the value of F of VQ [0] where F is the largest at the end of round 1. In Round 2, first, VQ [0]
Two packets p02 are output. In the VQ to which Fmax is given, F = Fmax, so that only one packet is output.

Next, packet p12 is output from VQ [1]. In this case, F (1) = L (p10) / W
(1) + L (p11) / W (1) + L (p12) / W
Since (1) is still equal to or smaller than the value of Fmax (2), the packet p13 is continuously output from VQ [1]. At this point, F (1) exceeds Fmax (2), and thus the packet transmission from VQ [1] in round 2 ends. Here, the first sent p
Since the packet length of No. 12 was short, p13 could also be transmitted.

Then, one packet p from VQ [2]
When 22 is output, F of VQ [2] is also Fmax (2).
And moves to the next round 3. Round 3 Fm
ax (3) becomes the value of F of VQ [1] where F is the largest at the end of round 2.

Thereafter, similarly, packets are output from each VC as shown in FIG. As described above, according to the present invention, the amount of increase in Fmax is adjusted according to the length of the packet actually transmitted, so that it is found that the burstiness of the packet output from each VQ is suppressed as much as possible.

For example, in FIG.
Since the packet length of the packet p13 output from [1] is long, Fmax (3) and Fmax in the next round
Although the difference in (2) is a large value, the difference between max (4) and Fmax (3) in the next round 4 is small because the packet lengths transmitted from each VQ in round 3 were all short. I understand. That is, in FIG. 2, it can be seen that only the difference between the Fmax of the round in which a long packet is transmitted and the Fmax of the next round increases, and the burst property is minimized.

On the other hand, when the conventional DRR is used, for example, in FIG. 2, packet transmission control is performed such that Fmax is fixedly increased every time by an amount corresponding to the length of the longest packet (p13). Even when only short packets have arrived, packets are output in bursts.

In addition, from the viewpoint of fairness, as can be seen from FIG. 2, the Fm of the F value of each VQ in each round.
It can be seen that the difference from ax is always within one packet.
In other words, it can be said that the packet transmission device of the present invention is fair with an error within one packet.

Next, the handling of F and Fmax of each VQ will be described. First, the initial values of F and Fmax of each VQ may all be set to 0, but the present invention is not limited to this, and the initial values of Fmax may be set to other values. F of VQ may be set to an appropriate value equal to or less than Fmax.

Next, when the F of a certain VQ is very small as compared with Fmax or another F,
If a packet is transmitted from the VQ using the small value of F as it is, a packet is temporarily output as a large burst from one VQ. Therefore, in such a case, for example, it is preferable to correct F so that the difference from Fmax is reduced. For example, when a packet arrives at an empty VQ, the VQ
Is compared with Fmax at that time, and if F is smaller, F can be corrected to Fmax.

Next, the F or Fmax of each VQ increases with the transmission of a packet, but any F reaches the maximum value that can be expressed by the storage area allocated for storing the F or Fmax of each VQ. In the case where it is not assumed, for example, when the device is initialized before F reaches the maximum value and F and Fmax return to the initial values, each VQ
F or Fmax may be increased as it is, but otherwise, it is preferable to perform processing so that F or Fmax of each VQ does not exceed the above-mentioned maximum value.

For example, if F exceeds the maximum value that can be taken, the excess from the maximum value is set to F, the excess flag is turned on, and Fmax and Fmax are set in consideration of the excess flag. Calculate F and VCI to SQ
If all the excess flags of the F of the VQ containing the “on” are turned on, a method of turning off all the excess flags (however, the value of F is also corrected as described above) If Fmax exceeds a certain value at the start time (or end time), SQ
A method in which the same value is subtracted from each of F and Fmax of the VQ containing the VCI (however, the value of F is also corrected as described above), the start time (or the end time) of a certain round ), The F of all VQs
Various methods such as a method of subtracting fmin from each of F and Fmax of all VQs (however, the lower limit is set to 0) such that the one that takes the smallest value (fmin) other than 0 is set to 0 There is a method.

The operation of subtracting the same value from each of F and Fmax is performed, for example, in each round.
It is easy to reduce the Fmax of the round from the updated F value of the VQ every time the VCI is dequeued from the SQ. In this case, Fmax of the next round becomes the maximum value of F after being reduced.

Next, two specific examples of the algorithm of the packet transmitting apparatus according to this embodiment will be described. This algorithm may be implemented, for example, as part of a selector,
Alternatively, it is implemented as a control unit for the selector.

The variable names used in the algorithm described below and their meanings are summarized below. vci: identifier of VQ F [vci]: amount of output packets Fmax: threshold value of F The initial value is arbitrary. Ftmp: A variable for determining Fmax. Initial value = 0 Cnt: Counter for measuring the round period. Initial value = 0 Weight [vci]: Weight. IWeight [vci]: Reciprocal of weight. A smaller band allows a larger band to be acquired. PWeight [vci]: Weight in packet number units. SQ_Len: SQ VCI number PQ_Len [vci]: VQ packet number VQ_Len [vci]: VQ cell number EoPi: Input cell is End of
Packet cell EoPo: The transmitted cell is End of
Packet cell BoP [vci]: The next arriving user cell is the first cell of the packet mode [vci]: The mode of vci. The vci mode includes, for example, the following CELL, CT, and SF modes. CELL: Cell-based scheduling CT: Cut-thru in packet-based scheduling
out SF: Store a in packet-by-packet scheduling
ndForward pWmode: a mode for setting a weight in packet number units. Only in this mode, pWeight can be set to 1 or more, and F, Fmax, Ftmp, Cnt,
Block not using iWeight blockSQ_flag: Flag indicating whether SQ is blocking blockSQ_start_time: Time when SQ starts blocking blockVQ-flag [vci]: Flag indicating whether VQ is blocking now: current time: time First, FIG. 3 shows an example of a flowchart of an enqueue operation when a packet arrives at the present packet transmitting apparatus.

This flowchart is executed when it is determined that the packet is not to be discarded as a result of the discard judgment of the arriving packet. First, an appropriate VQ is selected and a packet is enqueued to the VQ (step S11).

Next, a packet input process is performed. First,
It is determined whether the VQ has changed from idle to active (step S12). Therefore, it is determined whether the variable PQ_Len [vci] indicating the number of packets in the VQ is less than 1, that is, 0.

If it is determined that the VQ has changed from idle to active, the new F [vci] value is set to the larger of the previous F [vci] value and the current Fmax ( Step S13).

If it is determined that VQ has changed from idle to active, vci is enqueued to SQ and SQ_Len, which is the number of VCIs in SQ, is incremented (steps S14, S15).

Finally, PQ_Len [vci] is incremented (step S16), and the enqueue process ends. Next, FIG. 4 shows an example of a flowchart of a dequeue for transmitting a packet from the packet transmitting apparatus.

This flowchart is executed when at least one VQ managed by the SQ is active, that is, when SQ_Len is greater than zero. First, a round process is performed.

It is checked whether the value of the counter Cnt is 0 (step S21). If 0, it means the end of the round, which also means the start of a new round. Specifically, the process for starting a new round is to set the value of the variable Ftmp in Fmax and set Snt in Cnt.
The value of Q_Len is set (step S22). SQ_
Len is the number of VCIs in the SQ.

Next, assuming that the VCI value at the head of the SQ is vci, a packet is output from the VQ indicated by vci (step S23). When a packet is output, its VQ
F [vci], which is the value of F, is (output packet length) /
Weight [vci] is added (step S24).
Weight [vci] is the weight of vci, and the larger the weight, the higher the bandwidth.

Subsequently, a packet output process is performed. First, the value of F [vci] updated by the addition is compared with Fmax (step S25). If F [vci] is still F
If the value is equal to or less than the value of max, the packet may be transmitted from the same VQ next time, and the subsequent processing is simple.
What is necessary is just to decrement Len [vci] (step S31).

Rather, F [vc
If i] is larger, the next packet is output from a different VQ. At this time, the vci at the head of the SQ is dequeued from the SQ, and SQ_Len is decremented (step S26). If the VQ is still active (PQ_Len [vci]> 1) (step S2)
7), re-enqueue the vci to the end of the SQ, and increment SQ_Len (step S2)
8). Then, Ftmp, which is a variable for determining the maximum value of F, is updated to the value of F [vci] (step S2).
9) The variable Cnt for determining the end of the round is decremented (step S30). Finally, PQ_L
en [vci] is also decremented (step S3)
1).

The above is an example of the operation algorithm of the present packet transmitting apparatus. As described above, in the present embodiment, the selection of the VQ to which the packet to be transmitted belongs is performed by a simple SQ enqueue and dequeue operation, so that the processing amount is constant regardless of the number of VQs handled. Furthermore, in order to dynamically determine Fmax so that the burst property is as low as possible,
It can be expected that the discard rate characteristics of the network will be better than in the conventionally known DRR method.

Next, an example in which the present invention is applied to an ATM communication system will be described. FIG. 5 shows an example of an enqueue flowchart, and FIGS. 6 and 7 show an example of a dequeue flowchart.
FIG. 8 shows an example of a flowchart of the timeout determination process.

In the ATM communication system, a packet is divided into fixed-length and short packets called cells and transferred. A packet is divided using a sublayer function usually called AAL5. In AAL5, packet division information is put in the header of the ATM cell. Specifically, in the header of the user cell,
1 indicating whether this is the cell that transfers the last part of the packet
Bit information is written. Here, the cell for transferring the last part of the packet is referred to as EoP (End of Pa).
a) cell. Also, the cell that transfers the first part is called a BoP (Begin of Packet) cell.

The packet transmission device of the ATM communication system normally performs scheduling in units of cells. In other words, transmission scheduling is performed regardless of which part of the packet into which the cell is divided is transferred. However, in some cases, scheduling in units of packets may be required. For example, there is a case where a function called VC merge is required. VC merge is V
C is one of the methods for realizing a multipoint point connection or a multipoint multipoint connection.
In this topology, cells arriving from multiple input links are merged into each of the output links. In AAL5, a plurality of cells that essentially transfer one packet are:
When a plurality of cells that transfer another packet are interleaved and joined together, there is a problem that the original packet cannot be separated. The ATM switch that performs merging buffers the cells from each input link with separate VQs and transmits the packets in units of packets to the output link, thereby preventing interleaving and overcoming this problem.

The packet transmitting apparatus described below enables scheduling on a packet basis. 2 for packet-by-packet scheduling
There are different methods. Cut Through (CT) mode and Store and Forward (SF) mode.

The CT mode is a mode in which a packet can be transferred to an output link as soon as a BoP cell arrives. However, cells from other input links that are merged into the same VC cannot be transferred to the output link until the EoP cell of the packet arrives.

On the other hand, the SF mode is a mode in which transmission of a BoP cell is not started until an EoP cell arrives. Although the packet transfer delay is larger than in the CT mode, the cells constituting one packet for starting transmission can be continuously output up to the EoP cell, and the transfer of cells arriving from another input link can be performed. There is an advantage that there is little hindrance.

In the CT mode, Eo of the packet being output is
If the PCell does not arrive forever, packets from other input links will be blocked forever. Therefore, the upper limit of the blocking time is determined using a timer, and the block is not blocked any longer. A mechanism is required.

First, FIG. 5 shows an example of an algorithm of an enqueue operation when a cell arrives at the present packet transmitting apparatus. This algorithm is executed when it is determined that the packet is not to be discarded as a result of the discard determination of the arrival cell.

First, an appropriate VQ is selected, a cell is enqueued to the VQ, and a variable VQ_Len [vci] for counting the number of cells stored in the VQ is incremented (step S41).

Subsequently, a blocking process is performed. First, whether or not the VQ is blocking another VQ is confirmed by blockVQ_flag [vci] (step S42). If blocking, blockVQ_flag [vc] to unblock
i] and blockSQ_flag are set to false (step S43).

Then, based on the VQ mode and the arriving cell, it is determined whether or not to perform the packet input process (step S44). The VQ mode includes a CELL mode in which scheduling is performed in units of cells, a CT mode in which scheduling of Cut Through is performed in units of packets, and a Store mode in which scheduling is performed in units of packets.
SF for scheduling nd Forward
There is a mode. It is necessary to determine whether the cell is a BoP cell, which is the head cell of the packet, or whether it is an EoP cell, which is the cell at the end of the packet. CELL
In the mode, each time one cell arrives, it is regarded that a packet has arrived. In the case of the CT mode, when a BoP cell arrives, it is regarded that a packet has arrived. Then, in the SF mode, when the EoP cell arrives, it is regarded that the packet has arrived.

When it is determined that a packet has arrived, a packet input process is performed. This packet input process is the same as in FIG. 3 when pWeight [vci] is 1. That is, first, it is determined whether or not the VQ has changed from idle to active (step S4).
5). Therefore, a variable PQ indicating the number of packets in the VQ
It is determined whether _Len [vci] is less than 1, that is, 0.

If it is determined that the VQ has changed from idle to active, the new F [vci] value is set to the larger of the previous F [vci] value and the current Fmax ( Step S46).

When it is determined that VQ has changed from idle to active, vci is enqueued to SQ, and SQ_Len, which is the number of VCIs in SQ, is incremented (steps S47, S48).

In each case, PQ_Len
[Vci] is incremented (step S49).
Then, at the end of the enqueue, if the arriving cell is an EoP cell, a flag BoP [vc] for each VQ to make the next arriving user cell a BoP cell.
i] is set (step S50).

Next, FIG. 6 and FIG.
9 is an example of a flowchart of a dequeue for transmitting a packet from the present packet transmitting device when applied to communication. FIG. 7 is a portion subsequent to FIG.

When at least one VQ managed by this SQ is active, that is, SQ_Len is greater than 0 and blockSQ_flag is fa
If it is 1se, this flowchart is executed.

First, a round process is performed. This process is the same as in FIG. That is, it is checked whether the value of the counter Cnt is 0 (step S61).

If 0, it means the end of the round, which also means the start of a new round. In the process for starting a new round, specifically, the value of the variable Ftmp is set to Fmax, and the value of SQ_Len is set to Cnt (step S62). SQ_Len is the number of VCIs in the SQ.

Next, the head vci of the SQ is read (step S63). In this step, only the head of the SQ is read, and no dequeue is performed. Subsequently, a process related to a timeout is performed.

In the present embodiment, it is determined that a timeout has occurred by performing dequeue processing when VQ_Len [vci] is 0 (step S6).
4). If it is determined that a timeout has occurred, the following processing is performed (step S65). A dummy EoP cell is transmitted from the VQ. And blockV
Q_flag [vci] is set to false to set the block to be released. Also, BoP [vci] is set to t
ru (true). This is because the next arriving cell should be treated as a BoP cell following the dummy EoP cell just transmitted. Also, since a dummy EoP cell has been transmitted, EoPo is set to true.

If it is determined that there is no block, the following processing is performed (step S66). A cell is output from VQ indicated by vci at the head of SQ. VQ_Len [vci] indicating the number of cells in the VQ is decremented, and a variable EoPo is set depending on whether or not the output cell is an EoP cell.

In any case, when the cell is transmitted, the value of F of the VQ is updated (step S67). Although the same formula as that of FIG. 4 may be used, here, the fact that the cell length is constant in the ATM communication is used, and the reciprocal of the weight is used for the calculation.
ht is added.

When the processing in FIG. 6 ends, the process proceeds to the processing in FIG. In FIG. 7, first, it is determined whether or not to perform the packet output process. If the vci is in the CELL mode, or if the cell transmitted in the CT mode or the SF mode is an EoP cell, a packet output process is performed (step S68).

The contents are the same as those described with reference to FIG. 4 if pWeight [vci] is set to 1. That is, first, FWmode that is not pWmode and updated by addition
If [vci] is still equal to or smaller than the value of Fmax (step S69), PQ_Len [vci] is decremented (step S31) because a packet can be transmitted from the same VQ next time, and the processing here ends. Becomes

[0096] Instead, F [vc
If i] is larger or pWmode, then a packet is output from a different VQ. At this time, the head vci of the SQ is dequeued from the SQ,
Q_Len is decremented (step S70), and the VQ is still active (PQ_Len [vc)
i]> pWeight [vci]) (step S
71), the vci is re-enqueued to the end of the SQ, and SQ_Len is incremented (step S7)
2). Then, Ftmp, which is a variable for determining the maximum value of F, is updated to the value of F [vci] (step S7).
3) The variable Cnt for determining the end of the round is decremented (step S74). Finally, PQ_L
en [vci] is also decremented (step S75),
The process here ends.

Thereafter, a blocking detection process is performed.
In FIG. 7, PQ_Len [vci] is 1 in the CT mode,
That is, when there is a packet to be transmitted to this VQ, if VQ_Len [vci] is 0, that is, if the number of transmittable cells of the VQ is 0, it is determined that blocking has occurred (step S76).

If it is determined that blocking has occurred, a flag bl indicating whether the SQ is blocking
ockSQ_flag and a flag blockVQ_flag indicating whether or not VQ is blocking
true and blockSQ_start_time is set to the current time now (step S76).

BlockSQ_start_time
Is the time when the blocking started, and the value now when the blocking started as shown in this flowchart.
May be set, or as another method, a value now may be set when a BoP cell is output.

The dequeue processing has been described above. Next, FIG. 8 shows an example of a flowchart of the timeout determination process. This process may be executed at an appropriate frequency in accordance with the accuracy of the timeout time, and may be executed fewer times than the enqueue process of FIG. 5 and the dequeue processes of FIGS. 6 and 7 described above.

The contents of the processing are as follows: blockSQ_f
Check if the SQ is currently blocking by looking at the lag, and if so, blockSQ_start_t
The difference between the current time and the current time is calculated, and the threshold time_t is calculated.
It is determined whether or not h is larger than h (step S81).

If a timeout has occurred, blockSQ_flag is sent to transmit a dummy EoP cell.
Is set to false, and a setting is made as if a cell to be transmitted to the SQ is waiting (step S82).

PWeight used in the flowcharts of FIGS. 5 and 7 is a weight in units of the number of packets.
LWFQ (Y. Ohba: “QLWFQ: A Queu
eLength Based Weightde Fa
ir QueueingAlgorithm in A
TM Networks ", Infocom '97.)
When implementing a method of weighted scheduling of cells introduced by a scheduler called, a value of 1 or more may be set. In order to realize this QLWFQ, pWmode shown in FIG. 7 may be set to true. Even in such a setting, a scheduling function for each packet in the CT mode or the SF mode can be realized.

Although FIGS. 5 to 8 are flowcharts in consideration of pWmode, it is needless to say that pWmode is used.
When de is not used, FIGS. 5 to 8 show pWmode
May be modified to a flowchart that does not take into account.

The above functions can be realized also as software. Further, the present embodiment is a computer-readable computer that records a program for causing a computer to execute a predetermined procedure (or for causing the computer to function as predetermined means, or for causing the computer to realize a predetermined function). It can also be implemented as a recording medium. The present invention is not limited to the above-described embodiment, and can be implemented with various modifications within the technical scope.

[0106]

According to the present invention, it is possible to easily select a packet to be transmitted irrespective of the number of flows, and to use the amount and weight of packets actually transmitted at each time.
Since the amount of packets that each flow can transmit in one round is dynamically determined, the packet transmission of each flow can be suppressed to the minimum necessary burst.

[Brief description of the drawings]

FIG. 1 is a diagram illustrating a configuration example of a packet transmission device according to an embodiment of the present invention.

FIG. 2 is an exemplary view for explaining a transmission order of packets transmitted from the packet transmission device according to the embodiment;

FIG. 3 is a flowchart illustrating an example of an enqueue operation;

FIG. 4 is a flowchart illustrating an example of a dequeue operation;

FIG. 5 is a flowchart illustrating an example of an enqueue operation in the ATM communication system;

FIG. 6 is a flowchart illustrating an example of a dequeue operation in the ATM communication system;

FIG. 7 is a flowchart showing an example of a dequeue operation in the ATM communication system;

FIG. 8 is a flowchart illustrating an example of a timeout determination process in the ATM communication system;

[Explanation of symbols]

 2 ... packet queue for each group 4 ... identifier list 6 ... selector 8 ... list management unit

Claims (10)

[Claims] 1. A packet transmitting apparatus for transmitting packets based on a list of identifiers of groups in which packets to be transmitted are present, wherein in each round of packet transmission based on the list, the amount of packets transmitted for each group is determined. Means for transmitting packets from each group until the cumulative value taking into account the weight exceeds the threshold value in the corresponding round, and the maximum value of the cumulative value of each group at the end of the preceding round Means for setting a threshold value in a round. 2. The transmitting means, each time one packet is transmitted, calculates a value obtained by dividing a packet length of the packet by a predetermined weight for a group to which the packet belongs, and accumulating the value of the packet up to that time. 2. The packet transmitting apparatus according to claim 1, wherein a new cumulative value is obtained by adding to the value. 3. The transmitting means calculates a cumulative value of a group to which the packet belongs each time one packet is transmitted, and the calculated cumulative value is still equal to or less than a threshold value in the round, and 3. The packet transmitting apparatus according to claim 1, wherein, when packets to be transmitted still exist in the group, one packet is continuously transmitted from the group. 4. A means for setting a cumulative value of a group in which a packet to be newly transmitted is generated and whose identifier is input to the list to a threshold value in the round or a predetermined value smaller than the threshold value. The packet transmitting apparatus according to claim 1, further comprising: 5. The apparatus according to claim 1, further comprising means for converting said accumulated value and said threshold value to a smaller value at a predetermined timing while maintaining their relationship. 3. The packet transmitting device according to 1. 6. When one packet is composed of a plurality of cells having an order, when the last cell constituting the packet arrives, the apparatus further comprises means for treating the packet as having arrived. The packet transmitting apparatus according to claim 1, wherein: 7. When one packet is composed of a plurality of cells having an order, when the first cell constituting a packet belonging to a predetermined group arrives, the packet is treated as having arrived. Means for controlling transmission of packets belonging to another predetermined group until the first cell transmits the last cell constituting the arriving packet. The packet transmission device according to claim 1, wherein 8. A means for storing input packets for each group, a means for managing a list of identifiers of groups in which packets to be transmitted exist, and one group to be transmitted based on the list. Means for repeatedly selecting each time, and for each packet transmitted for the selected group, a value obtained by dividing the packet length of the packet by a predetermined weight for the group is added to the accumulated value of the group up to that time. Adding a new accumulated value to obtain a new accumulated value, and if the obtained new accumulated value is still equal to or smaller than the threshold value at the time, repeating a process of continuously transmitting one packet from the group; When the processing by the means is executed once for all the identifiers present in the list, the cumulative value of each group at that time is calculated. Packet transmitting apparatus characterized by comprising a means for setting the Chino maximum value as a new threshold. 9. A packet sending method in a packet sending apparatus for sending packets based on a list of identifiers of groups in which packets to be sent are present, wherein each round of packet sending based on the list includes Until the cumulative value considering the weight of the transmitted packet amount exceeds the threshold value in the round, packets are transmitted from each group, and the maximum value among the cumulative values of each group at the end of each round is A packet sending method characterized by using a threshold value in a round. 10. A packet transmitting method in a packet transmitting apparatus for transmitting packets based on a list of identifiers of groups to which packets to be transmitted exist in each round, wherein one group selected based on the list is provided. about,
Each time one packet is transmitted, a value obtained by dividing the packet length of the packet by a predetermined weight for the group is added to the cumulative value of the group so far to obtain a new cumulative value. If the new cumulative value is still less than or equal to the threshold value of the current round, one packet is continuously transmitted from the group. If the calculated cumulative value exceeds the threshold value of the current round, The process of selecting the next group based on the list is repeated until all identifiers present in the list are executed, and the maximum value among the accumulated values of each group at the end of this round is A packet transmission method characterized by using a threshold value in a round.