WO2014071636A1

WO2014071636A1 - Packet transmission method and device

Info

Publication number: WO2014071636A1
Application number: PCT/CN2012/084474
Authority: WO
Inventors: 孙艳平; 王运富; 黄雪峰
Original assignee: 华为技术有限公司
Priority date: 2012-11-12
Filing date: 2012-11-12
Publication date: 2014-05-15
Also published as: CN103229465A

Abstract

The present invention relates to the field of communications. Disclosed are a packet transmission method and device, capable of reducing the number of calculations of required tokens during the packet transmission process. The method comprises: receiving a packet, and calculating according to the packet length the number of tokens T_L used for transmitting the packet (101); acquiring the number of currently available tokens T_add in a token bucket (102); if the number of the currently available tokens T_add is smaller than the number of tokens T_L used for transmitting the packet, then calculating the delay time T_delay and starting to keep the time, the delay time being T_delay = (T_L-T_add)/V, wherein V represents the token filling rate (103); and transmitting the packet when the delay time expires (104). The present invention is applicable to network packet transmission.

Description

Method and device for transmitting messages

The present invention relates to the field of communications, and in particular, to a method and an apparatus for transmitting a message. Background technique

The token bucket algorithm is the most common algorithm used for traffic shaping (Traffic Shaping) and rate limiting ( Rate Limiting). The token bucket algorithm can also be used to control the number of data sent to the network and allow the transmission of bursty data.

The token bucket algorithm places tokens in the bucket according to the user's actual needs according to the speed set by the user. When the number of tokens reaches the bucket capacity, the number of tokens does not increase. When sending packets, if the number of tokens is sufficient to send packets. The packet is sent normally, and the number of tokens is also reduced during the sending process. If the number of tokens is too small to send a packet, the packet is discarded. A token bucket is also a tool for controlling data traffic. When a token bucket has no token, the packet is not sent. It can be sent only after a new token is generated in the bucket. This limits the traffic of the packet. It can only be less than or equal to the speed at which tokens are generated, and the purpose of limiting traffic is achieved. However, in the prior art, the method of filling a token by a token bucket is a timed padding. The padding method is simple to implement, but the operation is frequent. When a packet with a long length is sent and the rate is high, multiple times occur. After the padding is enabled, the packet is allowed to pass. This increases the number of times the system needs to calculate the token during the packet sending process. Summary of the invention

The embodiment of the invention provides a method and a device for transmitting a message, which can reduce the number of calculations of the token requirement in the message sending process.

In order to achieve the above objective, the embodiment of the present invention adopts the following technical solutions: The first aspect provides a method for sending a packet, including: receiving a packet, and calculating, according to the length of the packet, a request for sending the packet Number of cards TL; Get the number of currently available tokens in the token bucket T _add ; If the currently available number of tokens T _{ADD is} less than the number of tokens TL required to send the message, calculate the delay time T _delay and start Timing, the delay time is:! ^^ ^ - Ding ^)^, where V represents the filling rate of the token;

When the delay time ends, the message is sent. In a first possible implementation manner, in combination with the first aspect, the acquiring the number of currently available tokens in the token bucket, specifically includes:

Obtaining the number of tokens T _c and the filling time T _flll from the token information table, where T _c indicates that the number of tokens remaining in the token bucket of the previous message is sent, and T _flll indicates that the last packet is sent. a time interval dT at which the time 1^ of the received message is received and the filling time T _{flll is calculated} ; the time interval dT is multiplied by the filling rate V of the token, and then rounded down, plus The number of tokens T _{c is} obtained as the number of currently available tokens T _add , that is, D2, LdT*Vj.

In a second possible implementation, in combination with the first possible implementation, the method further includes:

The token flag is obtained from the token information table, and when the token flag is positive, the step of calculating the delay time is performed.

In a third possible implementation manner, combining the second possible implementation manner, the method specifically includes: if the current available number of tokens T _{ADD is} smaller than a number of tokens TL required to send the packet, The card flag is modified to be negative to indicate that the subsequent message is not processed; the token flag is modified to be positive at the end of the delay time to indicate that the subsequent message is processed. In a fourth possible implementation, in combination with the first possible implementation manner, the second possible implementation manner, or the third possible implementation manner, the method further includes:

Updating the padding time T _flll in the token information table with T _n +T _delay to LT _add + V*

T _delay -TjJ updates the number of tokens T _{e of} the token information table, at which time T _flll indicates that the message is sent

In a fifth possible implementation, in combination with the first possible implementation, the method further includes:

If the currently available number of tokens T _{add is} greater than or equal to the number of tokens T _L required to send the message, the padding time T _{flll of} the token information table is updated at the time T _n when the message is received. The token number T _{e of} the token information table is updated with T _add -TL, and the message is transmitted. The second aspect provides a device for sending a message, including: a token processor and a delay device, wherein: the token processor is configured to receive a packet, and calculate a sending station according to the length of the packet. The number of tokens TL required for the message; and the number of currently available tokens T _ADD in the token bucket;

The delay, if the token processor for the acquired current is less than the number of available tokens T _add the packet transmission required number of tokens T _L, the calculated delay time T _delay and start timing The delay time is: T _delay = ( T _L - T _add ) / V , where V represents the filling rate of the token; the token processor is further configured to send when the delay time ends The message. In a first possible implementation, in combination with the second aspect, the token processor further includes:

The token information obtaining unit is configured to obtain the number of tokens T _c and the filling time T _fm from the token information table, where T _c represents the number of tokens remaining in the token bucket after the last packet is sent, T _flll After the transmission time indicates a packet; timing unit, for calculating the packet receiving time T _n _T flll with the filling time interval dT;

a token calculation unit, configured to multiply the time interval dT by the filling rate V of the token, and then round down, and add the number of tokens T _c to obtain the currently available number of tokens T _Add , ie T _add =T _C + LdT *VJ.

In a second possible implementation manner, in combination with the first possible implementation manner, the delay device is further configured to obtain a token flag from the token information table, when the token flag is positive, Used to calculate the delay time _Tdelay . In a third possible implementation, in combination with the second possible implementation, the token processor is further configured to: if the currently available number of tokens T _{add is} smaller than a token required to send the packet When the number T _L is set, the token flag is modified to be negative to indicate that the subsequent message is not processed; and the token flag is modified to be positive at the end of the delay time to indicate that the subsequent message is processed continuously. .

In a fourth possible implementation, in combination with the first possible implementation manner, the second possible implementation manner, or the third possible implementation manner, the token processor is further configured to use T _N +T _DELAY updates the fill time T _FLLL in the token information table to LT _ADD + V *T _DELAY -T _L J Updating the number of tokens T _{c of} the token information table, at which time T _flll indicates the time at which the message is sent,

In a fifth possible implementation, in combination with the first possible implementation manner, the token processor is further configured to: if the currently available number of tokens T _{add is} greater than or equal to an order required to send the packet The number of cards TL, the filling time T _{FLLL of} the token information table is updated at the time T _{N of} receiving the message, the number of tokens T _{C of} the token information table is updated by T _ADD _-TL , and then the sending place Said message.

In a third aspect, an apparatus for transmitting a message is provided, comprising: at least one processor, a memory, a communication interface, and a bus, wherein the at least one processor, the memory, and the communication interface are connected by a bus and complete communication with each other;

The processor is configured to receive a packet by using the communication interface, and calculate a number of tokens T _L required to send the packet according to the length of the packet; the processor is further configured to obtain a current available in the token bucket. The number of tokens T _add ; the processor is further configured to calculate a delay time T _DELAY and start timing if the currently available number of tokens T _{add is} less than a number of tokens TL required to send the message, the delay The time is: T _delay = ( T _L - T _add ) / V, where V represents the filling rate of the token; the processor is further configured to: when the delay time ends, send the Message.

In a first possible implementation, in combination with the third aspect, the processor is further configured to obtain the token number T _c and the filling time T _fl11 from the token information table, where T _c indicates that the last message is sent. The number of tokens remaining in the token bucket, T _flll indicates the time when the last message is sent; and is also used to calculate the time interval dT between the time T _{n of} receiving the message and the padding time T _flll ; Multiplying the time interval dT by the filling rate V of the token, and then rounding down, plus the number of tokens T _c , to obtain the current available number of tokens T _add , that is, T _add = T _c + L dT*Vj.

In a second possible implementation, in combination with the first possible implementation, the processor is further configured to obtain a token identifier from the token information table, and when the token identifier is positive, The delay time _{Tdelay is} calculated. In a third possible implementation manner, in combination with the second possible implementation manner, the processor is further configured to: if the currently available number of tokens T _{add is} smaller than a number of tokens T _L required to send the packet, Then, the token flag is modified to be negative to indicate that the subsequent message is not processed; and the token flag is modified to be positive at the end of the delay time to indicate that the subsequent message is processed. In a fourth possible implementation, in combination with the first possible implementation manner, the second possible implementation manner, or the third possible implementation manner, the processor is further configured to use T _n +T _delay Updating the padding time T _flll in the token information table, and updating the token number T _{c of} the token information table by LT _add +v* T _delay -Tjj. At this time, T _FLLL indicates that the message is sent. At time, T _c represents the number of tokens remaining in the token bucket after the message is sent.

In a fifth possible implementation, in combination with the first possible implementation manner, the processor is further configured to: if the currently available number of tokens T _{ADD is} greater than or equal to the number of tokens required to send the packet, TL And updating the filling time T _{FM of} the token information table at the time T _{N of} receiving the message, updating the number of tokens T _{C of} the token information table by T _ADD -TL, and transmitting the message . The method and device for sending a message provided by the embodiment of the present invention can enable the release of the message after the token is filled in at least one time in the delay period, and the message is prevented from being filled for a plurality of times. After the permission is allowed to occur, the delay processing can reduce the number of calculations of the token requirement during the message transmission. DRAWINGS

BRIEF DESCRIPTION OF THE DRAWINGS The accompanying drawings, which are set forth in the description of the claims Other drawings can also be obtained from these drawings on the premise of creative labor. FIG. 1 is a schematic flowchart of a method for sending a message according to an embodiment of the present invention; FIG. 2 is a schematic flowchart of a method for sending a message according to another embodiment of the present invention;

FIG. 3 is a schematic structural diagram of an apparatus for sending a message according to an embodiment of the present invention; FIG. 4 is a schematic structural diagram of an apparatus for sending a message according to another embodiment of the present invention;

FIG. 5 is a schematic structural diagram of an apparatus for transmitting a message according to another embodiment of the present invention.

detailed description

The technical solutions in the embodiments of the present invention are clearly and completely described in the following with reference to the accompanying drawings in the embodiments of the present invention. It is obvious that the described embodiments are only a part of the embodiments of the present invention, but not all of the embodiments. General techniques in the art based on embodiments in the present invention All other embodiments obtained by a person without creative efforts are within the scope of the present invention.

The token bucket algorithm is one of the most commonly used algorithms in traffic shaping and rate limiting (Rate Limiting) in the communications field. Typically, the token bucket algorithm is used to control the amount of data sent to the network and to allow the transmission of bursty data. Therefore, some methods to limit network traffic need to be taken to avoid the problems that burst traffic brings to the network. The invention is based on a network processor or related software (for example, a programmable device such as a Field-Programmable Gate Array (FPGA)), and a method for delay processing by controlling a token bucket in a message processing process The method for sending a message, which is provided by the present invention, as shown in FIG. 1 , includes the following processes:

101. Receive a packet, and calculate a number of tokens TL required to send the packet according to the length of the packet. The number of tokens to be consumed in the current packet corresponds to the length of the text body. That is, the number of tokens to be consumed corresponding to the length L of the packet is T _L , where the number of tokens corresponding to the length of the packet is exactly an integer. The number of tokens required to obtain a packet is TL. When the number of tokens required for the packet length is a decimal number, the number of tokens required to be received is rounded up to obtain the required number of tokens. To ensure that the number of tokens that meet the minimum requirements for message delivery can be provided during the subsequent token filling process.

102. Obtain the number of currently available tokens in the token bucket! ^. Specifically, the number of tokens T _C remaining in the token bucket of the previous packet is obtained in the token bucket by querying the token information table. Here the token information is represented as shown in Table 1:

Table 1

The token information table contains the token flag T _P . _s , the number of tokens T _E and the padding time T _FLLL ; where T _E represents the number of tokens remaining in the token bucket of the previous message, and T _FLLL represents the time of transmitting the last message. As shown in Table 1, the token flag T _{P is indicated} . _s is positive. The number of tokens remaining in the token bucket of the previous message is T _C. The time at which the last message is sent is T _FLLL is 50 seconds. Of course, the unit of T _FLLL is only in seconds. To illustrate, in this embodiment, microseconds, milliseconds, or The number of cycles in the FPGA system is in units. The step 102 specifically includes:

102a, obtaining the number of tokens T _C and the filling time T _FLLL from the token information table;

Where T _C indicates the number of tokens remaining in the token bucket of the previous message, and T _FLLL indicates the time when the last packet was sent.

102b, the calculation time of the received packets with time T _n T _fill filling time interval dT; Specifically, the current token flag Tp. _{When s} is positive ("1"), the current filling time τ _η can be understood as the current message arrival time T _a „, where T _a „ is the current message arrival time, τ _η is the current filling time; The end time of the supplementary token can be understood as the time when the number of tokens required for the previous message is filled, if it is represented by T _fill ; if the time interval is represented by dT, the time interval is:

dT= T _n -T _fill .

102c, multiply the time interval dT by the filling rate V of the token, and then round down, and add the number of tokens T _c to obtain the currently available number of tokens T _add , that is, D ^ LdT * Vj.

The number of tokens filled in the time interval is: dT is multiplied by the filling rate V, and the original number of tokens in the token bucket is T _C (that is, the last message recorded in the token information table) The number of tokens TJ updated after transmission, then the current number of available tokens T _ADD is:

T _add = T _c + LdT * VJ ; where the product of dT * V may not be an integer. If it is not an integer, you can use rounding down (for example, 3.8 to take an integer down to 3) to calculate the D.

103. If the currently available number of tokens T _{ADD is} smaller than the number of tokens TL required to send the message, calculate the delay time T _dela y and start timing. The delay time is: T _delay = ( T _L - T _add ) / V , where V represents the filling rate of the token. The delay time here can be carried in decimals. For example, when the unit of calculating the delay time is calculated in seconds or seconds, the delay time is also a decimal, but when using FPGA as the computing environment, Since in the computing environment of the FPGA, the time value is generally in the form of a pulse, that is, the number of time periods is calculated. When an incomplete time period occurs, it is sometimes difficult to implement in the FP GA computing environment. So in the FPGA computing loop The delay time in the environment generally adopts the upward rounding strategy, and the upward rounding strategy ensures that the minimum number of tokens can be provided for the transmission of the message within the delay time. Specifically, the delay time in the FPGA computing environment The calculation formula can be adjusted to: T _DELAY = "( T _L - T _ADD ) / V 1 , where V represents the filling rate of the token, and "1 represents the rounding up.

104. When the delay time ends, the message is sent. The method for sending a message provided by the embodiment of the present invention can enable the release of the message after the token is filled in at least one time in the delay time period, and the message is prevented from being filled after being filled for a plurality of times. The case of allowing the passage occurs, so that the delay processing can reduce the number of calculations of the token requirement during the message transmission.

Specifically, as shown in FIG. 2, the method for sending a message provided by the embodiment of the present invention includes the following process:

201. Receive a packet, and calculate a number of tokens TL required to send the packet according to the length of the packet. The number of tokens that need to be consumed in the current packet corresponds to the length of the text body. That is, the number of tokens that need to be consumed corresponding to the length L of the packet is T _L . Here, when the number of tokens required for the packet is an integer, the number of tokens required to obtain the packet is TL; when the number of tokens required for the packet is a decimal, in order to meet the transmission requirement of the packet, Round up to get the required number of tokens "TL1. Therefore, the number of tokens required for the message can be obtained by rounding up.

202. Obtain a token flag from the token information table. When the token flag is positive, perform steps.

203.

203. Obtain a token number T _C and a padding time T _FLLL from the token information table, where T _C indicates that the number of tokens remaining in the token bucket of the previous message is sent, and T _FLLL indicates that the last packet is sent. Specifically, the number of currently available tokens in the token bucket can be obtained by querying the token information table; where the token information is represented by, for example, Table 2:

Table 2 The token information table contains the token flag T _P . _s , number of tokens T _E and filling time T As shown in Table 2, the token flag T _{p is} represented. _s is positive, sending the packet on a token remaining in the token bucket number T _c is 100, the time to transmit a complete packet _T flll 50 seconds, of course, where _T flll unit only seconds Example To illustrate, in this embodiment, microseconds, milliseconds, or the number of cycles in the FPGA system can also be used.

204, the received message is calculated time T _n and the filling time interval dT T _fill;

Specifically, the current token flag Tp. _s is positive ( "1"), the current fill time T _n will be appreciated that with the current packet arrival time T _a "- induced, where T _a" packet arrival time of the current, the current fill time T _n; Last The end time of the supplementary token can be understood as the time when the number of tokens required for the previous message is filled, if it is represented by T _fill ; if the time interval is represented by dT, the time interval is:

dT= T _n -Tim„

205. Multiply the time interval dT by the filling rate V of the token, and then round down, and add the number of tokens T _c to obtain the currently available number of tokens T _add , that is, T _add =T _c + Ld T*Vj .

The number of tokens filled in the time interval is obtained by multiplying the number of tokens by the filling rate V, and the number of original tokens in the token bucket is T _c (that is, the last packet transmission recorded in the token information table) After the number of tokens updated T _c ) , the current number of available tokens T _add is:

T _add =T _c + LdT *Vj ; where the product of dT * V may not be an integer, so it can be calculated by rounding down (for example, 3.8 is down to an integer of 3).

206. If the currently available number of tokens T _{ADD is} less than the number of tokens TL required to send the message, calculate the delay time T _dela y and start timing. The delay time is: T _delay = ( T _L - T _add ) / V, where V represents the filling rate of the token. The delay time may not be rounded. In actual situations, the delay time may also be a decimal number. The reason for calculating the delay time is as an example in the computing environment of the embodiment of the present invention. For seconds, seconds, the delay time is also decimal and can be calculated. However, in a computing environment such as FPGA, the delay time will be in the form of a pulse. If the decimal is taken, it may be difficult to implement, and when the delay time is rounded, the application is rounded up. The mode is an example, but it is not limited to the implementation that only rounds up. The delay time is: T _delay = "(T _L - T _add ) / V l , where V represents the filling rate of the token, and "1 represents rounding up to ensure that there are enough tokens to send the current message. .

207. Modify the token flag to be negative to indicate that the subsequent message is not processed.

If the current number of available tokens T _ADD < T _L , T _P will be. _{s is} set to "0" (previously mentioned, 0 means negative).

208. When the delay time ends, the message is sent.

209. Modify the token flag to be positive to indicate that the subsequent message is processed.

The delay time refers to adding the remaining number of tokens to the minimum time for satisfying the token consumption. When the delay time end token number is added to meet the minimum time for the token consumption, the token information in the token information table is The token flag is set to "1", where 1 is positive.

210. Update the padding time T _{FLLL of the} token information table by T _N +T _dela y to LT _ADD + V*

T _DELAY -T _L j updates the token number T _{c of the} token information table, U indicates rounding down.

Update the token table information: T _fll corpse T _fm +dT + T _delay ; that is T _fll corpse T _n + T _delay ;

T _c = LT _a dd + V* Tdelay-T _L J ;

T = 1

i pos 丄

In the updated token table, T _FLLL indicates the time at which the message is sent; T _C indicates the number of tokens remaining in the token bucket after the message is sent; T _P . _S indicates that the updated token flag of the token information table is positive, indicating that the subsequent four messages can be processed. Here, not limited in the sequence of steps 210, the present invention can be acquired at the end of step 206 for updating T _ADD filling time _{_{_{T fll T N + T dela y}}} , and the number of tokens T _C token update information table +V *T _dela y - T _L , so step 210 can be performed at any step between 206 and the time the message is sent. Optionally, the method further includes:

21 1. If the currently available number of tokens T _{ADD is} greater than or equal to the number of tokens TL required to send a message, a message is sent.

212. Update the padding time T _{FLLL of the} token information table at the time T _{N of} receiving the message, and update the token number _{Tc of the} token information table by T _ADD _-TL . Specifically: If the currently available number of tokens T _{ADD is} greater than or equal to the number of tokens TL consumed by the message, the remaining number of tokens T _c is calculated, and the padding time T _fl11 in the token information table is updated to release the message.

The remaining number of tokens T _c is:

T _c =T _add - min((T _add -L):H), that is, T _add -L and the token bucket depth H take the minimum value. Here, H is the depth of the token bucket, indicating that the currently available number of tokens T _{add is} greater than or equal to the number of tokens consumed by the packet. The depth H of the token bucket can be set according to the amount of burst data tolerated. When the token bucket depth H is set to be larger, the packet rate that is passed is closer to the set rate, so that the amount of burst data that the token bucket can tolerate is larger, that is, the number of tokens in the token bucket can fully satisfy the report. The request for sending packets, and thus the rate of filling the tokens (so that the traffic of the packets can only be less than or equal to the filling rate of the tokens), achieves the purpose of limiting packet traffic.

T _flll =T _flll +dT (because the processing scheme here is determined by the value as an integer, so when dT*V is rounded, the updated fill time in the actual situation: T _flll =T _flll + ( LdT*V_| / V); wherein _T flll represents the left side of equation _T flll updated, i.e., the packet transmission time; _T flll RHS represents _T flll before update, i.e. sending the packets over a time Gen 4 .

T _p . _s = " 1 " (1 means the token bucket flag is positive, 0 means negative, the same below).

Specifically, for example: In an FPGA environment, the system clock elk is 10MHz, V = 1Mbps, T _fill = 100, T _arr = 205, then:

T _n = 205 dT = T _n - T _fill = \Q5 dT * V = 105 * {\ I \0M) * \M = 10.5 In this case, the number of tokens added will be decimal, but the number of tokens added It is in integer, that is, in the case of a decimal, the token bucket can theoretically only add the number of integer-bit tokens with a decimal value, but if only the number of tokens is rounded down, In the calculation, the loss of the number of tokens is formed, and the corresponding adjustment is not performed in time. After several consecutive message transmissions, the system timing and the number of tokens are disordered. Therefore, the Tn is also processed accordingly. The whole Tn=200 is taken down, and all the subsequent steps are calculated according to Τη=200, and the number of tokens thus calculated is also rounded down correspondingly. Obtained by step 204: the time interval dT is:

dT= T _n -T _fill , so the calculation method is updated as: , = dT + T _fill = L llll + _Tfill = 100 + 100 = 200,

among them,

, where dT is during the interval

Here, the fill time is in accordance with D. . Updated, but since the calculation process is performed under the same clock system, the clock period that is ignored by the original calculation value and the number of tokens after the decimal point due to the ignored clock cycle will be calculated when the next message arrives. The calculation process is compensated, so in the long run, the time is not wasted, and the token bucket token is not consumed. The method for sending a message provided by the embodiment of the present invention can enable the release of the message after the token is filled in at least one time in the delay time period, and the message is prevented from being filled after being filled for a plurality of times. The case of allowing the passage occurs, so that the delay processing can reduce the number of calculations of the token requirement during the message transmission.

An embodiment of the present invention provides a device 3 for transmitting a message. Referring to FIG. 3, the method includes: a token processor 31 and a delay unit 32. The token processor 31 is configured to receive a packet according to the packet. The length of the token is TL; and the number of currently available tokens T _ADD in the token bucket is obtained;

The delay device 32 is configured to calculate the delay time _Tdelay and start timing if the currently available number of tokens _{Tadd is} less than the number of tokens _TL required to send the message, and the delay time is: _Tdelay = (T _L - T _add ) / V , where V represents the fill rate of the token;

The token processor 31 is further configured to send a message when the delay time ends. The device for sending a message according to the embodiment of the present invention can enable the release of the message after the token is filled in the delay time period by using the delay processing, thereby avoiding the release of the message. The case that the message is allowed to pass after being filled for a plurality of times occurs, so that the delay processing can reduce the number of calculations of the token requirement during the message sending process.

Further, referring to FIG. 4, the token processor 31 further includes: a token information acquiring unit 31 1 , a timing unit 312, and a token calculating unit 313, where:

The token information obtaining unit 31 1 is configured to obtain the number of tokens T _c and the filling time T _fl11 from the token information table, where T _c represents the number of tokens remaining in the token bucket of the previous message, T _flll Indicates the time at which the last article was sent;

The timing unit 312 is configured to calculate a time interval dT between the time T _{n of the} received text and the filling time T _fill ; the token calculating unit 313 is configured to multiply the time interval dT by the filling rate V of the token, and then round down, Together with the number of tokens T _c , the current number of available tokens T _{add is obtained} , that is, D2, LdT *VJ. Optionally, as shown in FIG. 4, the token processor 31 is further configured to update the time T _{N of} receiving the message if the currently available number of tokens T _{ADD is} greater than or equal to the number of tokens TL required to send the message. The padding time T _{FLLL of the} token information table updates the token number T _{C of the} token information table with T _ADD _-TL and transmits the message.

The device provided by the embodiment of the present invention supports sending the packet in a normal processing manner, because the number of currently available tokens is greater than or equal to the number of tokens that are required to be sent by the packet. Optionally, the delay device 32 is further configured to obtain a token flag from the token information table, and when the token flag is positive, to calculate the delay time T _dela y.

Further, the token processor 31 is further configured to: if the currently available number of tokens T _{add is} smaller than the number of tokens T _L required to send the packet, modify the token flag to be negative to indicate that the subsequent packet is not processed; The token flag is modified to be positive at the end of the delay time to indicate that the subsequent message continues to be processed. Further, the token processor 31 is further configured to update the token time T _flll in the T _n +T _delay update token information table, and update the token of the token information table with LT _add + V *T _dela y - T _L j The number T _c , at this time, T _flll represents the time at which the message is sent, and T _c represents the number of remaining tokens in the token bucket after the message is sent. The device for sending a message, which is provided by the embodiment of the present invention, can enable the release of the message after the token is filled in at least one time in the delay period, so as to avoid the message being filled after multiple times. Allow the passage of the situation, so that it can be processed by delay It is enough to reduce the number of calculations of the token demand during the message sending process.

An embodiment of the present invention provides a device 5 for transmitting a message. Referring to FIG. 5, the device 5 for transmitting a message includes: at least one processor 51, a memory 52, a communication interface 53, and a bus 54, the at least one processing The device 51, the memory 52 and the communication interface 53 are connected by a bus 54 and complete communication with each other.

The bus 54 may be an Industry Standard Architecture (ISA) bus, a Peripheral Component (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus. Wait. The bus 54 can be divided into an address bus, a data bus, a control bus, and the like. For ease of representation, only one thick line is shown in Figure 5, but it does not mean that there is only one bus or one type of bus. among them:

Memory 52 is for storing executable program code, including computer operating instructions. The memory 52 may include a high speed RAM memory and may also include a non-volatile memory, for example at least one disk memory.

The processor 51 may be a Central Processing Unit (CPU), or an Application Specific Integrated Circuit (ASIC), or one or more configured to implement the embodiments of the present invention. integrated circuit. The communication interface 53 is mainly used to implement communication between devices in this embodiment. The processor 52 is configured to receive the packet TL by the communication interface 53 and calculate the number of tokens TL required to send the packet according to the length of the packet. The processor 52 is further configured to obtain the currently available token number T _{add in the} token bucket. The processor 52 is further configured to calculate the delay time T _dela y and start timing if the currently available number of tokens T _{add is} less than the number of tokens TL required to send the message, and the delay time is: T _DELAY = ( T _L - T _Add ) / V; The processor 52 is further configured to send a message through the communication interface 54 when the delay time ends, where V represents the filling rate of the token. The device for sending a message, which is provided by the embodiment of the present invention, can enable the release of the message after the token is filled in at least one time in the delay period, so as to avoid the message being filled after multiple times. The case of allowing the passage occurs, so that the delay processing can reduce the number of calculations of the token requirement during the message transmission.

Further, the processor 52 is further configured to obtain the token number T _c and the padding from the token information table. Time T _flll , where T _c indicates the number of tokens remaining in the token bucket of the previous message, T _flll indicates the time when the last message was sent; and is also used to calculate the time 1^ and the padding time of the received message. T _flll interval dT; is also used to multiply the time interval dT by the token filling rate V, and then _round down, plus the number of tokens T _c , to obtain the current available number of tokens T _add , ie D ^ L dT *VJ. Optionally, the processor 52 is further configured to: if the currently available number of tokens T _{add is} greater than or equal to the number of tokens TL required to send the message, update the padding time T of the token information table at the time T _{N of} receiving the message. _fm, in order to update the token T _add -Tj ^ token information table number T _c, and message transmission. The device provided by the embodiment of the present invention supports sending the packet in a normal processing manner, because the number of currently available tokens is greater than or equal to the number of tokens that are required to be sent by the packet.

Optionally, the processor 52 is further configured to obtain a token flag from the token information table, and when the token flag is positive, used to calculate the delay time _Tdelay .

The processor 52 is further configured to: if the currently available number of tokens T _{add is} less than the number of tokens T _L required to send the message, modify the token flag to be negative to indicate that the subsequent packet is not processed; At the end, the token flag is modified to be positive to indicate that the subsequent message continues to be processed. The processor 52 is further configured to update the padding time T _{flll of the} token information table with T _n +T _dela y, and update the token number T _{c of the} token information table by LT _add + V* T _delay -TjJ. T _flll indicates the time at which the packet is sent, and T _c indicates the number of tokens remaining in the token bucket after the packet is sent. The device for sending a message, which is provided by the embodiment of the present invention, can enable the release of the message after the token is filled in at least one time in the delay period, so as to avoid the message being filled after multiple times. The case of allowing the passage occurs, so that the delay processing can reduce the number of calculations of the token requirement during the message transmission.

The device for sending a message, which is provided in the foregoing embodiments, may be used to perform the method corresponding to the embodiment shown in FIG. 1 and FIG. 2, and the details already described in the embodiment shown in FIG. 1 and FIG. 2 are not described herein. .

A person skilled in the art can understand that all or part of the steps of implementing the above method embodiments may be completed by using hardware related to program instructions, and the foregoing program may be stored in a computer readable storage medium, and the program is executed when executed. The foregoing steps include the steps of the foregoing method embodiments; and the foregoing storage medium includes: a medium that can store program codes, such as a ROM, a RAM, a magnetic disk, or an optical disk. The above is only the specific embodiment of the present invention, but the scope of the present invention is not limited thereto, and any person skilled in the art can easily think of changes or substitutions within the technical scope of the present invention. It should be covered by the scope of the present invention. Therefore, the scope of the invention should be determined by the scope of the appended claims.

Claims

Claim

A method for sending a message, comprising: receiving a message, calculating a number of tokens TL required to send the message according to the length of the packet; acquiring a currently available token in the token bucket The number T _add ; if the currently available number of tokens T _{ADD is} smaller than the number of tokens TL required to send the message, calculate the delay time T _delay and start timing, the delay time is: D ₍₁ = ( ^ -丁_£1(1(1 )/¥ , where V represents the fill rate of the token;

When the delay time ends, the message is sent.

2. The method according to claim 1, wherein the obtaining the number of currently available tokens in the token bucket! ^, specifically comprising: obtaining the number of tokens T _c and the filling time T _flll from the token information table, where T _c represents the number of tokens remaining in the token bucket after the last _message is sent, and T _flll indicates that the transmission is completed. a time interval of the previous message; calculating a time interval dT of the time 1^ of receiving the message and the filling time T _flll ; multiplying the time interval dT by the filling rate V of the token, and then taking down In addition, the number of tokens T _{c is added} , and the currently available number of tokens T _{add is obtained} , that is, the number of tokens TdT*Vj.

The method according to claim 2, wherein the method further comprises: acquiring a token flag from the token information table, and performing the calculation delay when the token flag is positive The steps of time.

The method according to claim 3, wherein the method further comprises: if the currently available number of tokens T _{ADD is} smaller than a number of tokens TL required to send the message, the token is The flag is modified to be negative to indicate that the subsequent message is not processed; the token flag is modified to be positive at the end of the delay time to indicate that the subsequent message is processed.

The method according to any one of claims 2 to 4, wherein the method further comprises:

Updating the padding time T _flll in the token information table with T _n +T _{delay, and} updating the token number T _{e of} the token information table by LT _add + V* T _delay -TjJ, at this time, T _flll represents Send the message

The method according to claim 2, wherein the method further comprises: If the currently available number of tokens T _{add is} greater than or equal to the number of tokens T _L required to send the message, the padding time T _{flll of} the token information table is updated at the time T _n when the message is received. Take

T _ADD -TL updates the token number ^ of the token information table and transmits the message.

A device for transmitting a message, comprising: a token processor and a delay device, wherein: the token processor is configured to receive a message, and calculate and send according to the length of the packet The number of tokens required by the packet is TL; and the number of currently available tokens T _ADD in the token bucket is obtained;

The delay device is configured to calculate a delay time T _DELAY and start timing if the current available token number T _add obtained by the token processor is smaller than a token number TL required to send the packet, The delay time is: T _delay = ( T _L - T _add ) / V , where V represents the filling rate of the token; the token processor is further configured to send the location when the delay time ends. Said message.

8. The device of claim 7, the token processor further comprising:

The device according to claim 8, wherein the delay device is further configured to acquire a token flag from the token information table, and when the token flag is positive, The delay time _{Tdelay is} calculated.

The device according to claim 9, wherein the token processor is further configured to: if the currently available number of tokens T _{ADD is} smaller than a number of tokens TL required to send the packet, The token flag is modified to be negative to indicate that the subsequent message is not processed; the token flag is modified to be positive at the end of the delay time to indicate that the subsequent message is processed.

The device according to any one of claims 8 to 10, wherein the token processor is further configured to update a filling time T _flll in the token information table with T _n +T _delay , Take LT _add + v *T _delay -Tjj updates the token number T _{c of} the token information table, at which time T _flll indicates

The device according to claim 8, wherein the token processor is further configured to: if the currently available number of tokens T _{add is} greater than or equal to the number of tokens TL required to send the packet, _Then , the padding time T _{FLLL of} the token information table is updated at the time T _N when the message is received, the number of tokens T _{c of} the token information table is updated by T _add _-TL , and then the message is transmitted.

13. A device for transmitting a message, comprising: at least one processor, a memory, a communication interface, and a bus, wherein the at least one processor, the memory, and the communication interface are connected by the bus and complete communication with each other, among them:

The processor is configured to receive a packet by using the communication interface, and calculate a number of tokens TL required to send the packet according to the length of the packet; If the currently available number of tokens T _{ADD is} smaller than the number of tokens TL required to send the message, the delay time T _delay is calculated and started, and the delay time is: T _delay = ( T _L - T _add ) /V, where V represents the fill rate of the token;

The processor is further configured to send the message through the communication interface when the delay time ends.

The device according to claim 13, wherein the processor is further configured to obtain the number of tokens T _c and the filling time T _fl11 from the token information table, where T _c indicates that the last packet is sent in the token bucket. The remaining number of tokens, T _flll represents the time when the last message was sent; and is also used to calculate the time interval dT between the time T _{n of} receiving the message and the filling time T _fill ; The interval dT is multiplied by the filling rate V of the token, and then rounded down, and the number of tokens T _{c is added} to obtain the currently available number of tokens T _add , that is, T _add =T _c + L dT *VJ.

The device according to claim 14, wherein the processor is further configured to acquire a token flag from the token information table, and when the token flag is positive, used to calculate the Delay time T _delay .

The device according to claim 15, wherein the processor is further configured to: if the current available number of tokens T _{ADD is} smaller than a number of tokens TL required to send the message, The card flag is modified to be negative to indicate that the subsequent message is not processed; the token flag is modified to be positive at the end of the delay time to indicate that the subsequent message is processed.

17. Apparatus according to any one of claims 14 to 16 wherein said The processor is further configured to update the padding time T _flll in the token information table with T _n +T _{delay, and} update the token number T _{e of} the token information table by LT _add +v* T _delay -TjJ , T _flll indicates that the message is sent.

The device according to claim 14, wherein the processor is further configured to: if the currently available number of tokens T _{ADD is} greater than or equal to the number of tokens TL required to send the message, receive the message updating the time T _n is the time of filling the token information table _T flll, the number of tokens T _add -T _L updating the token information table T _e, and transmits the packet.