KR20040018785A - Method for Queue Management Using Queuing Threshold - Google Patents
Method for Queue Management Using Queuing Threshold Download PDFInfo
- Publication number
- KR20040018785A KR20040018785A KR1020020050779A KR20020050779A KR20040018785A KR 20040018785 A KR20040018785 A KR 20040018785A KR 1020020050779 A KR1020020050779 A KR 1020020050779A KR 20020050779 A KR20020050779 A KR 20020050779A KR 20040018785 A KR20040018785 A KR 20040018785A
- Authority
- KR
- South Korea
- Prior art keywords
- queue
- packet
- qth
- xest
- queuing
- Prior art date
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
- H04L47/32—Flow control; Congestion control by discarding or delaying data units, e.g. packets or frames
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L49/00—Packet switching elements
- H04L49/90—Buffering arrangements
- H04L49/9005—Buffering arrangements using dynamic buffer space allocation
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L49/00—Packet switching elements
- H04L49/90—Buffering arrangements
- H04L49/9084—Reactions to storage capacity overflow
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
Description
본 발명은 TCP/IP를 기반으로 하는 데이터 통신망에서 큐 관리 방법에 관한 것으로 특히, 큐잉 임계치(Queuing Threshold)를 사용하여 폭주(Congestion) 상태가 지속되는 경우 패킷의 큐잉 지연(Delay)을 일정하게 감소시키는 큐 관리 방법에 관한 것이다.The present invention relates to a queue management method in a data communication network based on TCP / IP. In particular, a queuing delay of a packet is constantly reduced when a congestion state is maintained by using a queuing threshold. It relates to a queue management method.
일반적으로, 상호 연결된 큐들로 이루어진 망이라 할 수 있는 인터넷은 폭발적인 가입자 증가로 인해 발생하는 패킷 손실과 큐잉 지연이라는 문제에 직면하고있다. 상기 손실된 패킷은 자원을 낭비하고, 큐잉 지연은 인터넷 서비스의 질을 저하시키는 요인으로 작용하기 때문에 이에 대한 방안이 요구된다.In general, the Internet, a network of interconnected queues, faces problems such as packet loss and queuing delays caused by explosive subscriber growth. Since the lost packet wastes resources and the queuing delay acts as a factor for degrading the quality of Internet service, a solution is required.
이러한 방안으로 논의되는 것에는 Droptail, RED, BLUE, REM, GREEN등의 큐 관리 방법이 있다. 이 중, GREEN은 새로운 AQM(Active Queue Management) 알고리즘으로서 폭주 상태를 현재의 큐 사이즈가 아닌 지수적으로 평균한 패킷 도착 속도(packet arrival rate)를 기반으로 측정한다.What is discussed in this way include queue management methods such as Droptail, RED, BLUE, REM, and GREEN. Among these, GREEN is a new AQM (Active Queue Management) algorithm that measures the congestion status based on the exponentially averaged packet arrival rate rather than the current queue size.
그리고, 상기 측정된 패킷 도착 속도(Estimated Packet Arrival Rate : 이하 'Xest'라고 칭함)에 따라 패킷 폐기 확률(Packet Drop Probability : 이하 'Pb'라고 칭함)을 결정함으로써 즉, Xest가 큐에 저장된 패킷이 전송되어 나가는 속도(Transmittion Packet Rate)보다 큰 경우(폭주 상태) Pb를 증가시키고, 작은 경우(정상 상태) Pb를 감소시킴으로써, 패킷 손실(loss)을 줄이는 동시에 큐잉 지연(delay)을 감소시킨다.Then, by determining the packet drop probability (hereinafter referred to as 'Pb') according to the measured packet arrival rate (hereinafter referred to as 'Xest'), that is, the packet stored in the queue is stored in the queue. By increasing Pb if it is larger than the Transmission Packet Rate (congestion state) and decreasing Pb when it is small (steady state), it reduces packet loss and reduces queuing delay.
다시 말하면, Pb를 이용하여 큐에 오버플로우(Overflow)가 발생하기 전에 패킷을 폐기함으로써 패킷 손실을 줄이는 동시에, 트래픽 정상 상태에서는 Pb를 감소시켜 큐 사이즈를 증가시킴으로써 큐의 버퍼링(Buffering) 효과를 유도하고, 폭주 상태에서는 Pb를 증가시켜 큐 사이즈를 감소시킴으로써 큐잉 지연을 감소시킨다.In other words, Pb can be used to discard packets before overflow occurs to the queue, reducing packet loss, while reducing the Pb and increasing the queue size in a steady state of traffic, leading to the buffering effect of the queue. In the congestion state, Pb is increased to decrease the queue size, thereby reducing the queuing delay.
이하, 첨부한 도면 도 1을 참조하여 GREEN에 의한 큐 관리 방법을 설명하면 다음과 같다.Hereinafter, a queue management method using GREEN will be described with reference to FIG. 1.
먼저, 큐 제어부(도면에 도시되지 않음)는 아래의 수학식에 따라 Xest를 산출하는데(S10), 상기 Xest는 패킷이 큐에 도착할 때마다 계산되어 업데이트된다.First, the queue control unit (not shown) calculates Xest according to the following equation (S10). The Xest is calculated and updated every time the packet arrives in the queue.
여기서, Del은 패킷간 지연 시간(Inter-packet Delay), B는 패킷 사이즈, K는 시간 상수이다.Here, Del is an inter-packet delay, B is a packet size, and K is a time constant.
큐 제어부는 상기 계산된 Xest를 가지고 Pb를 계산하는데, 상기 Pb는 ΔT 시간마다 계산되어 업데이트된다.The queue controller calculates Pb with the calculated Xest, which is calculated and updated every ΔT time.
즉, Xest가 Ct(Target Link Capacity) 이상인지 여부를 확인하여(S11), Ct 이상인 경우 직전 Pb(이하 'Pb_old'라고 칭함)를 ΔPb만큼 증가시키고(S12), Ct 미만인 경우 Pb_old를 ΔPb만큼 감소시킨다(S13).That is, it is determined whether Xest is greater than or equal to Ct (Target Link Capacity) (S11), and if it is greater than or equal to Ct, the previous Pb (hereinafter referred to as 'Pb_old') is increased by ΔPb (S12), and if it is less than Ct, Pb_old is decreased by ΔPb. (S13).
상기 ΔT 및 ΔPb는 네트웍 상황을 고려한 상수이다.ΔT and ΔPb are constants in consideration of network conditions.
여기서, Ct는 큐 사이즈가 '0'으로 수렴하도록 하기 위해 C(Actual Link Capacity)보다 작은 값으로 셋팅되는데, 통상 C에 '0.97'을 곱한 값으로 셋팅된다.Here, Ct is set to a value smaller than C (Actual Link Capacity) so that the queue size converges to '0', and is usually set to C multiplied by '0.97'.
한편, Pb는 '0 ~ 1'사이의 값을 가지는 확률이므로 상기 계산된 Pb가 '0'미만인 경우 또는 '1'을 초과하는 경우에는 일률적으로 각각 '0' 또는 '1'로 대체한다.Meanwhile, since Pb is a probability having a value between '0' and '1', when the calculated Pb is less than '0' or exceeds '1', it is replaced with '0' or '1', respectively.
상기 계산된 Pb를 통해 큐 제어부는 패킷을 큐에 저장할 것인지 또는 폐기(drop)할 것인지 여부를 결정한다.Through the calculated Pb, the queue controller determines whether to store or drop the packet in the queue.
즉, 상기 계산된 Pb가 난수 발생기에서 발생한 '0 ~ 1'사이의 난수(Random Variable)인 R이상이고 '0'이 아닌 경우 패킷을 폐기하고(S14,S15), Pb가 R 미만이거나 '0'인 경우 패킷을 큐에 저장한다(S14,S16).That is, if the calculated Pb is equal to or greater than R, which is a random variable between '0' and '1' generated by the random number generator, and is not '0', the packet is discarded (S14, S15), and Pb is less than R or '0'. ', The packet is stored in the queue (S14, S16).
상기와 같이, GREEN은 Ct(target link capacity)를 C(actual link capacity)보다 낮게 셋팅하여 큐 사이즈를 '0'으로 수렴하게 하고, Xest에 따라 Pb를 결정하여, 결정된 Pb에 따라 패킷을 폐기함으로써, 큐 사이즈 즉, 큐에 쌓이는 패킷을 제한하여 큐잉 지연을 감소시킨다.As described above, GREEN sets the target link capacity (Ct) lower than the actual link capacity (C) to converge the queue size to '0', determines Pb according to Xest, and discards the packet according to the determined Pb. Therefore, the queue size, that is, the packets accumulated in the queue are limited to reduce the queuing delay.
그러나, Ct(target link capacity)를 C(actual link capacity)보다 낮게 셋팅하는 것은 링크 사용율을 감소시키는 원인이 될 뿐만 아니라 Xest가 C보다 낮게 셋팅된 Ct(C*0.97)를 기준으로 급변하는 경우에는 큐에 저장된 패킷이 전송되어 나가는 속도보다 큐에 도착하는 패킷의 평균값이 더 크기 때문에 실제 큐 사이즈가 '0'으로 수렴하지 않게 된다.However, setting the target link capacity (Ct) lower than the actual link capacity (C) not only causes a decrease in link utilization, but also causes a sudden change based on the Ct (C * 0.97) set lower than C. Because the average value of packets arriving in the queue is larger than the rate at which the queued packets are sent out, the actual queue size does not converge to '0'.
예를 들면, Ct를 C의 97%를 사용했을때 Xest가 다음과 같이 96% 110% 96% 110%...로 급변하는 경우 초기 Pb가 0이고 ΔPb가 0.01이라고 했을때 Pb는 다음과 같이 0 0.01 0 0.01...이 된다. 이 경우, 패킷이 큐에 들어가는 속도(Xest-Xest*Pb)는 96% 108.9% 96% 108.9%...가 되고 이들의 평균은 100%보다 크게 되어 큐 사이즈는 계속 증가하게 된다.For example, if Ct uses 97% of C and Xest rapidly changes to 96% 110% 96% 110% ... as follows, when Pb is 0 and ΔPb is 0.01, Pb is 0 0.01 0 0.01. In this case, the rate at which packets enter the queue (Xest-Xest * Pb) is 96% 108.9% 96% 108.9% ... and their average is greater than 100% so that the queue size continues to increase.
전술한 바와 같이, 종래 Ct를 C보다 낮게 셋팅하여 큐 사이즈를 감소시키는 GREEN 방법에 의하면 링크 사용율이 떨어지고, Xest가 급변하는 상황에서 큐 사이즈가 증가하는 문제점이 있었다.As described above, according to the GREEN method of reducing the queue size by setting Ct lower than C, there is a problem in that the queue utilization decreases and the queue size increases in a situation where Xest rapidly changes.
본 발명은 전술한 바와 같은 문제점을 해결하기 위한 것으로 그 목적은, 큐잉 임계치(Queuing Threshold)를 도입하여 패킷을 큐에 저장할 때 이를 이용함으로써, 타겟 링크 용량을 실제 링크 용량과 같게 설정하여 링크 사용율을 높이고 폭주가 지속되는 상황에서 큐 사이즈를 폭주시 큐잉 임계치 이하로 제한하여 큐잉 지연을 감소시키는데 있다.SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problems, and an object thereof is to introduce a queuing threshold and use it when storing a packet in a queue, thereby setting the link utilization by setting the target link capacity to be equal to the actual link capacity. It is to reduce the queuing delay by limiting the queue size below the queuing threshold during congestion in a situation where the congestion continues to increase.
도 1은 종래 GREEN에 의한 큐 관리 동작을 설명하기 위한 플로우챠트.1 is a flowchart for explaining a queue management operation by a conventional GREEN.
도 2는 본 발명에 따른 큐 관리 동작을 설명하기 위한 플로우챠트.2 is a flowchart illustrating a queue management operation according to the present invention.
도 3은 패킷 도착 속도와 큐잉 임계치의 관계를 도시한 도.3 illustrates the relationship between packet arrival rate and queuing threshold.
도 4는 큐잉 임계치와 큐 사이즈의 관계를 설명하기 위한 도.4 is a diagram for explaining a relationship between a queuing threshold and a queue size.
상술한 바와 같은 목적을 달성하기 위한 본 발명에 따른 큐 관리 방법은, 큐에 도착하는 패킷의 속도를 지수적으로 평균한 패킷 도착 속도(Xest)를 산출하는 과정과; 상기 패킷 도착 속도를 실제 링크 용량(C)과 동일하게 셋팅된 타겟 링크 용량(Ct)과 비교하여 패킷 폐기 확률(Pb)과 큐잉 임계치(Qth)를 산출하는 과정과; 상기 패킷 폐기 확률(Pb)과 큐잉 임계치(Qth)에 따라 큐에 도착하는 패킷을 폐기하는 과정을 포함하여 이루어지는 것을 특징으로 한다.The queue management method according to the present invention for achieving the above object comprises the steps of: calculating a packet arrival rate (Xest) that is an exponential average of the speed of packets arriving in the queue; Calculating a packet drop probability (Pb) and a queuing threshold (Qth) by comparing the packet arrival rate with a target link capacity (Ct) set equal to an actual link capacity (C); And discarding packets arriving in the queue according to the packet discard probability Pb and the queuing threshold Qth.
그리고, 상기 패킷의 폐기는 상기 패킷 폐기 확률(Pb)이 난수 발생기에서 발생된 난수 이상이거나 현재 큐 사이즈가 상기 큐잉 임계치(Qth)를 초과하는 경우 도착하는 패킷을 폐기하는 것을 특징으로 한다.The discarding of the packet may include discarding a packet arriving when the packet discard probability Pb is equal to or greater than a random number generated by a random number generator or when the current queue size exceeds the queuing threshold Qth.
또한, 상기 큐잉 임계치(Qth)는 패킷 도착 속도에 따라 폭주시 큐에 저장될 수 있는 패킷의 최대값인 폭주시 큐잉 임계치(Qth_C)와 물리적 큐 사이즈 사이의 값을 가지는 것을 특징으로 한다.In addition, the queuing threshold Qth is characterized by having a value between the congestion queuing threshold Qth_C, which is the maximum value of packets that can be stored in the queue upon congestion, and the physical queue size.
나아가, 상기 큐잉 임계치(Qth)는 상기 패킷 도착 속도(Xest)가 타겟 링크용량(Ct) 이상인 경우 폭주시 큐잉 임계치(Qth_C)이고, 상기 패킷 도착 속도(Xest)가 타겟 링크 용량(Ct) 미만인 경우 수학식 'Qth = Qsize - (Qsize - Qth_C) * Xest/Ct'에 의해 계산되는 것을 특징으로 한다.Furthermore, the queuing threshold Qth is a queuing threshold Qth_C when the packet arrival rate Xest is greater than or equal to the target link capacity Ct, and the packet arrival rate Xest is less than the target link capacity Ct. It is characterized in that it is calculated by the equation 'Qth = Qsize-(Qsize-Qth_C) * Xest / Ct'.
본 발명은 종래 GREEN 방법을 개선하여 타겟 링크 용량을 실제 링크 용량과 동일하게 셋팅하여 링크 사용율을 높이되, 큐잉 지연을 감소시키기 위해 큐잉 임계치를 도입하여 큐에 도착하는 패킷의 폐기 여부를 결정시 패킷 폐기 확률과 함께 큐잉 임계치를 고려하게 한 것이다.The present invention improves the conventional GREEN method to increase the link utilization by setting the target link capacity to be the same as the actual link capacity, but introduces a queuing threshold to reduce the queuing delay and determines whether to discard packets arriving at the queue. The queuing threshold is considered along with the discard probability.
이하, 본 발명에 따른 실시예를 첨부한 도면 도 2를 참조하여 상세하게 설명하면 다음과 같다.Hereinafter, with reference to the accompanying drawings an embodiment according to the present invention will be described in detail as follows.
큐 제어부(도면에 도시되지 않음)는 패킷이 큐에 도착할 때마다 Xest(Estimated Data Arrival Rate)를 산출하는데(S20), 그 계산 방법은 종래와 동일하므로 생략한다.The queue control unit (not shown) calculates an Estimated Data Arrival Rate (Xest) each time a packet arrives at the queue (S20), and the calculation method is the same as in the related art, and thus will be omitted.
큐 제어부는 상기 계산된 Xest를 기초로 ΔT 시간마다 Pb(Packet Drop Probability)와 Qth를 산출한다.The queue controller calculates Pb (Packet Drop Probability) and Qth every ΔT time based on the calculated Xest.
즉, Xest가 Ct(Target Link Capacity) 이상인지 여부를 확인하여(S21), Xest가 Ct 이상인 경우 Pb는 직전 Pb(이하 'Pb_old'라고 칭함)를 ΔPb만큼 증가시킨 값이 되고, 큐잉 임계치는 폭주시 큐잉 임계치(Congested_Q_Threshold : 이하 'Qth_C'라고 칭함)가 된다(S22).That is, it is determined whether Xest is greater than or equal to Ct (Target Link Capacity) (S21), and when Xest is greater than or equal to Ct, Pb is a value obtained by increasing the previous Pb (hereinafter referred to as 'Pb_old') by ΔPb, and the queuing threshold is congested. A time queuing threshold (Congested_Q_Threshold: hereinafter referred to as 'Qth_C') is obtained (S22).
그리고, Xest가 Ct 미만인 경우 Pb는 Pb_old를 ΔPb만큼 감소시킨 값이 되고, 큐잉 임계치는 아래의 수학식에 의해 산출된다(S23).When Xest is less than Ct, Pb is a value obtained by decreasing Pb_old by ΔPb, and the queuing threshold is calculated by the following equation (S23).
여기서, Qsize는 물리적인 큐 사이즈를 나타내고, Ct는 링크 사용율을 높이기 위해 실제 링크 용량 C(Actual Link Capacity)와 동일한 값으로 셋팅된다.Here, Qsize represents a physical queue size, and Ct is set to the same value as the actual link capacity C (Actual Link Capacity) in order to increase link utilization.
한편, Pb는 '0 ~ 1'사이의 값을 가지는 확률이므로 종래와 마찬가지로 상기 계산된 Pb가 '0'미만인 경우 또는 '1'을 초과하는 경우에는 일률적으로 각각 '0' 또는 '1'로 대체한다.On the other hand, Pb is a probability having a value between '0 and 1', so when the calculated Pb is less than '0' or exceeds '1' as in the prior art, it is replaced by '0' or '1' uniformly, respectively. do.
상기 Qth는 첨부한 도면 도 3에 도시된 바와 같이 Xest에 따라 Qth_C와 물리적인 큐 사이즈 사이에서 변하는 값으로서, Xest가 Ct이상인 폭주 상태에서 Qth_C로 유지된다.Qth is a value that varies between Qth_C and the physical queue size according to Xest as shown in FIG. 3, and is maintained at Qth_C in a runaway state where Xest is greater than or equal to Ct.
상기 계산된 Pb와 Qth를 통해 큐 제어부는 패킷을 큐에 저장할 것인지 또는 폐기(drop)할 것인지 여부를 결정한다.Through the calculated Pb and Qth, the queue controller determines whether to store or drop the packet in the queue.
즉, 상기 계산된 Pb가 난수 발생기에서 발생한 '0 ~ 1'사이의 난수(Random Variable)인 R이상이고 '0'이 아닌 경우 또는 현재의 큐 사이즈가 Qth를 초과하는 경우 도착하는 패킷을 폐기하고(S24,S25), 그 외의 경우 패킷을 큐에 저장한다(S24,S26).That is, if the calculated Pb is equal to or greater than R which is a random variable between '0' and '1' generated in the random number generator and is not '0', or if the current queue size exceeds Qth, discarding the arriving packet. (S24, S25), otherwise, the packet is stored in the queue (S24, S26).
전술한 바와 같이, 본 발명에 의한 큐 관리 방법에 의하면 실제 링크 용량을 100% 활용하여 링크 사용율을 높이는 동시에 폭주시에는 큐 사이즈를 강제적으로제한함으로써, 큐잉 지연을 감소시킨다.As described above, the queue management method according to the present invention utilizes 100% of the actual link capacity to increase the link utilization rate and forcibly limits the queue size during congestion, thereby reducing the queuing delay.
본 발명에 따라 큐잉 임계치가 정상 상태에서 폭주시 큐잉 임계치와 물리적 큐 사이즈 사이에서 결정되고, 폭주 상태에서는 폭주시 큐잉 임계치로 결정되면, 큐 사이즈는 상기 결정된 큐잉 임계치 이하로 제한되는데, 첨부한 도면 도 4에 도시된 바와 같이 정상 상태에서는 ②~③ 사이의 값이 되고, 폭주 상태에서는 ①~②사이의 값이 된다.According to the present invention, if the queuing threshold is determined between the queuing threshold and the physical queue size at the time of congestion in the normal state, and the queuing threshold at the time of congestion in the congestion state, the queue size is limited to below the determined queuing threshold, As shown in Fig. 4, the value is between ② and ③ in the normal state, and between ① and ② in the runaway state.
또한, 본 발명에 따른 실시예는 상술한 것으로 한정되지 않고, 본 발명과 관련하여 통상의 지식을 가진자에게 자명한 범위내에서 여러 가지의 대안, 수정 및 변경하여 실시할 수 있다.In addition, the embodiments according to the present invention are not limited to the above-described embodiments, and various alternatives, modifications, and changes can be made within the scope apparent to those skilled in the art.
특히, 패킷 도착 속도의 지수적 평균값에 따라 변하는 큐잉 임계치를 GREEN 이외의 다른 큐 관리 스킴(Scheme)에 도입함으로써, 폭주 상태에서 큐 사이즈를 폭주시 큐잉 임계치 이하로 유지하여 큐잉 지연을 일정 수준이하로 제한할 수 있다.In particular, by introducing a queuing threshold that changes according to the exponential average of packet arrival rates into a queue management scheme other than GREEN, the queue size is kept below the queuing threshold during congestion in a congested state, and the queuing delay is kept below a certain level. You can limit it.
이상과 같이, 본 발명은 큐잉 임계치를 도입하여 패킷 폐기시 이를 이용함으로써, 타겟 링크 용량을 실제 링크 용량까지 끌어올려 링크 사용율을 증가시키고, 폭주 상태에서 큐에 쌓이는 패킷을 폭주시 큐잉 임계치 이하로 제한하여 큐잉 지연을 감소시키는 효과가 있다.As described above, the present invention increases the link utilization by increasing the target link capacity to the actual link capacity by introducing a queuing threshold and using the packet drop, and limiting the packets accumulated in the queue in the congestion state to be below the queuing threshold during congestion. This reduces the queuing delay.
Claims (5)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| KR1020020050779A KR20040018785A (en) | 2002-08-27 | 2002-08-27 | Method for Queue Management Using Queuing Threshold |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| KR1020020050779A KR20040018785A (en) | 2002-08-27 | 2002-08-27 | Method for Queue Management Using Queuing Threshold |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| KR20040018785A true KR20040018785A (en) | 2004-03-04 |
Family
ID=37323993
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| KR1020020050779A KR20040018785A (en) | 2002-08-27 | 2002-08-27 | Method for Queue Management Using Queuing Threshold |
Country Status (1)
| Country | Link |
|---|---|
| KR (1) | KR20040018785A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2008083578A1 (en) * | 2006-12-27 | 2008-07-17 | Huawei Technologies Co., Ltd. | A service assembly method and system |
| CN107347039A (en) * | 2016-05-05 | 2017-11-14 | 深圳市中兴微电子技术有限公司 | A kind of management method and device in shared buffer memory space |
-
2002
- 2002-08-27 KR KR1020020050779A patent/KR20040018785A/en not_active Application Discontinuation
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2008083578A1 (en) * | 2006-12-27 | 2008-07-17 | Huawei Technologies Co., Ltd. | A service assembly method and system |
| CN107347039A (en) * | 2016-05-05 | 2017-11-14 | 深圳市中兴微电子技术有限公司 | A kind of management method and device in shared buffer memory space |
| CN107347039B (en) * | 2016-05-05 | 2020-02-21 | 深圳市中兴微电子技术有限公司 | Management method and device for shared cache space |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11070481B2 (en) | Predictive management of a network buffer | |
| US7069356B2 (en) | Method of controlling a queue buffer by performing congestion notification and automatically adapting a threshold value | |
| US7447152B2 (en) | Controlling traffic congestion | |
| KR101075724B1 (en) | Apparatus and method for limiting packet transmission rate in a communication system | |
| US20030223368A1 (en) | Retro flow control for arriving traffic in computer networks | |
| CN110138678B (en) | Data transmission control method and device, network transmission equipment and storage medium | |
| CN114938350A (en) | Congestion feedback-based data flow transmission control method in lossless network of data center | |
| US7756028B2 (en) | Pulsed backpressure mechanism for reduced FIFO utilization | |
| CN112995048B (en) | Blocking control and scheduling fusion method of data center network and terminal equipment | |
| US20050185581A1 (en) | Active flow management with hysteresis | |
| CN107770082B (en) | Transmission control method based on task flow characteristics in data center network | |
| Tang et al. | Adaptive queue management based on the change trend of queue size | |
| JP4838739B2 (en) | Router buffer management method and router using the management method | |
| KR20040018785A (en) | Method for Queue Management Using Queuing Threshold | |
| Irawan et al. | Performance evaluation of queue algorithms for video-on-demand application | |
| Almomani et al. | Simulation Based Performance Evaluation of Several Active Queue Management Algorithms for Computer Network | |
| Patel et al. | A new active queue management algorithm: Altdrop | |
| Singh et al. | Comparing different active queue management techniques | |
| Al-Raddady et al. | A new adaptive congestion control mechanism for the Internet based on RED | |
| Abdel-Jaber et al. | Evaluation of the Performance for IM-RED and IGRED Algorithms using Discrete-time Queues | |
| Kadhum et al. | The impact of the average packet arrival rate on the drop/mark probability of the Fast Congestion Notification (FN) mechanism | |
| Sivakumar et al. | Convex Optimized Lagrange Multiplier Based Algebraic Congestion Likelihood for Improved TCP Performance in MANET | |
| Li et al. | Adaptive rate control for TCP Incast based on selective ECN-marking | |
| Kadhum et al. | The drop activation function of the fast congestion notification (FN) mechanism | |
| Yaghmaee et al. | A New Fuzzy Logic Approach for TCP Congestion Control |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| N231 | Notification of change of applicant | ||
| WITN | Application deemed withdrawn, e.g. because no request for examination was filed or no examination fee was paid |