CN103618682A - Low power frequency modulation method based on traffic - Google Patents

Low power frequency modulation method based on traffic Download PDF

Info

Publication number
CN103618682A
CN103618682A CN201310485617.2A CN201310485617A CN103618682A CN 103618682 A CN103618682 A CN 103618682A CN 201310485617 A CN201310485617 A CN 201310485617A CN 103618682 A CN103618682 A CN 103618682A
Authority
CN
China
Prior art keywords
frequency
flow
traffic
frequency modulation
module
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN201310485617.2A
Other languages
Chinese (zh)
Other versions
CN103618682B (en
Inventor
周磊
李亮华
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Jiaxinghuabing Internet Of Things Technology Co Ltd
Original Assignee
Hangzhou Dianzi University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Hangzhou Dianzi University filed Critical Hangzhou Dianzi University
Priority to CN201310485617.2A priority Critical patent/CN103618682B/en
Publication of CN103618682A publication Critical patent/CN103618682A/en
Application granted granted Critical
Publication of CN103618682B publication Critical patent/CN103618682B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Landscapes

  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

本发明一种基于流量的低功耗调频方法，本发明通过计算网络数据流的大小来进行调频的。通过在特定时间内计算出网络设备接收到的数据包数，根据接收到的数据包数来反应当前网络中的流量，然后把这个流量作为选择各个频率的指标，从而实现基于流量的自适应调频。本发明可以实现自适应调频；避免了频率的振荡，调频灵敏度可调；降低模块功耗，节能。

Figure 201310485617

The invention discloses a traffic-based low-power frequency modulation method. The invention performs frequency modulation by calculating the size of network data flow. By calculating the number of data packets received by the network device within a specific time, the current traffic in the network is reflected according to the number of received data packets, and then this traffic is used as an indicator for selecting each frequency, so as to realize traffic-based adaptive frequency modulation . The invention can realize self-adaptive frequency modulation, avoid frequency oscillation, adjust frequency modulation sensitivity, reduce module power consumption and save energy.

Figure 201310485617

Description

一种基于流量的低功耗调频方法A flow-based low-power frequency modulation method

技术领域   technical field

本发明涉及通信领域的电源节能技术，特别是涉及通信设备的一种节能调频控制方法。     The invention relates to a power supply energy-saving technology in the communication field, in particular to an energy-saving frequency modulation control method for communication equipment. 

背景技术   Background technique

据相关研究组织调查表明：在以太网有线通信技术中，全球用于局域网中的网络接口设备有30亿个；据估计，欧洲互联网提供商（ISP）在2010年的全部网络电力功耗约为214亿千瓦时，如果不采用绿色节能技术，到2020年，这个消耗量预测将大幅增加到约358亿千瓦时。因此，在当今推崇的生态网络建设中，能源有效利用变得越来越重要，针对网络设备的功耗研究和改善也显得十分必要。 According to the survey of relevant research organizations, in Ethernet wired communication technology, there are 3 billion network interface devices used in local area networks worldwide; it is estimated that the total network power consumption of European Internet providers (ISPs) in 2010 is about 21.4 billion kwh. If green energy-saving technologies are not adopted, by 2020, this consumption forecast will increase substantially to about 35.8 billion kwh. Therefore, in the construction of the ecological network that is respected today, the efficient use of energy is becoming more and more important, and the research and improvement of the power consumption of network equipment is also very necessary.

当前Internet骨干网的带宽平均利用率为30％，带宽最高平均利用率小于45％；但是，为了网络数据流顺畅地流通以及应对其突发状况的出现，当前的网络设备一般都是全天候满负荷运行，造成了电力的极大耗费。针对这一现象，网络设备提供商和科研工作者提出了两种基本方法以使网络设备运行在合适的工作状态；一种就是让网络设备能在运行和休眠之间相互切换，另一种就是根据某种调频机制实现网络设备工作频率的自适应调整。由于网络数据流的难以预测性，网络设备在休眠和运行状态下切换时可能会对网络性能造成极大的影响；然而有效的自适应调频机制却能很好地解决这一问题，从而实现性能和功耗上的完美结合。 The average bandwidth utilization rate of the current Internet backbone network is 30%, and the highest average utilization rate of bandwidth is less than 45%. operation, resulting in a great consumption of electricity. In response to this phenomenon, network equipment providers and researchers have proposed two basic methods to make network equipment run in a suitable working state; one is to allow network equipment to switch between running and dormancy, and the other is to According to a certain frequency modulation mechanism, the adaptive adjustment of the working frequency of network equipment is realized. Due to the unpredictability of network data flow, network devices may have a great impact on network performance when switching between dormant and running states; however, an effective adaptive frequency modulation mechanism can solve this problem well, thereby achieving performance The perfect combination of power consumption.

针对网络设备的频率自适应调整，国内外也有相关的研究，主要的两种方式是根据当前数据流的缓存量或者通过上层软件的配置来实现设备调频[，这两种方法理论上可以达到调频的效果，但是存在一定不足。首先，软件配置调频，由于上层软件不能快速、高效的实现频率的动态调整，而且由于涉及到软件和硬件的相互通信，实现相对复杂。基于缓存量的自适应调频方式也有一定的缺点：在特定的频率下，特定的缓存量大小是可以作为路由器工作状态的指标，但是如果在不同频率下，用特定的缓存量来指示当前工作状态是没有明显意义的，而且这种调频方式还会造成相邻的频率反复地切换，从而影响路由器的稳定，并且在切换过程中会有大量数据包的丢失和额外的功耗浪费，虽然一定的算法（双阈值或多阈值）可以一定程度上解决这个问题，但还是会对网络设备的稳定性造成一定的潜在威胁。 There are also related researches at home and abroad on the frequency adaptive adjustment of network equipment. The main two methods are to realize equipment frequency modulation according to the current data stream buffer size or through upper-layer software configuration [ ] . These two methods can theoretically achieve frequency modulation. effect, but there are certain deficiencies. First of all, software configuration frequency modulation, because the upper layer software cannot quickly and efficiently realize the dynamic frequency adjustment, and because it involves the mutual communication between software and hardware, the implementation is relatively complicated. The adaptive frequency modulation method based on the buffer size also has certain disadvantages: at a specific frequency, the specific buffer size can be used as an indicator of the router's working status, but if it is at a different frequency, the specific buffer size is used to indicate the current working status It has no obvious meaning, and this frequency modulation method will also cause repeated switching of adjacent frequencies, which will affect the stability of the router, and there will be a lot of data packet loss and extra power consumption during the switching process, although certain Algorithms (double-threshold or multi-threshold) can solve this problem to a certain extent, but they still pose certain potential threats to the stability of network devices.

发明内容  Invention content

本发明的目的就是为了有效地解决频率调整的振荡现象以及软件调频的繁琐、缓慢的缺点，提供了一种基于流量的自适应调频机制来达到降低网络设备功耗的方法。 The purpose of the present invention is to provide a traffic-based self-adaptive frequency modulation mechanism to reduce the power consumption of network equipment in order to effectively solve the oscillation phenomenon of frequency adjustment and the cumbersome and slow shortcomings of software frequency modulation.

本发明通过计算网络数据流的大小来进行调频的。通过在特定时间内（设定一个特定的时间门限值）计算出网络设备接收到的数据包数，根据接收到的数据包数来反应当前网络中的流量，然后把这个流量作为选择各个频率的指标，从而实现基于流量的自适应调频。 The present invention performs frequency modulation by calculating the size of network data flow. By calculating the number of data packets received by the network device within a specific time (setting a specific time threshold), the current traffic in the network is reflected according to the number of received data packets, and then this traffic is used as the choice of each frequency indicators, so as to realize traffic-based adaptive frequency regulation.

本发明解决其技术问题所采用的技术方案是： The technical scheme that the present invention solves its technical problem adopts is:

步骤一：对网络设备流量的准确计算： Step 1: Accurate calculation of network device traffic:

当网络数据流进入流量统计模块的输入队列时，统计模块将数据流划分成多个数据包，并且计算出每个数据包的字节数，通过一个计数模块统计设定时间内所接收到的总的字节数，可以调整统计时间的大小来调整本调频方法的灵敏性。 When the network data flow enters the input queue of the traffic statistics module, the statistics module divides the data flow into multiple data packets, and calculates the number of bytes of each data packet, and counts the data received within the set time through a counting module The total number of bytes, you can adjust the size of the statistical time to adjust the sensitivity of this frequency modulation method.

步骤二：多个频率的产生 Step 2: Generation of multiple frequencies

通过使用FPGA厂商提供的时钟 IP核，或通过硬件描述语言的方法来产生全局时钟频率，全局时钟的要求的特点：一、负载能力特别强, 任何一个全局时钟驱动线都可以驱动芯片内部的触发器; 二是时延差特别小; 三是时钟信号波形畸变小, 工作可靠性好。  The global clock frequency is generated by using the clock IP core provided by the FPGA manufacturer, or through the method of hardware description language. The required characteristics of the global clock: 1. The load capacity is particularly strong, and any global clock drive line can drive the trigger inside the chip. The second is that the delay difference is very small; the third is that the distortion of the clock signal waveform is small, and the working reliability is good. the

步骤三：调频机制 Step 3: FM Mechanism

流量t相对于前一时刻，无非是增加或者减少； Compared with the previous moment, the flow t is nothing more than an increase or decrease;

当t增加时，频率和流量值的关系式如下所示：  When t increases, the relationship between frequency and flow value is as follows:

Figure 2013104856172100002DEST_PATH_IMAGE002
Figure 2013104856172100002DEST_PATH_IMAGE002

当t减小时, 频率和流量值的关系式如下所示： When t decreases, the relationship between frequency and flow value is as follows:

其中， f为工作频率，

Figure 2013104856172100002DEST_PATH_IMAGE006
 为第个频率，且
Figure 2013104856172100002DEST_PATH_IMAGE010
,
Figure 2013104856172100002DEST_PATH_IMAGE012
;
Figure 2013104856172100002DEST_PATH_IMAGE014
为流量统计模块接收到流量值，为设定的第
Figure 541653DEST_PATH_IMAGE008
个较小的流量值，
Figure 2013104856172100002DEST_PATH_IMAGE018
为设定的第
Figure 583427DEST_PATH_IMAGE008
个较大的流量值，，；
Figure 2013104856172100002DEST_PATH_IMAGE022
。 Among them, f is the working frequency,
Figure 2013104856172100002DEST_PATH_IMAGE006
for the first frequency, and
Figure 2013104856172100002DEST_PATH_IMAGE010
,
Figure 2013104856172100002DEST_PATH_IMAGE012
;
Figure 2013104856172100002DEST_PATH_IMAGE014
The flow value received by the flow statistics module, set for the first
Figure 541653DEST_PATH_IMAGE008
a smaller flow value,
Figure 2013104856172100002DEST_PATH_IMAGE018
set for the first
Figure 583427DEST_PATH_IMAGE008
a larger flow value, , ;
Figure 2013104856172100002DEST_PATH_IMAGE022
.

在相邻的流量区间之间留出缓冲空间，即设定相邻的流量值间隔为

Figure 2013104856172100002DEST_PATH_IMAGE024
。当模块工作在频率为
Figure 523754DEST_PATH_IMAGE006
时，数据流量t由小到大增加到超过时将工作频率切换到高频模式
Figure 2013104856172100002DEST_PATH_IMAGE028
（如果接收到的数据流足够大时至
Figure 2013104856172100002DEST_PATH_IMAGE030
，
Figure 2013104856172100002DEST_PATH_IMAGE032
等，频率也可以直接切换到
Figure 2013104856172100002DEST_PATH_IMAGE034
，，直到最高频率
Figure 2013104856172100002DEST_PATH_IMAGE038
）；同样，当模块工作在频率
Figure 712028DEST_PATH_IMAGE006
时，数据流量t由大到小减小到低于
Figure 2013104856172100002DEST_PATH_IMAGE040
时，频率就会被切换到低频模式
Figure 2013104856172100002DEST_PATH_IMAGE042
（如果接收到的数据流迅速变小至
Figure 2013104856172100002DEST_PATH_IMAGE044
，等，频率也可以直接切换到
Figure 2013104856172100002DEST_PATH_IMAGE048
，
Figure 2013104856172100002DEST_PATH_IMAGE050
，直到最低频率
Figure 2013104856172100002DEST_PATH_IMAGE052
）。当
Figure 2013104856172100002DEST_PATH_IMAGE054
时，即流量值介于缓冲区间内，如果当前频率工作在低频率
Figure 347540DEST_PATH_IMAGE042
时，设备频率继续保持在低频
Figure 305918DEST_PATH_IMAGE042
；如果当前频率工作在频率
Figure 442502DEST_PATH_IMAGE006
时，设备也继续维持在
Figure 298331DEST_PATH_IMAGE006
，
Figure 291695DEST_PATH_IMAGE006
与
Figure 2013104856172100002DEST_PATH_IMAGE056
是相邻的频率值。这样就有效的避免了频率的抖动。 Reserve a buffer space between adjacent flow intervals, that is, set the interval of adjacent flow values to
Figure 2013104856172100002DEST_PATH_IMAGE024
. When the module works at a frequency of
Figure 523754DEST_PATH_IMAGE006
When , the data flow t increases from small to large to more than Switch the operating frequency to high frequency mode when
Figure 2013104856172100002DEST_PATH_IMAGE028
(if the received data stream is large enough to
Figure 2013104856172100002DEST_PATH_IMAGE030
,
Figure 2013104856172100002DEST_PATH_IMAGE032
etc., the frequency can also be switched directly to
Figure 2013104856172100002DEST_PATH_IMAGE034
, , up to the highest frequency
Figure 2013104856172100002DEST_PATH_IMAGE038
); Similarly, when the module works at the frequency
Figure 712028DEST_PATH_IMAGE006
When , the data flow t decreases from large to small to below
Figure 2013104856172100002DEST_PATH_IMAGE040
, the frequency will be switched to low frequency mode
Figure 2013104856172100002DEST_PATH_IMAGE042
(if the received data stream rapidly dwindles to
Figure 2013104856172100002DEST_PATH_IMAGE044
, etc., the frequency can also be switched directly to
Figure 2013104856172100002DEST_PATH_IMAGE048
,
Figure 2013104856172100002DEST_PATH_IMAGE050
, until the lowest frequency
Figure 2013104856172100002DEST_PATH_IMAGE052
). when
Figure 2013104856172100002DEST_PATH_IMAGE054
, that is, the flow value is within the buffer zone, if the current frequency is working at a low frequency
Figure 347540DEST_PATH_IMAGE042
When the frequency of the device continues to be kept at a low frequency
Figure 305918DEST_PATH_IMAGE042
; If the current frequency is working at frequency
Figure 442502DEST_PATH_IMAGE006
, the device also continues to maintain
Figure 298331DEST_PATH_IMAGE006
,
Figure 291695DEST_PATH_IMAGE006
and
Figure 2013104856172100002DEST_PATH_IMAGE056
are adjacent frequency values. This effectively avoids frequency jitter.

有益效果：  Beneficial effect :

本发明对比已有技术具有以下显著优点：基于流量的调频方法可以实现自适应调频；避免了频率的振荡，调频灵敏度可调；降低模块功耗，节能。 Compared with the prior art, the present invention has the following significant advantages: the flow-based frequency modulation method can realize self-adaptive frequency modulation; frequency oscillation is avoided, frequency modulation sensitivity can be adjusted; module power consumption is reduced and energy is saved.

附图说明   Description of drawings

图1是本发明调频方法的结构框图；  Fig. 1 is the structural block diagram of frequency modulation method of the present invention;

图2是本发明的实现流程图； Fig. 2 is the realization flowchart of the present invention;

图3是本发明带缓冲和不带缓冲的两种调频机制； Fig. 3 is two kinds of FM mechanisms with buffering and without buffering in the present invention;

图4是本发明双门限带缓冲的调频状态转换图； Fig. 4 is the FM state transition diagram of double threshold band buffering of the present invention;

图5是本发明的实施方式结构图。 Fig. 5 is a structural diagram of an embodiment of the present invention.

具体实施方式  Detailed ways

如图1，图2所示，在本设计的调频方法是通过计算网络数据流的大小来进行调频的。通过在特定时间内（设定一个特定的时间门限值）计算出网络设备接收到的数据包数，根据接收到的数据包数来反应当前网络中的流量，然后把这个流量作为选择各个频率的指标，从而实现基于流量的自适应调频。 As shown in Fig. 1 and Fig. 2, the frequency modulation method in this design performs frequency modulation by calculating the size of the network data flow. By calculating the number of data packets received by the network device within a specific time (setting a specific time threshold), the current traffic in the network is reflected according to the number of received data packets, and then this traffic is used as the choice of each frequency indicators, so as to realize traffic-based adaptive frequency regulation.

本发明所采用的技术方案是： The technical scheme adopted in the present invention is:

1）对网络设备流量的准确计算： 1) Accurate calculation of network device traffic:

当网络数据流进入流量统计模块的输入队列时，统计模块将根据以太网控制器将数据流划分成一个个数据包，并且计算出每个数据包的字节数，通过一个计数模块统计设定时间内所接收到的总的字节数，可以调整统计时间的大小来调整本调频方法的灵敏性。 When the network data flow enters the input queue of the traffic statistics module, the statistics module will divide the data flow into data packets according to the Ethernet controller, and calculate the number of bytes of each data packet, and set the statistics through a counting module The total number of bytes received within the time, the size of the statistics time can be adjusted to adjust the sensitivity of the frequency modulation method.

2）多个频率的产生 2) Generation of multiple frequencies

通过使用FPGA厂商提供的时钟 IP核，或通过硬件描述语言的方法来产生全局时钟频率，全局时钟的要求的特点：一、负载能力特别强, 任何一个全局时钟驱动线都可以驱动芯片内部的触发器; 二是时延差特别小; 三是时钟信号波形畸变小, 工作可靠性好。  The global clock frequency is generated by using the clock IP core provided by the FPGA manufacturer, or through the method of hardware description language. The required characteristics of the global clock: 1. The load capacity is particularly strong, and any global clock drive line can drive the trigger inside the chip. The second is that the delay difference is very small; the third is that the distortion of the clock signal waveform is small, and the working reliability is good. the

3）调频机制 3) FM mechanism

流量

Figure DEST_PATH_IMAGE058
相对于前一时刻，无非是增加或者减少； flow
Figure DEST_PATH_IMAGE058
Compared with the previous moment, it is nothing more than an increase or decrease;

当t增加时，频率和流量值的关系式如下所示：  When t increases, the relationship between frequency and flow value is as follows:

Figure 609806DEST_PATH_IMAGE002
Figure 609806DEST_PATH_IMAGE002

当

Figure DEST_PATH_IMAGE060
减小时, 频率和流量值的关系式如下所示： when
Figure DEST_PATH_IMAGE060
When decreasing, the relationship between frequency and flow value is as follows:

Figure DEST_PATH_IMAGE062
  
Figure DEST_PATH_IMAGE062
  

其中， f为工作频率， 为第

Figure 680323DEST_PATH_IMAGE008
个频率，且,;
Figure 751550DEST_PATH_IMAGE014
为流量统计模块接收到流量值，为设定的第
Figure 356286DEST_PATH_IMAGE008
个较小的流量值，
Figure 22890DEST_PATH_IMAGE018
为设定的第
Figure 101705DEST_PATH_IMAGE008
个较大的流量值，
Figure 786633DEST_PATH_IMAGE020
，；
Figure 126228DEST_PATH_IMAGE022
。 Among them, f is the working frequency, for the first
Figure 680323DEST_PATH_IMAGE008
frequency, and , ;
Figure 751550DEST_PATH_IMAGE014
The flow value received by the flow statistics module, set for the first
Figure 356286DEST_PATH_IMAGE008
a smaller flow value,
Figure 22890DEST_PATH_IMAGE018
set for the first
Figure 101705DEST_PATH_IMAGE008
a larger flow value,
Figure 786633DEST_PATH_IMAGE020
, ;
Figure 126228DEST_PATH_IMAGE022
.

如图3、图4所示，在相邻的流量区间之间留出缓冲空间，即设定相邻的流量值间隔为

Figure 946416DEST_PATH_IMAGE024
。当模块工作在频率为
Figure 485851DEST_PATH_IMAGE006
时，数据流量t由小到大增加到超过
Figure 428399DEST_PATH_IMAGE026
时将工作频率切换到高频模式
Figure 804017DEST_PATH_IMAGE028
（如果接收到的数据流足够大时至
Figure 913049DEST_PATH_IMAGE030
，等，频率也可以直接切换到
Figure 561386DEST_PATH_IMAGE034
，
Figure 486616DEST_PATH_IMAGE036
，直到最高频率）；同样，当模块工作在频率
Figure 395590DEST_PATH_IMAGE006
时，数据流量t由大到小减小到低于时，频率就会被切换到低频模式
Figure 217101DEST_PATH_IMAGE042
（如果接收到的数据流迅速变小至，
Figure 115098DEST_PATH_IMAGE046
等，频率也可以直接切换到，
Figure 329228DEST_PATH_IMAGE050
，直到最低频率
Figure 895339DEST_PATH_IMAGE052
）。当
Figure 134690DEST_PATH_IMAGE054
时，即流量值介于缓冲区间内，如果当前频率工作在低频率
Figure 186609DEST_PATH_IMAGE042
时，设备频率继续保持在低频
Figure 511411DEST_PATH_IMAGE042
；如果当前频率工作在频率
Figure 68164DEST_PATH_IMAGE006
时，设备也继续维持在
Figure 896442DEST_PATH_IMAGE006
，
Figure 693497DEST_PATH_IMAGE006
与
Figure 990748DEST_PATH_IMAGE056
是相邻的频率值。这样就有效的避免了频率的抖动。不管当前频率是
Figure 101924DEST_PATH_IMAGE006
还是
Figure 299556DEST_PATH_IMAGE056
，调频设备都能够正常工作。 As shown in Figure 3 and Figure 4, a buffer space is reserved between adjacent flow intervals, that is, the interval between adjacent flow values is set to
Figure 946416DEST_PATH_IMAGE024
. When the module works at a frequency of
Figure 485851DEST_PATH_IMAGE006
When , the data flow t increases from small to large to more than
Figure 428399DEST_PATH_IMAGE026
Switch the operating frequency to high frequency mode when
Figure 804017DEST_PATH_IMAGE028
(if the received data stream is large enough to
Figure 913049DEST_PATH_IMAGE030
, etc., the frequency can also be switched directly to
Figure 561386DEST_PATH_IMAGE034
,
Figure 486616DEST_PATH_IMAGE036
, up to the highest frequency ); Similarly, when the module works at the frequency
Figure 395590DEST_PATH_IMAGE006
When , the data flow t decreases from large to small to below , the frequency will be switched to low frequency mode
Figure 217101DEST_PATH_IMAGE042
(if the received data stream rapidly dwindles to ,
Figure 115098DEST_PATH_IMAGE046
etc., the frequency can also be switched directly to ,
Figure 329228DEST_PATH_IMAGE050
, until the lowest frequency
Figure 895339DEST_PATH_IMAGE052
). when
Figure 134690DEST_PATH_IMAGE054
, that is, the flow value is within the buffer zone, if the current frequency is working at a low frequency
Figure 186609DEST_PATH_IMAGE042
When the frequency of the device continues to be kept at a low frequency
Figure 511411DEST_PATH_IMAGE042
; If the current frequency is working at frequency
Figure 68164DEST_PATH_IMAGE006
, the device also continues to maintain
Figure 896442DEST_PATH_IMAGE006
,
Figure 693497DEST_PATH_IMAGE006
and
Figure 990748DEST_PATH_IMAGE056
are adjacent frequency values. This effectively avoids frequency jitter. Regardless of the current frequency
Figure 101924DEST_PATH_IMAGE006
still
Figure 299556DEST_PATH_IMAGE056
, FM equipment can work normally.

本发明的具体实施方式参见图5，实施方式选用的NetFPGA开发平台上的参考路由器。为了确保调频方法能够准确地计算出当前的网络流量，所以调频模块的加入必须要嵌入在整个路由器设计的前端模块，这样才能保证不会因为后续处理模块的处理性能不够而影响对流量计算的准确性。 Referring to Fig. 5 for the specific embodiment of the present invention, the reference router on the NetFPGA development platform selected by the embodiment. In order to ensure that the frequency modulation method can accurately calculate the current network traffic, the frequency modulation module must be embedded in the front-end module of the entire router design, so as to ensure that the accuracy of traffic calculation will not be affected by the insufficient processing performance of the subsequent processing modules sex.

在参考路由器网络层User_Data_Path的前端，即数据包接收模块（Rx_Quence）中加入调频判定机制。Rx_Quence模块介于链路层处理模块和网络层处理模块之间，它的用途是接收链路层的输入数据并封装成数据包，然后经过FIFO(First In First Out)缓存后转发给网络层模块（User_Data_Path）进行路由处理。 The frequency modulation decision mechanism is added to the front end of User_Data_Path in the network layer of the reference router, that is, the data packet receiving module (Rx_Quence). The Rx_Quence module is between the link layer processing module and the network layer processing module. Its purpose is to receive the input data of the link layer and encapsulate it into a data packet, and then forward it to the network layer module after being buffered by FIFO (First In First Out) (User_Data_Path) for routing processing.

调频判定机制首先根据流量统计模块统计出的网络流量来置位相应的调频标志信号，多频率生成模块用于生成各种频率并且根据调频标志信号的不同选取一个合适的频率，然后通过全局时钟线传递到路由器的各个功能模块。多频率模块共有五个时钟频率可以选择，为了明显地区分各个频率下路由器的处理能力和功耗需求，本设计的五个频率为倍频关系且是通过Xilinx DCM IP核生成的，大小分别为7.813MHz，15.625MHz，31.25MHz，62.5MHz和最高频率125MHz。调频判定机制有相应的拥有五个调频标志信号，分别对应着不同的频率。数据流量统计模块则用于统计网口接受到的数据流量，统计数据包模块的原理是当Rx_Quence接收到数据帧时，把数据帧封装成一个个数据包并且计算出数据包的大小，并且还要为每个数据包产生一个时钟周期长度的包标志信号，然后在一定的时间门限区间内采集包标志信号来实现对数据流量的计算。为了防止误统计，数据包的标志信号必须在一个时钟周期后复位（本文中的包标记信号为1，复位为0）。而且，为了使得统计的正确性，数据流量统计模块所用到的时钟信号不会随调频信号的变化而变化，它的大小为125MHz。 The frequency modulation determination mechanism first sets the corresponding frequency modulation flag signal according to the network traffic counted by the traffic statistics module. Passed to each functional module of the router. The multi-frequency module has a total of five clock frequencies to choose from. In order to clearly distinguish the processing capability and power consumption requirements of the router at each frequency, the five frequencies in this design are multiplied and generated by the Xilinx DCM IP core. The sizes are respectively 7.813MHz, 15.625MHz, 31.25MHz, 62.5MHz and the highest frequency 125MHz. The FM judgment mechanism has correspondingly five FM flag signals corresponding to different frequencies. The data traffic statistics module is used to count the data traffic received by the network port. The principle of the statistical data packet module is that when Rx_Quence receives a data frame, it encapsulates the data frame into a data packet and calculates the size of the data packet, and also It is necessary to generate a packet flag signal with a clock cycle length for each data packet, and then collect the packet flag signal within a certain time threshold interval to realize the calculation of the data flow. In order to prevent false statistics, the flag signal of the data packet must be reset after one clock cycle (the packet flag signal in this article is 1, and the reset is 0). Moreover, in order to make the statistics correct, the clock signal used by the data traffic statistics module will not change with the frequency modulation signal, and its size is 125MHz.

本发明实施例参考路由器调频自适应调频方法，调频方法的具体运行步骤如下： The embodiment of the present invention refers to the router frequency modulation adaptive frequency modulation method, and the specific operation steps of the frequency modulation method are as follows:

步骤1.通过Xilinx DCM IP核产生五个不同的频率； Step 1. Generate five different frequencies through the Xilinx DCM IP core;

步骤2.从NetFPGA网口接收数据流，通过以太网控制器划分为一个个数据包，并且统计数据包的字节数，产生数据包标志位，同时开启定时； Step 2. Receive the data flow from the NetFPGA network port, divide it into data packets by the Ethernet controller, and count the number of bytes of the data packet, generate the data packet flag bit, and start the timing at the same time;

步骤3.在定时1s的时间内，计算出所产生的数据标志位总数； Step 3. In the time of timing 1s, calculate the total number of generated data flags;

步骤4.根据不同的标志总数，确定相应的频率； Step 4. Determine the corresponding frequency according to the total number of different signs;

步骤5.把确定的频率作为参考路由器各个功能模块的运行频率。 Step 5. Use the determined frequency as the operating frequency of each functional module of the reference router.

Claims (1)

1. A low-power consumption frequency modulation method based on flow is characterized by comprising the following steps:
the method comprises the following steps: accurate computation of network device traffic
When a network data stream enters an input queue of a flow counting module, the counting module divides the data stream into a plurality of data packets, calculates the byte number of each data packet, counts the total received byte number within a set time through a counting module, and can adjust the sensitivity of the frequency modulation method by adjusting the counting time;
step two: generation of multiple frequencies
The global clock frequency is generated by using a clock IP core provided by an FPGA manufacturer or by a hardware description language method, and the required characteristics of the global clock are as follows: the chip has the advantages that the chip has extremely strong load capacity, any one global clock drive line can drive a trigger in the chip, the time delay difference is extremely small, the waveform distortion of a clock signal is small, and the working reliability is good;
step three: frequency modulation mechanism
The flow t is not increased or decreased relative to the previous time;
when t increases, the frequency and flow values are related as follows:
Figure DEST_PATH_IMAGE002
when t decreases, the frequency and flow values are related as follows:
Figure DEST_PATH_IMAGE004
wherein, fin order to be the working frequency of the frequency converter,is as follows
Figure DEST_PATH_IMAGE008
A frequency of
Figure DEST_PATH_IMAGE010
,
Figure DEST_PATH_IMAGE012
;
Figure DEST_PATH_IMAGE014
In order for the flow statistics module to receive the flow value,
Figure DEST_PATH_IMAGE016
is a set one
Figure 783442DEST_PATH_IMAGE008
The flow rate value of the flow rate is smaller,is a set one
Figure 66525DEST_PATH_IMAGE008
The value of the flow rate that is greater,，
Figure 404227DEST_PATH_IMAGE012
；
Figure DEST_PATH_IMAGE022
；
buffer space is reserved between adjacent flow intervals, namely adjacent flow value intervals are set as
Figure DEST_PATH_IMAGE024
(ii) a When the module operates at a frequency of
Figure 394049DEST_PATH_IMAGE006
Time of day, data traffictIncreases from small to large to exceed
Figure DEST_PATH_IMAGE026
While switching the operating frequency to a high frequency modeIf the received data stream is large enough to
Figure DEST_PATH_IMAGE030
，Etc., the frequency may also be switched directly to
Figure DEST_PATH_IMAGE034
，
Figure DEST_PATH_IMAGE036
Up to the highest frequency
Figure DEST_PATH_IMAGE038
(ii) a Also, when the module is operating at frequencyTime of day, data traffictDecreases from large to small to belowThen the frequency is switched to the low frequency modeIf the received data stream rapidly gets smaller
Figure DEST_PATH_IMAGE044
，
Figure DEST_PATH_IMAGE046
Etc., the frequency may also be switched directly to
Figure DEST_PATH_IMAGE048
，
Figure DEST_PATH_IMAGE050
Up to the lowest frequency(ii) a When in use
Figure DEST_PATH_IMAGE054
When the current frequency is in the buffer interval, if the current frequency is in low frequency
Figure 982912DEST_PATH_IMAGE042
While the frequency of the device continues to remain at a low frequency(ii) a If the current frequency is operating at the frequency
Figure 342535DEST_PATH_IMAGE006
While the equipment continues to remain at
Figure 652293DEST_PATH_IMAGE006
，
Figure 660701DEST_PATH_IMAGE006
And
Figure DEST_PATH_IMAGE056
are adjacent frequency values.
CN201310485617.2A 2013-10-16 2013-10-16 A kind of low power consumption frequency modulation based on flow Active CN103618682B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201310485617.2A CN103618682B (en) 2013-10-16 2013-10-16 A kind of low power consumption frequency modulation based on flow

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201310485617.2A CN103618682B (en) 2013-10-16 2013-10-16 A kind of low power consumption frequency modulation based on flow

Publications (2)

Publication Number Publication Date
CN103618682A true CN103618682A (en) 2014-03-05
CN103618682B CN103618682B (en) 2017-06-06

Family

ID=50169386

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201310485617.2A Active CN103618682B (en) 2013-10-16 2013-10-16 A kind of low power consumption frequency modulation based on flow

Country Status (1)

Country Link
CN (1) CN103618682B (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105049365A (en) * 2015-06-15 2015-11-11 国家计算机网络与信息安全管理中心 Adaptive frequency modulation energy-saving method for multi-core multi-thread intrusion detection device
CN113132272A (en) * 2021-03-31 2021-07-16 中国人民解放军战略支援部队信息工程大学 Network switching frequency dynamic adjustment method and system based on flow perception and network switching chip structure

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1581844A (en) * 2003-08-15 2005-02-16 上海贝尔阿尔卡特股份有限公司 Packet switching network distributed adaptive dither buffer adjusting method
CN101527672A (en) * 2008-03-07 2009-09-09 瑞昱半导体股份有限公司 Network flow control device and method

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1581844A (en) * 2003-08-15 2005-02-16 上海贝尔阿尔卡特股份有限公司 Packet switching network distributed adaptive dither buffer adjusting method
CN101527672A (en) * 2008-03-07 2009-09-09 瑞昱半导体股份有限公司 Network flow control device and method

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
汪漪 等: ""快速自适应调频机制及其在NetFPGA上的实现"", 《计算机学报》 *

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105049365A (en) * 2015-06-15 2015-11-11 国家计算机网络与信息安全管理中心 Adaptive frequency modulation energy-saving method for multi-core multi-thread intrusion detection device
CN105049365B (en) * 2015-06-15 2019-05-31 国家计算机网络与信息安全管理中心 A kind of self-adapting frequency modulation power-economizing method of Multi-core intrusion detection device
CN113132272A (en) * 2021-03-31 2021-07-16 中国人民解放军战略支援部队信息工程大学 Network switching frequency dynamic adjustment method and system based on flow perception and network switching chip structure
CN113132272B (en) * 2021-03-31 2023-02-14 中国人民解放军战略支援部队信息工程大学 Network switching frequency dynamic adjustment method and system based on flow perception and network switching chip structure

Also Published As

Publication number Publication date
CN103618682B (en) 2017-06-06

Similar Documents

Publication Publication Date Title
US9531596B2 (en) System and method for dynamically power and performance optimized server interconnects
US9525593B2 (en) Oversubscribing a packet processing device to adjust power consumption
CN104767694B (en) A kind of stream compression forwarding method towards Fat Tree data center network architectures
CN106533960A (en) Data center network routing method based on Fat-Tree structure
CN103929372B (en) Active queue management method and apparatus
CN103139093B (en) Based on the express network data stream load equalization scheduling method of FPGA
CN107948103A (en) A kind of interchanger PFC control methods and control system based on prediction
CN102291389A (en) Cross-layer congestion control method in satellite network
CN101582852A (en) Method and system for network congestion management
CN108989236B (en) A flow control method, device and system
CN102752192B (en) Bandwidth allocation method of forwarding and control element separation (ForCES) transmission mapping layer based on stream control transmission protocol (SCTP)
CN104883710A (en) Congestion control method for wireless sensor network in star configuration
CN102739510B (en) A kind of flow control method and device
CN103618682B (en) A kind of low power consumption frequency modulation based on flow
CN107332785B (en) A kind of effective discharge control method based on dynamic duty threshold value
CN103825833B (en) A kind of method for controlling network flow that statistics is deflected based on packet
Meng et al. Greening the internet using multi-frequency scaling scheme
CN107800711A (en) A kind of method that OpenFlow controllers resist ddos attack
JP5177004B2 (en) Operation mode changing device and communication device
CN106792831B (en) Congestion avoidance module and method for wireless node level and wireless link level
CN104052619A (en) A method for energy saving of network equipment supporting frequency modulation and consumption reduction
CN105049365B (en) A kind of self-adapting frequency modulation power-economizing method of Multi-core intrusion detection device
CN102377670B (en) Dynamic route adjustment method of user QoS (Quality of Service) oriented to cognitive network
CN102088414B (en) Method for managing passive queue by virtue of combining random drop and drop front
CN115665053A (en) DPU-oriented token bucket algorithm current limiting system based on FPGA

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant
TR01 Transfer of patent right
TR01 Transfer of patent right

Effective date of registration: 20200509

Address after: 314000 Room 101, floor 1, building 2, No. 87, South Street, Changfeng street, Wuzhen Town, Tongxiang City, Jiaxing City, Zhejiang Province

Patentee after: Jiaxinghuabing Internet of things Technology Co., Ltd

Address before: Hangzhou City, Zhejiang province 310018 Xiasha Higher Education Park No. 2 street

Patentee before: HANGZHOU DIANZI University