CN103428104B - A kind of jamming control method based on content center network - Google Patents

Abstract

The invention discloses a congestion control method based on a content-centric network, comprising: calculating a CIB value in a router for a Data packet to be sent; wherein, the CIB value is used to reflect the current congestion degree of the router; and comparing the calculated CIB value and the CIB value before the Data packet to be sent, and the CIB value that is used to represent the worst router congestion state in the two is used as the CIB value of the Data packet, and then sends the Data packet; the client receives the CIB value After the Data packet, adjust the size of the sending window according to the CIB value of the Data packet; wherein, when the CIB value reflects that the router is relatively idle, increase the size of the sending window, and when the CIB value reflects that the router When more crowded, reduce the size of the sending window.

Description

A Congestion Control Method Based on Content-Centric Network

technical field

The invention relates to the field of network communication, in particular to a content-centric network-based congestion control method.

Background technique

The Internet has developed tremendously in the past few decades. The traditional Internet mainly focuses on end-to-end communication, but today's Internet mainly focuses on the distribution and acquisition of content. Under this premise, Content-Centric Network (CCN, Content-Centric Network) came into being. CCN is a brand-new network architecture, which completely abandons the way that IP networks use IP addresses to name each host. CCN does not name the host, but the content. In the CCN network, each file is split into several fixed-size blocks (Chunk), and each block is assigned a fixed name, such as: ccnx://hpnl.ioa.ac.cn/video/filename/ _chunknum/_timestamp. There are two kinds of data packets in CCN, request packet (Interest) and data packet (Data). The Interest package contains the content name and other relevant information (such as version, permission, etc.), while the Data package contains the content name, other relevant information and payload.

CCN routers are different from traditional routers: CCN routers have a cache function (the content is cached in the ContentStore), and it can cache Data packets passing through it according to a certain strategy; in addition, CCN routers route Interest packets according to their names (the forwarding information The table (FIB) stores the corresponding information of the name and the interface), and transmits the Data package according to the reverse path of the Interest package (the state information of the Interest is stored in the waiting request table (PIT)).

During the operation of the CCN network, the end user sends an Interest packet, and the CCN router routes the Interest packet according to the name. If there is such content in the cache on a router node in the routing path, it will directly return the corresponding Data packet. If there is no such content on the router along the way, the Interest is finally forwarded to the terminal server. The terminal server returns this content, and the router along the path will cache this content. If the router receives this Interest packet again, it can directly return Data. Therefore, the CCN network saves user download time and reduces repeated transmission of resources.

In order for the network to operate efficiently, congestion control must be performed on the network. There are mainly two congestion control methods that have been proposed in the CCN network:

1. Reference 1 "N. Rozhnova and S. Fdida, "An effective hop-by-hop interestshaping mechanism for ccn communications," in IEEE NOMEN Workshop, co-located with INFOCOM, 2012" proposes a method that each node in the network (including Terminals and routers) make dynamic predictions on future congestion according to their current information (queue length, egress bandwidth, RTT (Round-Trip Time, round-trip delay)), adjust the rate at which they send Interest at the moment according to the prediction results, and then Control the receiving rate of Data to achieve the purpose of eliminating network congestion.

Figure 1 is a schematic diagram of such a method in operation. As shown in Figure 1, C(t) represents the available bandwidth for nodes to send data out at time t, e(t) represents the amount of data in the node cache at time t, B represents the cache size, and r is a set threshold. A(t) represents the delay from when the Interest is sent from this node to when the corresponding Data is received. γ(t) represents the Interest sending rate at time t. The formula for adjusting the Interest sending rate proposed by this method is as follows:

Among them, h is a set weight value.

2. Reference 2 "G. Carofiglio, M. Gallo, and L. Muscariello, "Icp: Design and evaluation of an interest control protocol for content-centric networking," inIEEE NOMEN Workshop, co-located with INFOCOM, 2012" The method does not require routers to participate in congestion control, and all congestion control is completed on the request initiator (that is, the host that sends out the Interest). It regards the Data packet as the ACK packet in TCP, starts a timer after sending the Interest packet, and cancels the timer when the Data packet is received. If the timer expires, it is considered that packet loss has occurred in the network and congestion occurs. The amount of Interest sent is halved. If there is no packet loss, it is considered that there is no congestion in the network, and the number of Interest packets should be increased (according to the comparison result of the sending window and the set threshold size, an exponential increase or a linear increase method is used). The formula for setting the timeout time T in this method is as follows:

T=RTT _min +(RTT _max -RTT _min )δ

Among them, δ is an adjustable parameter.

The above two methods in the prior art have certain problems in practice.

First, in CCN, Interest may hit any node along the path, so the RTT of each Interest packet is different and varies widely, and there is no correlation between RTTs. Both reference 1 and reference 2 adopt RTT-based design ideas (such as the parameter A(t) in reference 1, and the timeout time T in reference 2). The method of using the measured RTT as the RTT of the Interest packet to be sent in Reference 1 does not take into account the fact that the above-mentioned RTT varies widely, which may cause Data packets to arrive at a certain node concentratedly, thereby causing congestion. The method of using a value between RTT _min and RTT _max as the timeout time in Reference 2 will cause those Data packets whose RTT is greater than the timeout time T to be misjudged as packet loss, so that they will be mistaken for packet loss when no congestion occurs. Congestion occurs.

Secondly, in the method described in reference 1, each node only adjusts the number of Interest sent according to its current state. If a node is congested, the Data packets it sends to the upstream node will be reduced accordingly, so that the queue length of the upstream node e(t) decreases. At this time, the upstream node will think that the network is in good condition, and then increase the interest sending rate, causing the interest to gather at the congested node. Too much interest will cause the buffer of the congested node to overflow, resulting in packet loss. In the method described in reference 2, the network responds to congestion only when the network has lost packets (the timer expires), and there will be more packets during the period from packet loss to terminal response thrown away.

Finally, the methods discussed in the above two references do not consider the issue of fairness of streams, and the bandwidth cannot be allocated fairly between streams. In addition, if congestion occurs, both the high-rate flow and the low-rate flow enter a congested state, which is unfair to the low-rate flow.

Contents of the invention

The purpose of the present invention is to overcome the disadvantages of the existing network congestion control method, such as easy misjudgment and poor fairness, so as to provide a timely and fair network congestion control method.

In order to achieve the above object, the present invention provides a content-centric network-based congestion control method, including:

Step 1), calculating the CIB value for the Data packet to be sent in the router; wherein, the CIB value is used to reflect the current congestion level of the router;

Step 2), comparing the CIB value obtained in step 1) with the CIB value before the Data packet to be sent, the CIB value used to represent the worst state of router congestion in the two as the CIB value of the Data packet, Then send the Data packet;

Step 3), after the client receives the Data packet, adjust the size of the sending window according to the CIB value of the Data packet; wherein, when the CIB value reflects that the router is relatively idle, increase the size of the sending window, When the CIB value reflects that the router is relatively congested, reduce the size of the sending window.

In the above-mentioned technical scheme, described CIB value has N, wherein some CIB values represent that the client reduces the Interest sending window, and some CIB values represent that the client increases the Interest sending window; Described calculation CIB value comprises: setting M Thresholds, using the M thresholds to divide the queue length into N intervals, each interval corresponding to a CIB value, where N=M+1.

In the above technical solution, the CIB value is represented by 2 bits, including "00", "01", "10", and "11", which are respectively used to indicate that the current congestion level of the router is "excellent", "good", " Medium", "poor"; the calculated CIB value includes:

Step 1-1), calculating the virtual queue maximum length threshold maxQ and the virtual queue length minimum threshold minQ;

Among them, B represents the cache size of the router, F represents the number of flows contained in the router, and maxRatio and minRatio are two parameters determined according to actual conditions;

Step 1-2), smoothing the queue length Q of the flow where the Data packet to be sent is located, to obtain avgQ:

Among them, Wq is the weight factor; n indicates the calculation of avgQ for the nth time;

Step 1-3), compare avgQ with maxQ and minQ, if avgQ≤minQ, then CIB is set to "00", if avgQ≥maxQ, then CIB is set to "11", if minQ<avgQ<maxQ, then execute Next step;

Steps 1-4), calculating the marking probability P:

Then set the CIB to "00" with the probability of (1-P); set the CIB to one of "01" or "10" with the probability of P, if Set CIB to "01", otherwise, set to "10".

In the above-mentioned technical scheme, described step 3) comprises:

Step 3-1), determine the size of the CIB value, if the CIB value is "00", perform the next step, if the CIB value is "01", perform step 3-3), if the CIB value is "10", perform step 3 -4), if the CIB value is "11", perform steps 3-5);

Step 3-2), comparing the current sending window size Iswnd with the preset threshold ssthresh, if Iswnd is smaller than ssthresh, then the sending window size Iswnd becomes larger with the first amplitude value, if the Iswnd is greater than or equal to ssthresh, Then the window size Iswnd becomes larger with the second amplitude value; the first amplitude value is greater than the second amplitude value;

Step 3-3), keep sending window size Iswnd unchanged;

Step 3-4), the sending window size Iswnd becomes smaller with the third amplitude value;

Step 3-5), the sending window size Iswnd becomes smaller with a fourth amplitude value; the absolute value of the fourth amplitude value is greater than the absolute value of the third amplitude value.

In the above technical solution, the size of the first amplitude value is 1; the size of the second amplitude value is 1/Iswnd _old , and Iswnd _old represents the size of the original sending window; the size of the third amplitude value is 0.25; The magnitude of the fourth amplitude value is 0.5.

In the above technical solution, when there are multiple Data packets to be sent by the router and belong to multiple streams, in the cache of the router, a virtual queue is maintained for each stream, and the router serves each virtual queue cyclically; The above service includes: taking out the data packet at the head of the queue from the virtual queue, calculating the CIB value for the Data packet, comparing and adding it to the Data packet.

The advantages of the present invention are:

1. Do not use an RTT-based congestion control mechanism to avoid the impact of dynamic changes in RTT on congestion control;

2. Adjust the terminal Interest sending rate before the network packet loss occurs, so that the terminal responds to congestion in a more timely manner, reducing or even avoiding packet loss;

3. Fairly allocate bandwidth to each flow, and at the same time, perform congestion detection on each flow separately. When network congestion occurs, reduce the bandwidth of high-rate flows and protect the bandwidth of low-rate flows.

Description of drawings

FIG. 1 is a schematic diagram of a congestion control method in the prior art during operation;

FIG. 2 is a schematic diagram of a network structure of a content-centric network;

Fig. 3 is a flowchart of the congestion control method of the present invention;

Fig. 4 is the schematic diagram of the virtual queue maintained in the cache of router;

Fig. 5 is a topological diagram of a simulation experiment;

Fig. 6 is a schematic diagram of the Iswnd changes of each terminal in a simulation experiment;

Fig. 7 is in a simulation experiment, the schematic diagram of the size of total queue and virtual queue in Router2;

FIG. 8 is a schematic diagram of bandwidth conditions of terminals and routers in a simulation experiment.

detailed description

For ease of understanding, the meanings of some symbols involved in the present invention are firstly described in a unified manner below.

B: The total buffer size of the router interface;

F: the number of flows in the current router;

Q: the length of the virtual queue of the flow;

Wq: weight factor, 0<Wq<1;

avgQ: smoothed virtual queue length;

maxQ: the maximum threshold of the virtual queue length;

minQ: the minimum threshold of the virtual queue length;

maxRatio: the ratio of maxQ to B/F;

minRatio: The ratio of minQ to B/F.

The present invention will be further described now in conjunction with accompanying drawing.

FIG. 2 is a schematic diagram of a network structure of a content-centric network. As shown in the figure, the network includes a client, a server, and a router between the client and the server. For the convenience of illustration, only one client, one server and three routers are shown in this figure, but those skilled in the art can easily understand that in an actual network, the number of clients, servers and routers By no means limited to the above numbers, the content-centric network shown in Figure 2 is used as an example in the following to describe the congestion control method of the present invention applied to the network, but the congestion control method of the present invention is also applicable to other content centers network.

Referring to FIG. 3 , the implementation steps of the congestion control method of the present invention will be described in detail below.

Step 1), the router calculates the CIB value for the Data packet to be sent.

As mentioned in the background art, after the client sends out the Interest packet, the router in the network receives the Interest packet. There are two situations at this time, one is that the router has previously stored the content of the Data packet corresponding to the Interest packet, and the other is that the router has not yet stored the content of the Data packet corresponding to the Interest packet. For the latter case, the router needs to forward the Interest packet, and wait for the server to return a corresponding Data packet according to the received Interest packet. Whether the router has stored the Data packet before or newly received the Data packet returned by the memory, it needs to return the Data packet to the client through the network. Before returning to the client, the Data packet needs to be Calculate the CIB value.

The router uses the CIB value to inform the client of its current congestion level. In this embodiment, the CIB is represented by 2 bits, and the CIB values can be "00", "01", "10", "11", these values are used to indicate the current congestion level of the router as "excellent", "good", "medium" and "poor", "excellent" means that the current router is idle, and "poor" means that the current router is very busy. Those skilled in the art can easily understand that the CIB can also be represented by more bits, and at this time, the congestion state of the router can be further subdivided.

Before the router calculates the CIB value for the Data packet, it needs to read the Data packet from the cache of the router. Generally speaking, a router will serve multiple streams at the same time, and the "flow" refers to a collection of data packets with the same name prefix but different chunk sequence numbers. As a preferred implementation, in this embodiment, in order to serve each flow fairly, as shown in Figure 4, in the cache of the router, a virtual queue is maintained for each flow. Take out the Data packet at the head of the queue from the virtual queue of the next flow, calculate the CIB value for the Data packet, add the CIB value to the Data packet, and transmit it to the link; then take out the Data packet at the head of the queue from the virtual queue of the next flow, and calculate And send out after adding CIB information.... The router round-robin serves each flow so that each flow has an equal chance of being sent, which in turn allows for a fair allocation of bandwidth among the flows. For those low-rate flows, the length of the virtual queue is small, and the queuing delay of the Data packets is relatively small. In other embodiments, only one queue is maintained in the cache of the router, and Data packets belonging to different flows are processed in the same queue, but this will obviously reduce efficiency and easily cause the problem of unfair bandwidth allocation.

The calculation method of the CIB value is described in detail below.

Assuming that when calculating the CIB value for a Data packet, there are F flows in the router, the buffer size allocated for each flow should be B/F, and two thresholds can be calculated using maxRatio and minRatio—maxQ and minQ:

Wherein, the values of maxRatio and minRatio are determined according to actual conditions. In this embodiment, maxRatio=0.7 and minRatio=0.3.

Smooth the queue length Q of the current flow to get avgQ:

Wherein, the value of the weight factor Wq is determined according to the actual situation, and in this embodiment, Wq=0.8. n indicates the nth calculation of avgQ.

Compare avgQ with maxQ and minQ, if avgQ≤minQ, then CIB should be set to "00"; if avgQ≥maxQ, then CIB should be set to "11"; if minQ<avgQ<maxQ, further calculation is required, first Calculate a marking probability P:

Set the CIB to "00" with the probability of (1-P); set the CIB to one of "01" or "10" with the probability of P. The specific setting depends on avgQ and size, if Set CIB to "01", otherwise, set to "10".

The above CIB calculation method is for the CIB value in this embodiment including four possible situations. If in other embodiments, there are more than 4 possible values of CIB, when calculating CIB, in addition to the thresholds minQ and maxQ already used in this embodiment, other thresholds can also be set, and a total of M (M≥ 1) thresholds, so that the size of the queue length is divided into N intervals (N=M+1), and each interval corresponds to a CIB value. Some of the obtained N CIB values indicate that the client should increase the Interest sending window, some CIB values indicate that the client should reduce the Interest sending window, and one value can be used (or not) to indicate that the size of the Interest sending window remains unchanged. Different CIB values represent different increase (or decrease) ranges.

Step 2), after the router calculates the CIB value for the Data packet to be sent, compare the calculated CIB value with the CIB value carried before the Data packet, and use the worst CIB value for representing the router congestion state as the Data packet The CIB value carried, and then send the Data packet to the link.

As mentioned before, the CIB value is used to inform the client of the congestion status of the router, so the CIB value carried by the Data packet should be used to indicate the congestion information of the router with the most serious congestion on the link. In this step, it is necessary to compare the CIB value calculated in step 1) with the CIB value carried before the Data packet, and decide whether to keep the CIB value carried in the Data packet according to the comparison result, or use the newly calculated CIB value The CIB value is used as the CIB value carried in the Data packet. If the current router is the closest router to the server, the Data packet returned by the server does not contain the CIB value, and the calculated CIB value must be added to the Data packet.

Step 3), after receiving the Data packet, the client adjusts the size of the sending window according to the CIB value.

In this step, the size of the sending window can be represented by Iswnd, which represents the maximum number of Interest packets that the client has sent but has not received the corresponding Data packet. The adjustment of Iswnd can achieve the purpose of adjusting the sending rate of Interest packets, and then achieve The purpose of adjusting the rate of receiving Data packets.

When adjusting the size of the sending window, a preset threshold ssthresh needs to be used. The size of the threshold can be determined according to actual needs. In this embodiment, the size is 512.

Iswnd is initialized to 1, and then adjusted according to the CIB in the received Data packet. Table 1 lists the adjustment method of Iswnd for four different CIB values in one embodiment.

Table 1

In other embodiments, when adjusting the value of Iswnd, the magnitude of each adjustment is not limited to the values mentioned in the above table, and can be modified according to actual needs. For example, when the value of CIB is "10", Iswnd = Iswnd _old -0.5; when the CIB value is "11", Iswnd = Iswnd _old -1. As long as the trend that the value of Iswnd becomes larger when the congestion state is "excellent" and becomes smaller when the congestion state is "bad" is generally satisfied.

In order to verify the specific effect of this method, the applicant verified the method based on the OMNet++ simulation platform, and the experimental topology is shown in Figure 5.

The three clients request three files respectively, and the file size is 10MB, but the time to start the request is different: 0 seconds, 4 seconds, and 8 seconds respectively. In the experiment, the size of the Interest packet and the size of the Data packet adopt the standard size in CCN: 512B and 4KB respectively. The buffer size (B) of each router interface is set to the size of 500 Data packets (ie 500×4KB=2000KB), maxRatio=0.7, minRatio=0.3, ssthresh=128. In the experiment, the link bandwidth between Router1 and Router2 is relatively small (10Mbps), so Router2 may be congested.

The experimental results record the size change of each terminal's Interest sending window (Iswnd) and the data receiving rate, the change of Router2 queue length and the data sending rate. The experimental results are shown in Figure 6. It can be seen from the figure that after the third terminal starts to initiate a request, the Iswnd of the first two terminals are appropriately reduced to avoid sending too many Interests. At the same time, Iswnd is relatively balanced overall, without large fluctuations, which avoids large fluctuations in the sending rate and can improve user experience. It can be seen from Figure 7 that during the experiment, the size of each virtual queue was relatively balanced, and the total queue size never exceeded the maximum value (the maximum value was set to 500), and no packet loss occurred in the experiment. It can be seen from Figure 8 that in the experiment, each terminal can always allocate bandwidth fairly, and the sending bandwidth of the router is also close to its maximum value, and the bandwidth utilization rate is high.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than limit them. Although the present invention has been described in detail with reference to the embodiments, those skilled in the art should understand that modifications or equivalent replacements to the technical solutions of the present invention do not depart from the spirit and scope of the technical solutions of the present invention, and all of them should be included in the scope of the present invention. within the scope of the claims.

Claims (4)

1. A congestion control method based on content-centric network, comprising: Step 1), calculate the CIB value for the Data packet to be sent in the router; wherein, the CIB value is used to reflect the current congestion level of the router; there are N CIB values, and some of the CIB values represent that the client decreases Interest sending window, some CIB values indicate that the client increases the Interest sending window; the calculation CIB value includes: setting M thresholds, using the M thresholds to divide the size of the queue length into N intervals, each interval Corresponding to a CIB value, where N=M+1; The CIB value is represented by 2 bits, including "00", "01", "10", and "11", which are respectively used to indicate that the current congestion level of the router is "excellent", "good", "medium" and "bad". "; the calculated CIB value includes: Step 1-1), calculating the virtual queue maximum length threshold maxQ and the virtual queue length minimum threshold minQ; $\left\{\begin{array}{c}\mathrm{<span\; class="notranslate">\; max</span>}\mathrm{<span\; class="notranslate">\; Q</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; B</span>}}{\mathrm{<span\; class="notranslate">\; f</span>}}<span\; class="notranslate">\; \&times;</span>\mathrm{<span\; class="notranslate">\; max</span>}\mathrm{<span\; class="notranslate">\; R</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; o</span>}<span\; class="notranslate">\; ,</span>\\ \mathrm{<span\; class="notranslate">\; min</span>}\mathrm{<span\; class="notranslate">\; Q</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; B</span>}}{\mathrm{<span\; class="notranslate">\; f</span>}}<span\; class="notranslate">\; \&times;</span>\mathrm{<span\; class="notranslate">\; min</span>}\mathrm{<span\; class="notranslate">\; R</span>}\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; t</span>}\mathrm{<span\; class="notranslate">\; i</span>}\mathrm{<span\; class="notranslate">\; o</span>}<span\; class="notranslate">\; .</span>\end{array}\right.$ Among them, B represents the cache size of the router, F represents the number of flows contained in the router, and maxRatio and minRatio are two parameters determined according to actual conditions; Step 1-2), smoothing the queue length Q of the flow where the Data packet to be sent is located, to obtain avgQ: ${\mathrm{<span\; class="notranslate">\; wxya</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}<span\; class="notranslate">\; =</span>\left\{\begin{array}{cc}{\mathrm{<span\; class="notranslate">\; Q</span>}}_{\mathrm{<span\; class="notranslate">\; 1</span>}}& \mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; =</span>\mathrm{<span\; class="notranslate">\; 1</span>}\\ <span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; W</span>}}_{\mathrm{<span\; class="notranslate">\; q</span>}}<span\; class="notranslate">\; )</span><span\; class="notranslate">\; *</span>{\mathrm{<span\; class="notranslate">\; wxya</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; 1</span>}}<span\; class="notranslate">\; +</span>{\mathrm{<span\; class="notranslate">\; W</span>}}_{\mathrm{<span\; class="notranslate">\; q</span>}}<span\; class="notranslate">\; *</span>{\mathrm{<span\; class="notranslate">\; Q</span>}}_{\mathrm{<span\; class="notranslate">\; no</span>}}& \mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; ></span>\mathrm{<span\; class="notranslate">\; 1</span>}\end{array}\right.<span\; class="notranslate">\; ;</span>$ Among them, Wq is the weight factor; n indicates the nth calculation of avgQ; Q _n indicates the actual length of the queue when the nth calculation of avgQ is performed; avgQ _n indicates the avgQ obtained during the nth calculation; Step 1-3), compare avgQ with maxQ and minQ, if avgQ≤minQ, then CIB is set to "00", if avgQ≥maxQ, then CIB is set to "11", if minQ<avgQ<maxQ, then execute Next step; Steps 1-4), calculating the marking probability P: $\mathrm{<span\; class="notranslate">\; P</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; a</span>}\mathrm{<span\; class="notranslate">\; v</span>}\mathrm{<span\; class="notranslate">\; g</span>}\mathrm{<span\; class="notranslate">\; Q</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; min</span>}\mathrm{<span\; class="notranslate">\; Q</span>}}{\mathrm{<span\; class="notranslate">\; max</span>}\mathrm{<span\; class="notranslate">\; Q</span>}<span\; class="notranslate">\; -</span>\mathrm{<span\; class="notranslate">\; min</span>}\mathrm{<span\; class="notranslate">\; Q</span>}}$ Then set the CIB to "00" with the probability of (1-P); set the CIB to one of "01" or "10" with the probability of P, if Set CIB to "01", otherwise, set to "10"; Step 2), comparing the CIB value obtained in step 1) with the CIB value before the Data packet to be sent, the CIB value used to represent the worst state of router congestion in the two as the CIB value of the Data packet, Then send the Data packet; Step 3), after the client receives the Data packet, adjust the size of the sending window according to the CIB value of the Data packet; wherein, when the CIB value reflects that the router is relatively idle, increase the size of the sending window, When the CIB value reflects that the router is relatively congested, reduce the size of the sending window. 2. the content-centric network-based congestion control method according to claim 1, is characterized in that, described step 3) comprises: Step 3-1), determine the size of the CIB value, if the CIB value is "00", perform the next step, if the CIB value is "01", perform step 3-3), if the CIB value is "10", perform step 3 -4), if the CIB value is "11", perform steps 3-5); Step 3-2), comparing the current sending window size Iswnd with the preset threshold ssthresh, if Iswnd is smaller than ssthresh, then the sending window size Iswnd becomes larger with the first amplitude value, if the Iswnd is greater than or equal to ssthresh, Then the window size Iswnd becomes larger with the second amplitude value; the first amplitude value is greater than the second amplitude value; Step 3-3), keep sending window size Iswnd unchanged; Step 3-4), the sending window size Iswnd becomes smaller with the third amplitude value; Step 3-5), the sending window size Iswnd becomes smaller with a fourth amplitude value; the absolute value of the fourth amplitude value is greater than the absolute value of the third amplitude value. 3. The congestion control method based on content-centric network according to claim 2, characterized in that, the size of the first amplitude value is 1; the size of the second amplitude value is 1/Iswnd _old , where Iswnd _old represents The size of the original sending window; the size of the third amplitude value is 0.25; the size of the fourth amplitude value is 0.5. 4. The congestion control method based on content-centric network according to claim 1, characterized in that, when the Data packet to be sent by the router has a plurality and belongs to a plurality of streams, in the cache memory of the router, for each stream A virtual queue is maintained, and the router serves each virtual queue cyclically; the service includes: taking out the data packet at the head of the queue from the virtual queue, calculating the CIB value for the Data packet, comparing, and adding to the Data packet.