A ubiquitous network dynamic fair rate cooperative control method

Abstract

The application relates to the technical field of network communication, and discloses a ubiquitous network dynamic fair rate cooperative control method. The method is aimed at the time-varying, topology reconstruction and multi-source and multi-path concurrent characteristics of the space-air-ground-sea integrated network link. A routing node periodically collects indexes such as request satisfaction rate, queue length and multi-hop non-queuing proportion, obtains link transmission quality and dynamically updates an interest packet forwarding table; an upper limit of a queue is derived, a rate shaping value is calculated in combination with an interest packet forwarding proportion, and an over-limit interest packet is implemented with delayed sending; when the interest queue exceeds a threshold value, a congestion feedback value is generated and is sent down with a data packet, and a non-congestion interface obtains released bandwidth through probability compensation; if the whole path is still congested, a consumer performs multiplicative contraction on a sending window according to a feedback proportion. The method can guarantee fair distribution of bandwidth resources, significantly improve throughput and reduce queuing delay.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of network communication, in particular to a ubiquitous network dynamic fair rate cooperative control method. BACKGROUND

[0002] With the continuous advancement of 5G enhancement, B5G transition and 6G vision, the global communication system is evolving from the "ground cell-core network" two-layer architecture to the ubiquitous network form of "space-ground-sea": low-orbit satellite constellation provides large-range low-latency backhaul, high-altitude platform (HAP / UAV) and unmanned aerial vehicle relay construct flexible relay and disaster blind link, vehicle-to-everything (V2X) and edge computing nodes bear near-ground high-concurrent services, and ocean buoys and underwater acoustic networks complete sea monitoring coverage. Due to the diversity of access media, significant differences in link characteristics, and frequent reconstruction of topology, the traditional IP system with endpoint addressing and single-path TCP congestion control as the core faces significant expansion bottlenecks in dynamic fusion scenarios. Named Data Networking (NDN) is considered by the research community as a potential core architecture for future ubiquitous networks, thanks to its three mechanisms: content naming, built-in multi-source multi-path protocol, and on-the-way caching. Around NDN, fundamental issues such as multi-hop routing, cache replacement, and data packet encryption and authentication have made some progress, but "cooperative rate control" to ensure content retrieval fairness, throughput rate, and end-to-end delay under high dynamic and heterogeneous link conditions is still an international frontier hotspot. In recent years, schemes such as HoBHIS, PCON, and JAQMCC have attempted to migrate the RED / CoDel, AIMD ideas of the IP era to the NDN system, and combined with the multi-path capability, they give per-hop queue management or consumer-side window adjustment strategies, providing an experimental basis for unified congestion criteria and cross-domain collaboration.

[0003] Currently, congestion control in NDN is divided into per-hop congestion control and consumer-driven congestion control. Per-hop congestion control adjusts the interest packet forwarding rate of different interfaces according to the difference between the actual forwarding rate of interest packets and the fair rate, achieving fair allocation of bandwidth. However, this type of scheme is limited in terms of fairness and network utilization in multi-source forwarding scenarios, and its forwarding rate limit is relatively strict, making it difficult to maximize network throughput while ensuring fairness. Consumer-driven congestion control usually adds forwarding interface labels to interest packets before they are forwarded by routing nodes. Producers extract node label information from interest packets and encapsulate it in data packets to return to consumers. Consumers obtain a set of paths from multiple sources based on this information and set the interest packet sending rate for each path individually. However, this method can cause path label inflation and increase transmission overhead. At the same time, since content can come from multiple sources (including content sources and caches), the appearance or disappearance of sources makes the routing unstable for a short time, which also leads to inefficient congestion control information. SUMMARY

[0004] (I) The technical problem solved: In view of the deficiencies of the prior art, the present application provides a ubiquitous network dynamic fair rate cooperative control method, which solves the above problems.

[0005] (II) Technical solution: In order to achieve the above object, the present application provides the following technical solution: a ubiquitous network dynamic fair rate cooperative control method, characterized in that the method comprises: S1: In each statistical period, the routing node initializes the interest packet forwarding table, so that the initial forwarding probability of interest packets belonging to the same content stream on all interfaces is equal; and collects link load information to update the interest packet forwarding table; the routing node distributes interest packets to the corresponding interfaces according to the updated interest packet forwarding table; the link load information includes the request satisfaction rate, the next hop data packet queue length and the proportion of non-queued data packets.

[0006] S2: The routing node calculates the shaping rate of each flow according to a set shaping formula based on the maximum congestion threshold of the interface data queue, the interest packet forwarding ratio, the current data queue length of each flow and the round-trip delay; when the interest packet forwarding rate is greater than the shaping rate, the corresponding interest packet is delayed and forwarded until the actual rate is not higher than the shaping rate.

[0007] S3: The routing node sets a minimum value for the interest queue of each flow, calculates a maximum value for the interface interest queue capacity according to the interest packet forwarding ratio And determines the interest queue threshold; when detecting that the interest queue length exceeds the interest queue threshold, a congestion feedback value is generated according to the exceeding ratio and carried to the downstream node with the returned data packet to notify the downstream node to reduce the forwarding of interest packets.

[0008] S4: When the routing node detects the congestion feedback value in the returned data packet, if there is an interface that has not occurred congestion, the forwarding probability of the congested interface is reduced by the congestion feedback value ratio and the released traffic is evenly compensated according to the number of non-congested interfaces; if all interfaces are congested, a correction feedback value is generated according to the maximum of the local feedback value and the feedback value converted according to the upstream forwarding probability, and the correction feedback value is encapsulated in the downstream data packet after recalculating the forwarding probability of each interface according to the correction feedback value.

[0009] S5: The consumer node dynamically adjusts its interest packet sending window according to the congestion feedback value carried in the returned data packet, so as to suppress the interest packet injection rate from the source and cooperate with the network to relieve congestion.

[0010] Further, the S1 comprises the following steps: S11: For each available interface , the routing node initializes the interest packet forwarding table so that the initial forwarding probability of each content flow On all available interfaces is equal.

[0011] S12: Statistics Interface Number of forwarded interest packets and from the interface Number of data packets received and with As the request satisfaction rate .

[0012] S13: Statistics on the PIT table of undetermined interests for the interface Number of unsatisfied interest packages and historical values and with Calculate the next-hop packet queue length ;in, This is a smoothing factor.

[0013] S14: Proportion of unqueued packets with a maximum of 4 upstream nodes extracted from the packet header. and in weighted sum manner Calculate the proportion-weighted sum of unqueued packets .

[0014] S15: Based on the request satisfaction rate according to the following formula... Next-hop packet queue length Weighted sum of proportions of unqueued packets Calculate link transmission quality : ;in, For interface The bandwidth of the corresponding link, These are weighting coefficients, and .

[0015] S16: Normalize the LTQ values ​​according to the following formula to update the interest packet forwarding table: ;In the formula, Indicates the number of available interfaces.

[0016] Furthermore, S14 specifically includes: S141: Obtaining the set in the data packet header. ,in , Indicates the distance to the routing node The proportion of queuing-free packets at hop nodes. S142: Calculate the proportion of queuing-free packets based on the following formula. : ;In the formula, Indicates in the interface Above, targeting the flow of The proportion of packets that skip queues.

[0017] S143: Obtain the number of local unqueued packets from the local statistics of the routing node. Number of forwarded packets The following formula will be used to... Insert the set First, and also delete the old field for the furthest hop count: .

[0018] Furthermore, S2 includes the following step: S21: Setting a queuing delay threshold. The maximum congestion threshold of the interface data queue is calculated according to the following formula. : ;In the formula, This represents the average size of the data packets. It is the bandwidth of the data packet output link.

[0019] S22: Record the interest packet forwarding ratio with an initial value of 1 in the interest packet header. When interest packages are accessed via the interface When forwarding, update the forwarding ratio of the interest package according to the following formula. : ;In the formula, It is a flow Interest packages from the interface The probability of forwarding.

[0020] S23: Update the interest pack forwarding ratio Maximum congestion threshold for interface data queue According to the formula Calculated flow The data packet queue threshold; where, This represents the number of concurrent streams.

[0021] S24: Calculate the shaping rate according to the following formula. When the forwarding rate of interest packets is greater than the shaping rate At that time, subsequent interest packets will be queued and delayed until the interest packet forwarding rate does not exceed the reshaping rate. Then forward it again: ;in, Represents a stream Current data queue length, Fixed parameters for round-trip time delay Set to 0.7; It is the bandwidth of the data packet output link.

[0022] Furthermore, S3 includes the following steps: S31: Setting the minimum value of the interest queue. Calculate the maximum value of the interest queue according to the following formula. : ; where, is the interest queue capacity on the interface, and is a parameter used to adjust the available interest queue size; is the number of concurrent flows.

[0023] S32: Set the interest queue threshold according to the following formula : .

[0024] S33: Monitor the actual interest queue length of the flow in real time. When , calculate the congestion feedback value according to the following formula : .

[0025] S34: Load the congestion feedback value into the congestion feedback field of the return data packet and send it to the downstream node to inform it to reduce the forwarding of interest packets of the corresponding flow.

[0026] Further, the S4 includes the following steps: S41: When the routing node R receives a data packet carrying a congestion feedback value from an interface , calculate and take the maximum value between the congestion feedback value and as the correction feedback value .

[0027] S42: If there is at least one interface that has not received a congestion feedback, update the forwarding probability of the congested interface to , and update each non-congested interface to , and then forward the data packet after setting the congestion feedback field in the data packet to zero; wherein represents the number of interfaces that have not experienced congestion.

[0028] S43: If all interfaces have received a congestion feedback, first adjust the forwarding probability of the interface according to the formula at the downstream node G, then correct the forwarding probability of the congested interface according to the formula in the routing node R, and correct the forwarding probability of the other interfaces according to the formula , and finally retain the correction feedback value .

[0029] Further, the S5 is specifically: when it is detected that the congestion feedback value is greater than zero, multiply the interest packet sending window of the consumer by the following formula : .

[0030] (Three) beneficial effects: compared with the prior art, the present application provides a ubiquitous network dynamic fair rate cooperative control method, which has the following beneficial effects: 1, the ubiquitous network dynamic fair rate cooperative control method, the routing node is based on multi-dimensional link transmission quality LTQ real-time update forwarding probability and consumer side interest package sending window cooperative adjustment, thereby reaching the effect of actively inhibiting queue expansion in dynamic topology, effectively avoiding congestion; first, the request satisfaction rate, the next hop queue length and the multi-hop non-queue ratio are used to construct the LTQ index, and the interest packets are periodically guided to the surplus link according to the probability; at the same time, the maximum data packet queue derived from the queuing delay threshold is referenced at the interface side, and the interest packet forwarding ratio IFP is subdivided to each flow, and the interest packet rate is limited by the rate shaping formula. If the queue threshold still appears, the routing node immediately generates a congestion feedback based on the exceeding ratio, and reallocates the interface probability according to the "bad road reduces weight, good road compensates" strategy; when the congestion feedback is transmitted to the consumer end, the source window is multiplied to shrink, and the remaining congestion amount is reduced from the root. The double-loop closed control makes the link queuing delay and packet loss rate maintain in the controllable interval, which significantly reduces the probability of network congestion triggering.

[0031] 2, the ubiquitous network dynamic fair rate cooperative control method, through the joint action of rate shaping speed limit and forwarding probability dynamic compensation, the effect of higher throughput is realized while ensuring the fair allocation of ubiquitous network resources; shaping ensures the minimum bandwidth share of different content streams on each domain link; the probability compensation link only quantitatively reduces the weight of the blocked interface when single-path congestion is detected, and releases the flow to the non-congested interface for average compensation, so that the bandwidth of high-quality link is not idle. BRIEF DESCRIPTION OF DRAWINGS

[0032] Figure 1 It is a system flowchart of the method of the present application.

[0033] Figure 2 It is a system framework diagram of the method of the present application.

[0034] Figure 3 It is an NDN network interest packet and data packet format diagram of the method of the present application.

[0035] Figure 4 It is a rate shaping process diagram of the method of the present application.

[0036] Figure 5 It is a S4 flowchart of the method of the present application.

[0037] Figure 6 It is a flow comparison diagram before and after the forwarding probability adjustment of the method of the present application.

[0038] Figure 7 Figure 1 is a single-hop topology experimental scenario for the method of the present application.

[0039] Figure 8 Figure 2 is a dumbbell topology experimental scenario for the method of the present application.

[0040] Figure 9 Figure 3 is an iridium star topology experimental scenario for the method of the present application.

[0041] Figure 10 Figure 4 is the throughput result of the method of the present application in a single-hop topology experimental scenario.

[0042] Figure 11 Figure 5 is the throughput result of the method of the present application in a dumbbell topology experimental scenario.

[0043] Figure 12 Figure 6 is the fairness index result of the method of the present application in a dumbbell topology experimental scenario.

[0044] Figure 13 Figure 7 is the average throughput result of the method of the present application in an iridium star topology experimental scenario. DETAILED DESCRIPTION

[0045] The technical solutions in the embodiments of the present application will be apparently and completely described below with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all the other embodiments obtained by those skilled in the art without creative work belong to the protection scope of the present application.

[0046] In order to make the persons in the technical field better understand the present application, the present application will be further described below with the drawings and specific embodiments.

[0047] Please refer to Figures 1-3 , Figure 1 Figure 8 is a system flowchart of the method of the present application; Figure 2 Figure 9 is a system framework diagram of the method of the present application; Figure 3 Figure 10 is a NDN network interest packet and data packet format diagram of the method of the present application; the present application discloses a ubiquitous network dynamic fair rate cooperative control method, which comprises the following steps: S1: in each statistical period, a routing node initializes an interest packet forwarding table, so that the initial forwarding probability of the interest packets belonging to the same content stream on all interfaces is equal; and collects link load information to update the interest packet forwarding table; the routing node distributes the interest packets to the corresponding interfaces according to the updated interest packet forwarding table; the link load information comprises a request satisfaction rate, a next hop data packet queue length and a proportion weighted sum of non-queuing data packets.

[0048] S2: The routing node calculates the shaping rate of each flow according to a set shaping formula based on the interface data queue maximum congestion threshold, the interest packet forwarding ratio, the current data queue length of each flow and the round-trip delay; when the interest packet forwarding rate is greater than the shaping rate, the corresponding interest packet is delayed for forwarding until the actual rate is not higher than the shaping rate and then the forwarding is resumed.

[0049] S3: The routing node sets the minimum value for the interest queue of each flow, calculates the maximum value of the interface interest queue capacity according to the interest packet forwarding ratio and determines the interest queue threshold; when detecting that the interest queue length exceeds the interest queue threshold, a congestion feedback value is generated according to the exceeding ratio and is carried to the downstream node with the returned data packet to inform the downstream node to reduce the forwarding of the interest packet.

[0050] S4: When the routing node detects the congestion feedback value in the returned data packet, if there is an interface that has not occurred congestion, the forwarding probability of the congested interface is reduced according to the congestion feedback value ratio and the released traffic is compensated according to the number of non-congested interfaces; if all interfaces are congested, a correction feedback value is generated according to the maximum of the local feedback value and the feedback value converted according to the upstream forwarding probability, and the correction feedback value is encapsulated in the downstream data packet after the forwarding probability of each interface is recalculated according to the correction feedback value.

[0051] S5: The consumer node dynamically adjusts the interest packet sending window according to the congestion feedback value carried in the returned data packet to suppress the interest packet injection rate from the source and relieve the congestion of the network.

[0052] Further, the S1 includes the following steps: S11: For each available interface , the routing node initializes the interest packet forwarding table, so that the initial forwarding probability on all available interfaces is equal.

[0053] S12: The number of forwarded interest packets to the interface and the number of data packets received from the interface are counted , and the ratio of the number of interest packets to the number of data packets is taken as the request satisfaction rate .

[0054] S13: The number of unsatisfied interest packets in the pending interest table PIT for the interface and the historical value are counted , and the next hop data packet queue length is calculated ; wherein, is a smoothing factor.

[0055] ​​​S14: Extract the non-queued packet ratio of the most 4-hop upstream node from the data packet header , and calculate the non-queued packet ratio weighted sum in a weighted sum manner .

[0056] S15: Calculate the link transmission quality based on the request satisfaction rate , the next-hop data packet queue length , and the non-queued packet ratio weighted sum , according to the following formula: ; wherein, is the bandwidth of the interface corresponding to the link, is the weighted coefficient, and .

[0057] S16: Normalize the LTQ value to update the interest packet forwarding table according to the following formula: ; in the formula, represents the number of available interfaces.

[0058] Further, S14 is specifically S141: Obtain the set from the data packet header, wherein , represents the non-queued packet ratio of the node at a distance of hops from the routing node. S142: Calculate the non-queued packet ratio sum based on the following formula: ; in the formula, represents the non-queued packet ratio sum of hops for flow on interface .

[0059] S143: Obtain the local non-queued packet number and the number of forwarded packets from the local statistical data of the routing node, insert at the first position of the set according to the following formula, and delete the old field of the farthest hop number: .

[0060] Specifically, in order to further understand S1 and its sub-steps, in the initial stage, the routing node lacks knowledge of the status of the ubiquitous network, therefore, in order to collect the load information of all links to evaluate the transmission performance of the interface, the same forwarding probability is allocated to each available forwarding interface, so that the interest packet can be forwarded to all available interfaces. For example, when there are three message flows and three forwarding interfaces, the interest packet forwarding table can be represented as follows: ​​.

[0061] In fact, , m message streams represent the set of content names requested by consumers, The set of forwarding interfaces available for forwarding interest packets. The elements in the forwarding table represent the flow The probability that an interest packet is forwarded to interface , wherein, .

[0062] After running for a period of time, the routing node collects the necessary load information to update the forwarding table. First, the number of data packets received on interface in the time period , and the number of interest packets forwarded from interface in the time period are counted, then the request satisfaction rate , which represents the satisfaction of interest packets on the interface (the completion of the request), is calculated; then the next hop data packet queue length , which represents the data packet queue length of the next hop node corresponding to the interface, is calculated; the non-queued data packet ratio aims to estimate the actual unused capacity in the link, specifically, the value represents the proportion of data packets that are enqueued and then dequeued directly. Considering that the delay of the ubiquitous network is long, the field only records the non-queued data packet ratio from the upstream at most four-hop nodes, rather than from the producer node to the metric estimation node (the current node), to ensure the availability of information. If the routing node receives data packets about flow from multiple interfaces at the same time, that is, multiple sets, the value is calculated according to these data packets respectively, and the set with the largest value is selected to be transmitted downstream.

[0063] Further, please refer to Figure 4 , Figure 4 for the rate shaping process of the method of the present application; the figure fully shows how to suppress the interest packet flood peak through rate shaping on a single interface first, and then trigger the FBTD congestion feedback when the queue is still above the threshold; the S2 includes the following steps: S21: set the queuing delay threshold , and calculate the maximum congestion threshold of the interface data queue according to the following formula : ; in the formula, is the average size of data packets, is the data packet output link bandwidth.

[0064] S22: Record the interest packet forwarding ratio with an initial value of 1 in the interest packet header. When interest packages are accessed via the interface When forwarding, update the forwarding ratio of the interest package according to the following formula. : ;In the formula, It is a flow Interest packages from the interface The probability of forwarding.

[0065] S23: Update the interest package forwarding ratio Maximum congestion threshold for interface data queue According to the formula Calculated flow The data packet queue threshold; where, This represents the number of concurrent streams.

[0066] S24: Calculate the shaping rate according to the following formula. When the forwarding rate of interest packets is greater than the shaping rate At that time, subsequent interest packets will be queued and delayed until the interest packet forwarding rate does not exceed the reshaping rate. Then forward it again: ;in, Represents a stream Current data queue length, Fixed parameters for round-trip time delay Set to 0.7; It is the bandwidth of the data packet output link.

[0067] S3 includes the following steps: S31: Set the minimum value of the interest queue. Calculate the maximum value of the interest queue according to the following formula. : ;In the formula, The parameter represents the capacity of the interest queue on the interface. Used to adjust the size of the available interest queue; This represents the number of concurrent streams.

[0068] S32: Set the interest queue threshold according to the following formula. : .

[0069] S33: Real-time monitoring of flow Actual length of the interest queue ,when At that time, the congestion feedback value is calculated according to the following formula. : .

[0070] S34: Transfer the congestion feedback value The congestion feedback field of the returned data packet is loaded and sent to a downstream node to inform it to reduce the forwarding of interest packets of the corresponding flow.

[0071] Specifically, for further understanding S2, S3 and their sub-steps, in S2, in order to avoid overflow of the data packet queue, the data packet queue of the current interface of the ubiquitous network routing node has a last data packet with a queuing time less than , is a data packet queuing time threshold (referring to the CoDel algorithm, the threshold is 5 ms), when the threshold is exceeded, congestion may be caused, and therefore , further calculation is as follows: Therefore, the maximum congestion threshold of the interface data queue can be obtained as .

[0072] In order to ensure the fairness of bandwidth resource allocation, the application adds an interest packet forwarding ratio (IFP) field in the interest packet. The interest packet forwarding ratio (IFP) field indicates that the interest packets of the flow forwarded by the node account for the proportion of the total interest packets of the flow issued by the consumer, wherein The field implies the number of interest packets that the routing node will receive from the consumer. The initial value of the field is set to 1 in the consumer, and each time the interest packet passes through the interface of the node for forwarding, the interest packet field is updated according to the formula in S22 step; the shaping rate of the flow is calculated according to the formula in S24 step When the actual forwarding rate of the interest packet on the routing node exceeds the shaping rate , the routing node will delay the forwarding of the interest packet, and the delayed interest packet will be queued in the interest packet. Until the actual forwarding rate of the interest packet is less than the shaping rate , the interest packet is forwarded.

[0073] In the ubiquitous network routing node, only the rate shaping mechanism cannot fundamentally solve the congestion, and the generation of interest packets must be reduced, so it is necessary to inform the consumer of the ubiquitous network to reduce the number of interest packets sent to the network. Therefore, in S3, the routing node maintains two values for the interest packet queue of the flow : and . is set to 1, is the maximum available queue, and then the interest queue threshold of the flow is set; the routing node will monitor the interest queue length, and when the interest queue length of the flow When the interest queue length exceeds the interest queue congestion threshold, the flow must reduce the number of interest packets to prevent congestion . But the routing node cannot control the arrival of interest packets and can only be passive reception. For this purpose, the present invention adds a congestion feedback field in the data packet header to inform the downstream to reduce the forwarding of interest packets.

[0074] Further, please refer to Figure 5 and Figure 6 , Figure 5 S4 flowchart of the method of the present invention; Figure 6 is a schematic diagram of the flow comparison before and after the adjustment of the forwarding probability of the method of the present invention; S4 includes the following steps: S41: when the routing node R receives a data packet carrying a congestion feedback value from the interface , calculate and take the maximum value between the congestion feedback value and as the correction feedback value .

[0075] S42: if there is at least one interface that has not received congestion feedback, update the forwarding probability of the congested interface to , and update each non-congested interface to , and then forward the data packet after setting the congestion feedback field in the data packet to zero; wherein represents the number of interfaces that have not occurred congestion.

[0076] S43: if all interfaces have received congestion feedback, first adjust the forwarding probability of the interface according to the formula at the downstream node G, then correct the forwarding probability of the congested interface in the routing node R according to the formula , and correct the forwarding probability of the other interfaces according to the formula ; finally, retain the correction feedback value .

[0077] Specifically, for further understanding S4 and its sub-steps, when the data packet FBTD field value received by the ubiquitous network routing node is greater than zero, the node first checks whether there is another interface that can divert the traffic. If the routing node has more than one available forwarding interface, but there is no interface that can divert the traffic (i.e. the available interfaces all receive congestion feedback within the time interval), the congestion notification is continued to be forwarded, and the influence of the reduced traffic downstream of the routing node on the non-congestion interface upstream of the routing node needs to be considered. The present application increases the traffic of the non-congestion interface by compensating the forwarding probability of the non-congestion interface, so as to realize that the traffic of the non-congestion interface is not affected when the downstream node receives the congestion feedback and reduces the traffic. If the routing node R itself needs to adjust the traffic (the node detects the congestion feedback on the local forwarding interface), the maximum value of the received congestion feedback and the local feedback value is selected as the traffic that needs to be adjusted. If the routing node receives the congestion feedback, and there is no other interface that can forward the interest packet, or the other forwarding interfaces all receive the congestion feedback within the time interval, the congestion feedback is continued to be transmitted downstream. If the consumer receives the congestion feedback, the interest packet sending rate will be reduced to alleviate the network congestion.

[0078] Further, S5 is specifically: when it is detected that the congestion feedback value is greater than zero, the interest packet sending window of the consumer is multiplicatively shrunk according to the following formula : .

[0079] When the data packet FBTD field received by the consumer of the ubiquitous network is greater than zero, it indicates that the forwarding probability adjustment in the network cannot alleviate the congestion, and the consumer adjusts the window size according to the FBTD. Although the consumer window is reduced, the routing node avoids unnecessary reduction of the traffic of the non-congestion path by compensating the forwarding probability of the non-congestion interface.

[0080] To further verify the beneficial effects of the present application, the method of the present application is DFRCC. Please refer to Figures 7-13 , Figure 7 is the single-path topology experimental scene graph of the method of the present application; Figure 8 is the dumbbell topology experimental scene graph of the method of the present application; Figure 9 is the iridium star topology experimental scene graph of the method of the present application; Figure 10 is the throughput result of the method of the present application in the single-path topology experimental scene; Figure 11 is the throughput result of the method of the present application in the dumbbell topology experimental scene; Figure 12 is the fairness index result of the method of the present application in the dumbbell topology experimental scene; Figure 13 is the average throughput result of the method of the present application in the iridium star topology experimental scene. The smoothing factor is 0.6, and the weighting coefficient The results show that the throughput of DFRCC and HoBHIS converges better than PCON and JAQMCC under different link delays. Except for DFRCC and HoBHIS, other schemes are greatly affected by the change of link delay. Second, the fairness of network resource cost allocation in the multi-path transmission scenario is verified using the dumbbell topology. The results show that DFRCC can effectively achieve the fair allocation of bandwidth between consumer C1 and consumer C2. Specifically, the fairness indexes of DFRCC, PCON, HoBHIS, and JAQMCC are 0.99, 0.96, 0.53, 0.78, and 0.90, respectively. Third, the throughput of consumers under different schemes is compared in the iridium star topology. Taking the arctic ground station (AGS) as an example, compared with PCON, HoBHIS, and JAQMCC, the average throughput of DFRCC is increased by 102.70%, 8.69%, 31.94%, and 48.13%, respectively. Taking the Kashgar ground station (KGS) as an example, compared with PCON, HoBHIS, MIRCC, and JAQMCC, the average throughput of DFRCC is increased by 16.15%, 83.15%, 3.34%, and 101.61%, respectively.

[0081] It should be noted that the relational terms herein such as first and second and the like are used solely to distinguish one from another entity or action without necessarily requiring or implying any actual relationship or order between such entities or actions. Moreover, the terms "comprises", "comprising", or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but can include other elements not expressly listed or inherent to such process, method, article, or apparatus. An element proceeded by "comprises... a" does not, without more constraints, exclude the existence of additional identical elements in the process, method, article, or apparatus that comprises the element.

[0082] While the embodiments of the application have been illustrated and described, it will be understood by those skilled in the art that various changes, modifications, substitutions, and alterations can be made therein without departing from the spirit and scope of the application in its broadest form. The scope of the application is defined by the appended claims and their equivalents.

Claims

1. A method for dynamic fair rate coordination in ubiquitous networks, characterized in that, The method comprises: S1: in each statistical period, the routing node initializes the interest packet forwarding table so that the initial forwarding probability of the interest packets belonging to the same content stream on all interfaces is equal; and collects link load information to update the interest packet forwarding table; the routing node distributes the interest packets to the corresponding interfaces according to the updated interest packet forwarding table; the link load information comprises a request satisfaction rate, a next-hop data packet queue length and a proportion-weighted sum of non-queued data packets; S2: the routing node calculates the shaping rate of each flow according to a set shaping formula based on an interface data queue maximum congestion threshold, an interest packet forwarding proportion, a current data queue length of each flow and a round-trip delay; when the interest packet forwarding rate is greater than the shaping rate, the corresponding interest packet is delayed to be forwarded until the actual rate is not higher than the shaping rate; S3: the routing node sets a minimum value for the interest queue of each flow, calculates a maximum value for the interface interest queue capacity according to the interest packet forwarding proportion and determines an interest queue threshold; when it is detected that the interest queue length exceeds the interest queue threshold, a congestion feedback value is generated according to an exceeding proportion and is carried in a returned data packet to a downstream node to inform the downstream node to reduce the forwarding of the interest packets; S4: when the routing node detects the congestion feedback value in the returned data packet, if there is an interface that has not occurred congestion, the forwarding probability of the congested interface is reduced according to the congestion feedback value proportion and the released traffic is compensated according to the number of non-congested interfaces; if all interfaces are congested, a modified feedback value is generated according to the maximum of the local feedback value and the feedback value converted according to the upstream forwarding probability, and the modified feedback value is encapsulated in the downstream data packet after the forwarding probability of each interface is recalculated according to the modified feedback value; S5: the consumer node dynamically adjusts the interest packet sending window according to the congestion feedback value carried in the returned data packet, so as to suppress the interest packet injection rate from the source and relieve the congestion of the network.

2. The method of claim 1, wherein, Said S1 Includes the following steps: S11: For each available interface The routing node initializes the interest packet forwarding table so that each content stream... The initial forwarding probability is equal on all available interfaces; S12: Statistical direction to interface Number of forwarded interest packets and from the interface Number of data packets received and with As the request satisfaction rate S13: Statistics on the PIT table of undetermined interests for the interface Number of unsatisfied interest packages and historical values and with Calculate the next-hop packet queue length ;in, S14: The proportion of unqueued packets with a maximum of 4 upstream nodes extracted from the packet header; (S14 is the smoothing factor.) and in weighted sum manner Calculate the proportion-weighted sum of unqueued packets S15: Based on the request satisfaction rate according to the following formula. Next-hop packet queue length Weighted sum of proportions of unqueued packets Calculate link transmission quality : ;in, For interface The bandwidth of the corresponding link, These are weighting coefficients, and S16: Normalize the LTQ values ​​according to the following formula to update the interest packet forwarding table: ;In the formula, Indicates the number of available interfaces.

3. The method of claim 2, wherein, The S14 specifically is: S141: obtaining a set in the data packet header wherein , represents the distance from the routing node the queuing-free data packet ratio of the hop node S142: Calculate the proportion of queuless data packets based on the following formula. : In the formula, Indicates in the interface Above, targeting the flow of S143: Obtain the local number of no-queue packets from the local statistics of the routing node. Number of forwarded packets The following formula will be used to... Insert the set First, and also delete the old field for the furthest hop count: .

4. The method of claim 1, wherein, Said S2 Includes the following steps: S21: Set queuing delay threshold The maximum congestion threshold of the interface data queue is calculated according to the following formula. : ;In the formula, This represents the average size of the data packets. This refers to the output link bandwidth of the data packet; S22: Record the interest packet forwarding ratio with an initial value of 1 in the interest packet header. When interest packages are accessed via the interface When forwarding, update the forwarding ratio of the interest package according to the following formula. : ;In the formula, It is a flow Interest packages from the interface Probability of forwarding; S23: Proportion of forwarding the updated interest package Maximum congestion threshold for interface data queue According to the formula Calculated flow The data packet queue threshold; where, S24: Calculate the shaping rate according to the following formula, where S24 is the number of concurrent flows. When the forwarding rate of interest packets is greater than the shaping rate At that time, subsequent interest packets will be queued and delayed until the interest packet forwarding rate does not exceed the reshaping rate. Then forward it again: ;in, Represents a stream Current data queue length, Fixed parameters for round-trip time delay Set to 0.7; It is the bandwidth of the data packet output link.

5. The method of claim 1, wherein, Said S3 comprising the following steps: S31: setting the interest queue minimum value , calculating the interest queue maximum value according to the following formula: ; in the formula, is the interest queue capacity on the interface, the parameter is used to adjust the available interest queue size; is the number of concurrent flows; S32: setting the interest queue threshold value according to the following formula: ; S33: monitoring the actual interest queue length of the flow in real time, when , calculating the congestion feedback value according to the following formula: ; S34: loading the congestion feedback value into the congestion feedback field of the return data packet and sending it to the downstream node to inform it to reduce the interest packet forwarding of the corresponding flow.

6. The method of claim 1, wherein, Said S4 Includes the following steps: S41: When routing node R from interface Received congestion feedback value When calculating data packets And retrieve congestion feedback values and The maximum value between them is used as the correction feedback value. S42: If at least one interface has not received congestion feedback, then the congested interface will be... The forwarding probability is updated to And for each non-congested interface Updated to Then, the congestion feedback field in the data packet is set to zero before forwarding; among which, This indicates the number of interfaces that are not congested; S43: If all interfaces have received congestion feedback, then first, in the downstream node G, according to the formula... Adjust the interface The forwarding probability is then calculated in the routing node R according to the formula. Correct the forwarding probability of congested interfaces, and simultaneously follow the formula. Correct the forwarding probability of other interfaces; finally, retain the correction feedback value. .

7. The method of claim 1, wherein the method further comprises: The S5 is specifically: when detecting that the congestion feedback value is greater than zero, multiplying the consumer's interest package sending window according to the following formula : .

Images