CN114401226B - Method and system for controlling route flow of stream media data - Google Patents
Method and system for controlling route flow of stream media data Download PDFInfo
- Publication number
- CN114401226B CN114401226B CN202210154844.6A CN202210154844A CN114401226B CN 114401226 B CN114401226 B CN 114401226B CN 202210154844 A CN202210154844 A CN 202210154844A CN 114401226 B CN114401226 B CN 114401226B
- Authority
- CN
- China
- Prior art keywords
- media data
- bucket
- streaming media
- priority
- node
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 47
- 230000008569 process Effects 0.000 claims abstract description 23
- 238000002347 injection Methods 0.000 claims description 20
- 239000007924 injection Substances 0.000 claims description 20
- 230000007704 transition Effects 0.000 claims description 17
- HRULVFRXEOZUMJ-UHFFFAOYSA-K potassium;disodium;2-(4-chloro-2-methylphenoxy)propanoate;methyl-dioxido-oxo-$l^{5}-arsane Chemical compound [Na+].[Na+].[K+].C[As]([O-])([O-])=O.[O-]C(=O)C(C)OC1=CC=C(Cl)C=C1C HRULVFRXEOZUMJ-UHFFFAOYSA-K 0.000 claims description 7
- 238000012163 sequencing technique Methods 0.000 claims description 6
- 230000001172 regenerating effect Effects 0.000 claims description 3
- 230000006735 deficit Effects 0.000 claims 1
- 230000006870 function Effects 0.000 description 6
- 230000005540 biological transmission Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000007246 mechanism Effects 0.000 description 2
- 238000009825 accumulation Methods 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 230000001010 compromised effect Effects 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000009977 dual effect Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 238000007726 management method Methods 0.000 description 1
- 238000013441 quality evaluation Methods 0.000 description 1
- 238000007493 shaping process Methods 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L47/00—Traffic control in data switching networks
- H04L47/10—Flow control; Congestion control
- H04L47/215—Flow control; Congestion control using token-bucket
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L45/00—Routing or path finding of packets in data switching networks
- H04L45/302—Route determination based on requested QoS
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L45/00—Routing or path finding of packets in data switching networks
- H04L45/302—Route determination based on requested QoS
- H04L45/306—Route determination based on the nature of the carried application
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
Abstract
The invention relates to a route forwarding control method and a system of stream media data, which are characterized in that each router node in a network classifies received stream media data according to a set priority rule and puts the received stream media data into various corresponding cache queues; the method comprises the steps that a token bucket and an overflow bucket are set for a cache queue of a router, the generated count value of the overflow bucket is counted in a preset observation window, the use state of bandwidth resources of a local router node in a current period of time can be determined, for routing table items which are sent by different neighbor nodes and have the same destination address and the shortest distance, the neighbor node with the highest overflow bucket count is used as a strong connection node, the priority of the strong connection node in the routing process is higher than that of other common nodes which are not in strong connection, more data flows can be distributed to the strong connection node with lower bandwidth utilization rate in the routing path selection process, the effective distribution of streaming media data is realized, and the congestion problem is reduced.
Description
Technical Field
The invention relates to the field of internet streaming media service, in particular to a method and a system for controlling routing flow of streaming media data.
Background
Streaming media refers to media formats such as audio, video or multimedia files that are played over the internet in a streaming manner. The key technology for realizing streaming media is streaming, wherein streaming mainly refers to that multimedia files such as whole audio, video, three-dimensional media and the like are analyzed into compressed packets by a specific compression mode, and the compressed packets are transmitted in a segmented mode in a network in a streaming mode, so that the service of broadcasting under the client terminal can be realized, and longer delay brought by downloading the whole media files to the local place by adopting a traditional transmission mode is avoided.
With the rapid development of broadband network technology, streaming media applications are increasingly wide, and real-time playing and high bandwidth requirements make streaming media distribution a great challenge. The streaming media transmission has the characteristics of real-time performance and continuity. Real-time requires that each video frame data transmitted in the network must arrive at the receiver before a specific play time limit; continuity requires that video frame data must be played in a certain order, which presents a greater challenge to the network in order to meet these characteristics.
In order to guarantee the quality of service (QoS) requirements of streaming media, how to effectively control network traffic under the condition of limited network resources is one of the main factors for meeting QoS requirements, so as to provide a service of "best effort delivery" and reduce the occurrence frequency of congestion. There are two types of smoother current limiting algorithms in common use today: a Leaky Bucket algorithm (leak Bucket) and a Token Bucket algorithm (Token Bucket). Because the leak rate of the leaky bucket is a fixed parameter, it lacks efficiency for traffic with bursty characteristics. For many application scenarios, the traffic actually streamed in different periods cannot always be consistent, especially in the case of unstable network state, and the traffic with burst characteristics is unavoidable. The token bucket algorithm can solve this problem exactly as compared to the leaky bucket algorithm.
The principle of the token bucket algorithm is that the system will put tokens into the bucket at a constant rate, and if the request needs to be processed, it will need to first get a token from the bucket, and if no token is desirable in the bucket, it will refuse to serve. If the traffic of the request service is bursty, all tokens remained in the bucket can be taken away at one time, thereby effectively solving the bursty event. Another benefit of the token bucket is that the speed can be conveniently changed, and once the speed needs to be increased, the speed of the tokens placed in the bucket is increased as needed. A certain number of tokens are typically added to the bucket at regular time, and some variant algorithms calculate in real time the number of tokens that should be added.
However, the token bucket algorithm does not consider the case where local network resources are idle waiting. Some routing forwarding nodes in the network are in idle states for a long time, i.e. no or a small amount of traffic passes in a certain period of time, so that bandwidth resources of the network nodes are wasted, and other routing forwarding nodes are in busy states for a long time, so that the resource utilization is unbalanced. The token bucket can only control the data flow when a certain node is congested, but cannot distribute the data transmission task to the network node in idle state in the network, and the data flow limitation relieves the load burden in the local area network at the cost of reducing the service quality, so the token bucket does not belong to an ideal streaming media data transmission strategy.
Disclosure of Invention
In order to solve the technical problems, the invention changes the traditional single data flow limiting mode, uses an improved routing protocol to provide service for forwarding the streaming media data, realizes the dual functions of data flow control and routing forwarding control, greatly reduces the congestion degree and satisfies the service quality of the media data of each level. The invention provides a route forwarding control method of stream media data, which specifically comprises the following steps:
each router node in the network classifies the received streaming media data according to a set priority rule and puts the streaming media data into various corresponding cache queues;
setting a token bucket and an overflow bucket for each cache queue, setting an upper limit value for loading the token bucket into the token and a token injection rate according to the priority of the cache queue, injecting the token overflowed from the token bucket into the overflow bucket, and generating an empty overflow bucket after the previous overflow bucket is full, and continuously injecting the token;
counting the counts of all levels of stream media data overflow barrels in a router node in a preset period, and broadcasting all levels of count values outwards along with local routing table data to be forwarded;
in the process of updating a local routing table, a routing path algorithm utilizing a distance vector selects a neighbor node with the highest overflow bucket count value at each level as a strong connection node with corresponding priority for routing table items which are sent by different neighbor nodes and have the same destination address and shortest distance, and updates the routing table items sent by the strong connection node to the local;
And forwarding the streaming media data of the corresponding level according to the path designated by the updated routing table item.
Further preferably, the priority rule adopts an IP priority or DSCP classification policy supporting QoS, and classifies the priority class of the streaming media data with the number marked by the differentiated services field in the IP packet header.
Further preferably, the counting rule of each stage of overflow buckets of streaming media data specifically includes:
setting an observation window consisting of n time cells, each time cell being a time period t, moving the observation window forward by one time cell every time period t,
t x ＝a x /r x
wherein a represents the upper limit value of the token bucket which can be filled with the token, r represents the injection rate of the token bucket, x represents the number of stages of the priority corresponding to the streaming media data, and the upper limit value of the overflow bucket which can be filled with the token is the same as the token bucket of the corresponding priority class;
counting overflow barrel counts generated by each time cell in the observation window in real time, and summing all count values:
wherein c token_i Representing the overflow amount of the token in the period of the ith time cell, C bucket Representing the sum of overflow bucket counts generated by all time cells in the observation window, and adding the count value C to bucket As a result of the count of the overflow buckets of the x-level streaming media data.
Further preferably, the counting rule of each stage of overflow buckets of streaming media data further comprises a flow breaking process:
counting the flow f of the x-level streaming media data forwarded by the local router node in the period of the ith time cell x_i ；
Flow f x_i And the value a x Comparing if f x_i ≤a x Then calculate the valueAs the flow loss of the ith time cell, the loss difference of the observation window is calculated by the following formula:
if D bucket If the value is negative, returning a zero value as the counting result of the overflow bucket of the x-level streaming media data, otherwise, directly returning a difference value D bucket As the counting result of the x-level stream media data overflow barrel;
if f x_i ＞a x Then calculate the valueAs the flow loss of the ith time cell, the loss difference of the observation window is calculated by the following formula:
wherein k represents f x_i ＞a x The number of consecutive occurrences of the condition before the ith time cell, k.gtoreq.1, if D bucket If the value is negative, returning a zero value as the counting result of the overflow bucket of the x-level streaming media data, otherwise, directly returning a difference value D bucket As a result of the count of the overflow buckets of the x-level streaming media data.
Further preferably, the specific operation procedure of updating the local routing table of the corresponding priority by using the strong connection nodes of each level is as follows:
Receiving a routing table sent by a neighbor node X, wherein the distance value from the neighbor node X to a destination address D is D 1 If the destination address D does not exist in the local routing table, adding the routing table entry into the local routing table;
otherwise, judging whether the neighbor node X is the same as the next hop node Y to the destination address D recorded in the local routing table, and if so, updating the routing table entry of the neighbor node X to the local;
if it is different, the distance value d 1 Distance value D to destination address D recorded in local routing table 2 Comparing, if d 1 +1＜d 2 Updating the routing table entry of the neighbor node X to the local;
otherwise, if d 1 +1＝d 2 Extracting a next-hop node Y to a destination address D recorded in a local routing table, comparing the priority of a neighbor node X with the priority of the next-hop node Y, updating the routing table entry of the neighbor node X to the local if the priority of the neighbor node X is greater than the priority of the next-hop node Y, and not updating the local routing table entry if the priority of the neighbor node X is not greater than the priority of the next-hop node Y;
otherwise, if d 1 +1＞d 2 The local routing table entry is not updated;
the node priority of the common router is less than the node priority of the strong connection.
Further preferably, the method further comprises a dynamic adjustment process of the priority of the streaming media data:
when the overflow bucket count in the observation window of the mth-level streaming media data is lower than a set threshold alpha m And the overflow bucket count generated by the last time cell is zero, further judging that the overflow bucket count in the m-1 level streaming media data observation window is higher than a set threshold alpha m-1 And when the overflow bucket count generated by the last time cell is not zero, the m-th level streaming media data buffer queue is transited to the m-1-th level streaming media data buffer queue, whether the condition meeting the transition is still met or not is repeatedly monitored every interval time t, if yes, the transition state of the m-th level streaming media data buffer queue is maintained, otherwise, the m-th level streaming media data buffer queue is restored, wherein m represents the stage number identification of streaming media data, m is more than or equal to 2, and the higher the stage number is, the lower the priority of the streaming media data is corresponding.
In order to realize the method, the invention also provides a route forwarding control system of the streaming media data, which comprises a plurality of router nodes distributed in a network; the router node supports RIP routing protocol, comprising: the system comprises a classifier, a cache, a token controller, an overflow bucket meter, a scheduler, a routing module and a routing data forwarding module;
A classifier: classifying the received streaming media data by adopting IP priority or DSCP classification rules supporting QoS;
a cache memory: temporarily storing stream media data to be forwarded, and distributing independent buffer spaces for the stream media data with different priority classes to form buffer queues corresponding to the classes;
a token controller: setting a token bucket and an overflow bucket for each cache queue, setting an upper limit value of the token bucket which can be filled with tokens and a token injection rate according to the priority of the cache queue, synchronously generating tokens according to the set token injection rate, injecting the tokens overflowed from the token bucket into the overflow bucket, and regenerating an empty overflow bucket after the current overflow bucket is full;
overflow bucket meter: counting the counts of all levels of stream media data overflow barrels in a router node in a preset period, and broadcasting all levels of count values outwards along with local routing table data to be forwarded;
a scheduler: sequencing the streaming media data of the obtained tokens according to the sequence of the acquisition time, and directing the sequencing queue to the corresponding output port;
and a routing module: in the process of updating a local routing table, a routing path algorithm utilizing a distance vector selects a neighbor node with the highest overflow bucket count value at each level as a strong connection node with corresponding priority for routing table items which are sent by different neighbor nodes and have the same destination address and shortest distance, and updates the routing table items sent by the strong connection node to the local;
A routing data forwarding module: and forwarding the streaming media data of the corresponding level according to the path designated by the updated routing table item.
Further preferably, the device further comprises a clock controller for performing clock synchronization on the token controller to which each priority class belongs.
Further preferably, the system further comprises a priority adjustment module, wherein the priority adjustment module is used for executing transition operation from the low-level streaming media data buffer queue to the high-level streaming media data buffer queue according to the set transition condition, so as to complete dynamic adjustment of the priority of the streaming media data.
The improved method and system for controlling the route forwarding of the streaming media data have the advantages that:
1. the method comprises the steps that a token bucket and an overflow bucket are set for a cache queue of a router, the generated count value of the overflow bucket is counted in a preset observation window, the use state of bandwidth resources of a local router node in a current period of time can be determined, for routing table items which are sent by different neighbor nodes and have the same destination address and the shortest distance, the neighbor node with the highest overflow bucket count is used as a strong connection node, the priority of the strong connection node in the routing process is higher than that of other common nodes which are not in strong connection, more data flows can be distributed to the strong connection node with lower bandwidth utilization rate in the routing path selection process, and effective flow distribution of streaming media data is achieved.
2. The invention adopts an improved RIP protocol algorithm, a router with a strong connection relation is preferentially selected from neighbor nodes under the same path selection condition as a next hop node for forwarding data, the shortest distance which is defined in a distance vector routing path algorithm and is used for reaching a destination address is reserved as a standard, the data delay of the whole network topological structure is reduced as much as possible, the invention also has the shunting function of routing forwarding data, and the flow limiting function of streaming media data is realized by using a token bucket, so that congestion is reduced.
3. The invention respectively carries out overflow bucket counting on each level of stream media data, so different forwarding paths can be generated for different levels of stream media data, namely different next hop nodes and ports can be selected for transmitting data at a certain forwarding node, and the routing control method is more flexible compared with the traditional routing protocol which can only implement all levels of data interaction with only unique routing information because the selection result of the routing path depends on the resource utilization rate of the certain level of data on each adjacent node.
4. According to different streaming media data with different QoS requirements, the streaming media data is classified, in order to enable packets in a high priority queue to have more opportunities to be served, a token bucket algorithm is combined with a weighted fair queuing mechanism WFQ (Weighted Fair Queuing), different token capacities and injection rates are set according to different priority levels of a token bucket, and the streaming media demands of various applications on different service qualities are compatible by adjusting the output and forwarding rates of various streaming media data, so that user experience is improved, and the network has more robust performance robustness.
5. According to the set transition condition, the transition operation from the low-level streaming media data buffer queue to the high-level streaming media data buffer queue is executed, and the dynamic adjustment of the priority of the streaming media data is completed, so that the network resources reserved for the high-quality service can be distributed to the low-level streaming media data for use in an idle state, the throughput of network load is increased, the utilization efficiency and task delivery capacity of the network resources are greatly improved, and the occurrence probability of network congestion is further reduced.
Drawings
Fig. 1 is a flow chart of a method for controlling route forwarding of streaming media data according to the present invention;
FIG. 2 is a schematic diagram of the TOS field of the header fixed part of an IP message;
FIG. 3 is a block diagram of implementing streaming media data streaming based on a modified token bucket algorithm;
FIG. 4 is a schematic view of a view window for overflow bucket counting as provided in the embodiment;
FIG. 5 is a flow chart of operations for updating local routing tables using a modified RIP protocol;
FIG. 6 is a schematic diagram of routing links of router nodes R2 through R10 in one embodiment;
FIG. 7 is a schematic diagram of routing links of router nodes R4 through R9 in another embodiment;
FIG. 8 is a view of a window movement state for overflow bucket counting as set forth in the embodiments;
Fig. 9 is a flow chart of a route forwarding control system provided by the present invention.
Detailed Description
Embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. While certain embodiments of the present disclosure have been shown in the accompanying drawings, it is to be understood that the present disclosure may be embodied in various forms and should not be construed as limited to the embodiments set forth herein, but are provided to provide a more thorough and complete understanding of the present disclosure. It should be understood that the drawings and embodiments of the present disclosure are for illustration purposes only and are not intended to limit the scope of the present disclosure.
As shown in fig. 1, the method for controlling route forwarding of streaming media data provided by the present invention specifically includes:
step 1), each router node in the network classifies received streaming media data according to a set priority rule and puts the streaming media data into various corresponding cache queues;
step 2) setting a token bucket and an overflow bucket for each cache queue, setting an upper limit value of the token bucket which can be filled with tokens and a token injection rate according to the priority of the cache queue, injecting the tokens overflowed from the token bucket into the overflow bucket, and generating an empty overflow bucket after the previous overflow bucket is full, and continuously injecting the tokens;
Step 3) counting the counts of all levels of stream media data overflow barrels in the router node in a preset period of time, and broadcasting all levels of count values outwards along with the local routing table data to be forwarded;
step 4) selecting the neighbor node with the highest overflow bucket count value of each level as a strong connection node with corresponding priority for the routing table items with the same destination address and shortest distance sent by different neighbor nodes in the process of updating the local routing table by using a routing path algorithm of the distance vector, and updating the routing table items sent by the strong connection node to the local;
and 5) forwarding the streaming media data of the corresponding level according to the path designated by the updated routing table item.
Quality of service, relative to network traffic, may be compromised while guaranteeing the quality of service for certain types of traffic. Because network resources are always limited, quality of service requirements may arise whenever there is a situation where network resources are robbed. The invention adopts the traditional QoS service quality evaluation standard to classify the streaming media data with different formats and different authorities. Specifically defined by DiffServ (Differentiated Service) model, in which data for different services are classified according to service requirements, messages are prioritized by class, and then services are provided differentially.
In the invention, the traffic classification is carried out by using an IP priority mode or redefined DSCP differential service code point (Differentiated Services Code Point) in the old IPv4 protocol message so as to expand the compatibility of data classification. Marking a message with the first three bits (i.e., IP priority) of the TOS (Type of service) field of the IP message header can divide the message into at most 2 3 A class; while using DSCP can use the first 6 bits of TOS domain, then at most can be divided into 2 6 A class (as shown in fig. 2). After the message is classified, other QoS characteristics can be applied to different classifications, and the processes of class-based congestion management and traffic shaping are realized.
And independently setting a token bucket and an overflow bucket which are matched with each classification, distributing various independently used buffer spaces in a local buffer, and temporarily storing classified band forwarding packets. If 3, each buffer queue corresponds to a token bucket, the classified packets firstly enter the corresponding buffer queue, then the packets with the earlier sequence get the tokens from the packets, enter the authorized queue, and are transferred out from the local node packets after being scheduled by the relevant forwarding process. If the token bucket of a certain stage is full of tokens, an overflow bucket is generated for storing overflowed tokens, after the current overflow bucket is full, another empty overflow bucket is automatically generated, and the tokens are continuously stored until all the tokens overflowed at the current moment are fully loaded into the overflow bucket.
The weighted fair queuing WFQ plays an important role in the quality of service architecture, and many router products currently incorporate the WFQ scheduling function. However, WFQ can only set different "weights" for streaming media data according to priorities and perform with the "weightsThe service duration of the current data flow does not have the current limiting function, and the port is easy to be congested when the multi-level data is scheduled. The invention integrates the weighting selection mechanism into the token bucket structure, sets different bucket capacities and token generation rates for token buckets serving different priorities, so that service data with high priority can obtain more bandwidth resources in the same time, and can provide a current limiting function through the token bucket. For example: token bucket B for two different priority levels 1 And B 2 Token bucket B 1 Higher priority than token bucket B 2 Setting the barrel capacity a of two barrels 1 、a 2 And injection rate r 1 And r 2 Wherein a is 1 ＞a 2 ，r 1 ＞r 2 Assuming that the time t passes after the two buckets are filled with tokens, under the condition of no consideration of overflow (the tokens are generated and used at once), the highest token numbers which can be allocated by the two buckets are respectively as follows:
c token_1 ＝a 1 +r 1 t
c token_2 ＝a 2 +r 2 t
from the above calculation results, c token_1 ＞c token_2 Instant sign barrel B 1 The number of tokens available for allocation is greater than token bucket B 2 The quality of service of high priority applications is guaranteed.
Regarding the counting problem of the overflow bucket of the streaming media data, the invention takes the overflow bucket count generated in the set latest time period as an accumulated result (not all-time statistics). In order to facilitate statistics, the present invention sets a corresponding observation window for each level of streaming media data, as shown in fig. 4, and in this embodiment, one observation window is formed by 10 time cells. Each time cell identifies a fixed time period t x ，t x ＝a x /r x I.e. from the ratio of bucket capacity to token injection rate, at t x At most, an overflow barrel is generated in the duration, when the count result of the overflow barrel is decimal, no rounding operation is performed, the whole observation window accumulation operation is directly participated, and when t is reached x No token overflow for the duration is counted zero.
The observation window is at intervals of time t x Move forward by one time unit, the count value is at every time period t x And (5) re-counting.
Counting overflow barrel counts generated by each time cell in the observation window in real time, and summing all count values:
wherein c token_i Representing the overflow amount of the token in the period of the ith time cell, C bucket Representing the sum of overflow bucket counts generated by 10 time cells within the observation window.
The statistics may be forwarded to the neighboring nodes along with the local routing table data. Routing tables forwarded based on RIP protocol can be adjusted to provide the table entry structure shown in the following table:
The counting weight field is added in the routing table item, and the counting result of the overflow bucket is filled in the counting weight field, so that the counting result can be sent together with the updated routing table item, the resource waste caused by repeated filling of the IP packet header due to independent forwarding is avoided, the occupied space of redundant data is reduced, and meanwhile, a timely and accurate data basis is provided for the decision of subsequent routing.
When a router node receives routing tables sent by all neighbor nodes, firstly selecting a next-hop node with a shortest path in a destination address according to a shortest path algorithm of RIP protocol, and if the node is unique, updating the routing table entry sent by the node to the local; if there are a plurality of next hop nodes on the shortest path, the local count weights recorded under different priority classes need to be extracted from the routing table items of the nodes one by one, the maximum value of the local count weight field under each class is obtained by comparison, the router node corresponding to the maximum value is used as the next hop node of the streaming media data of the current priority class, and the routing information is updated to the local. The invention takes the shortest path as the priority principle, and selects the next-hop router node with the least service to the data as the forwarding node of the data aiming at the data of the class of service quality requirement when a plurality of next-hop router nodes meeting the shortest path condition exist. Therefore, the invention provides an independent route path for the class data of each class of service quality requirement, so that the forwarding paths of two different classes of data to the same destination can be greatly different. This mainly takes into account: compared with other neighbor nodes, the strong connection node established for certain type of data has the lowest utilization rate of the resource of the type in the current period of time and the longest idle time, so the strong connection node is used as the node on the data distribution optimal path.
RIP (Routing Information Protocol) is the most known routing protocol in the domain, and because it adopts a distributed routing path algorithm based on distance vector, compared with OSPF (Open Shortest Path First) adopting a link state algorithm, it has the characteristics of simple operation and convenient maintenance, and becomes a widely used routing protocol for internal networks of various enterprises at present. Therefore, the present invention uses the RIP protocol as a routing policy of the intra-domain network, and adjusts the routing condition of the RIP protocol according to the priority, as shown in fig. 5, the specific operation procedure of completing the update of the local routing table of the corresponding priority by using each level of strong connection node is as follows:
receiving a routing table sent by a neighbor node X, wherein the distance value from the neighbor node X to a destination address D is D 1 If the destination address D does not exist in the local routing table, adding the routing table entry into the local routing table;
otherwise, judging whether the neighbor node X is the same as the next hop node Y to the destination address D recorded in the local routing table, and if so, updating the routing table entry of the neighbor node X to the local;
if it is different, the distance value d 1 With destination address D recorded in the local routing table Distance value d 2 Comparing, if d 1 +1＜d 2 Updating the routing table entry of the neighbor node X to the local;
otherwise, if d 1 +1＝d 2 Extracting a next-hop node Y to a destination address D recorded in a local routing table, comparing the priority of a neighbor node X with the priority of the next-hop node Y, updating the routing table entry of the neighbor node X to the local if the priority of the neighbor node X is greater than the priority of the next-hop node Y, and not updating the local routing table entry if the priority of the neighbor node X is not greater than the priority of the next-hop node Y;
otherwise, if d 1 +1＞d 2 The local routing table entry is not updated;
the node priority of the common router is less than the node priority of the strong connection.
In the local network topology shown in fig. 6, three shortest paths of the router nodes R2 to R10 are respectively: r2- & gt R3- & gt R6- & gt R10, R2- & gt R5- & gt R8- & gt R10, R2- & gt R7- & gt R9- & gt R10; the overflow bucket of a certain class is set to be 2000 tokens, wherein each data packet to be forwarded corresponds to one token, the packet is fragmented into fixed-length packets to be forwarded outwards, the token injection rate is set to be 1000 pieces/min, and for convenience of comparison, the load quantity of the data of the certain class between any two adjacent nodes is 200 packets within each minute. In the initial state, the overflow bucket count in the observation windows of all router nodes is zero, and the observation window length is set to 10 time cells (10 minutes).
At the end of the 5 th minute, node R2 is ready to send data to node R10. At this time, for the node R3, since it transmits data simultaneously with five neighboring nodes, the bandwidth resource consumed per minute is 5×200=1000 packets, which is equivalent to the number of tokens generated per minute, so that no overflow value is generated and the count is zero. Similarly, node R5 may generate 5 x (1000-4 x 200) =1000 token overflows, which translates to 1/2 of the overflow bucket count, node R7 may generate 5 x (1000-3 x 200) =2000 token overflows, which translates to 1 of the overflow bucket count, where node R2 sets node R7 as a strong connection node on the connection path with R10, distributes the data to be transmitted to node R7, and since node R7 is only R9 from the shortest path of R10, forwards the data to R9 and delivers it directly to destination node R10 through node R9, and the final selected optimal routing path is the bold line shown in fig. 6.
As shown in fig. 7, the node R4 is ready to send data to the node R9, and five shortest paths between the two paths are respectively: r4- & gt R3- & gt R2- & gt R7- & gt R9, R4- & gt R3- & gt R5- & gt R8- & gt R9, R4- & gt R3- & gt R6- & gt R8- & gt R9 and R4- & gt R3- & gt R6- & gt R10- & gt R9; wherein the node R3 is a necessary router node, and the following routing paths of the node R3 are selected according to the setting conditions described in the embodiment of fig. 6, which are unchanged: the operation process according to the previous embodiment can be known that the router node with the least number of neighbor node connections is the strong connection node satisfying the condition as long as the router node with the least number of neighbor node connections is selected on the shortest route. On a plurality of shortest paths from the node R3 to the node R9, the adjacent node of the next-hop node R6 is least in connection, so that the node is defined as a strong connection node on the paths from the node R3 to the node R9, and data is forwarded to the node R6; for a strong connection node selection R10 on the path of nodes R6 to R9, the data is finally delivered to the destination node R9 by node R10.
The implementation process just indicates the route selection process by setting a simplest network topology structure and network working state, and the routing tables of the received neighbor nodes do not arrive at the same time under the actual network structure and working state, so when the data needing to be forwarded appears, the routing algorithm does not confirm all levels of strong connection nodes on the forwarding path in real time, but the data distribution is completed by the determined strong connection nodes in the locally updated routing information of each router node before the data needing to be forwarded appears, namely the routing table updating process and the data forwarding process are mutually independent, and the working efficiency is maximized.
When the network topology structure is unstable, router nodes are frequently deleted or added in the network, the newly laid router nodes can generate a cold start problem in a short time, the accumulated value of overflow barrels in an initial state is zero, and complete statistics can be obtained after the time of an observation window is needed. For the router node R to remain stable in operation in the network, it is assumed that one class of overflow buckets is set to capacity 1000 packets (or 1000 tokens, for ease of understanding and hereinafter collectively referred to as packets), packets are fragmented for fixed length forward-out, and the token injection rate is also set to 1000 per minute. As shown in fig. 8, the router node R fills the token bucket before time t1, and instantly sends 500 packets at time t2, t3 and t4, then 1000 packets are necessarily accumulated between time t1 and time t2 to be injected into the overflow bucket (1 bucket count), and 500 packets in the token bucket are taken away at time t 2; filling the empty space of the token bucket in the time t2-t3, injecting 500 packets into the overflow bucket (1/2 bucket count), and taking out 500 packets in the token bucket at the time t 3; similarly, there are still 500 packets injected into the overflow bucket between times t3-t 4. In this way, when the bucket count is accumulated by 1+9/2=5.5 for time t11, if a new node is added to the shortest path with the same destination address as the router node R, the router node R must be selected as a data forwarding node in the routing process because the new node overflows the bucket count value to be zero in the initial state. However, this allocation is not reasonable because there may not be traffic at the new node at this point, and router node R has 500 packets per unit time.
For this reason, the present invention considers the output flow per unit time as a compromise of the overflow bucket count. Taking the router node R as an example, 1000 packet injection overflow buckets (1 bucket count) are generated since no traffic output occurs between time t1-t2, 500 packet injection overflow buckets (1/2 bucket count) are generated between time t2-t3, and since there is 500 packet traffic output at time t2, the overflow bucket count between time t1-t3 is 1+ (1/2) - (1/2) =1, and so on, and the bucket overflow count is still 1 at time t 11. When a new node is added to the shortest path with the same destination address as the router node R at time t11, the new node only needs to wait for a unit time to time t12 to replace the router node R as a newly established strong connection node, because the overflow bucket count of the router node R becomes zero at time t12, and the overflow bucket count of the new node becomes 1 if there is no data traffic between times t11-t 12. If the above-mentioned break operation is not set, the new node needs to wait for at least 6 units of time before being replaced by the strongly connected node (in the case that the new node has no data output all the time).
Consider another case: the router node R has no flow output all the time in the whole observation window range between the time t1 and the time t11, and the overflow bucket count is 10; whereas traffic of bursty nature is generated at the end of time t11 and the output is continuous, no tokens overflow after time t 11. If the priority is required to be adjusted to the lowest state without considering the traffic loss calculation, 10 unit time is required to be spent on clearing the count of the observation window, and a large amount of shunt data is jammed on the R port of the router node within the time range, and even if the loss calculation mode is adopted, 5 unit time is required to be waited. In order to solve the problem, the invention designs another new damage operation mode:
if f x_i ＞a x Then calculate the valueAs the flow loss of the ith time cell, the loss difference of the observation window is calculated by the following formula:
wherein k represents f x_i ＞a x The number of consecutive occurrences of the condition before the ith time cell, k.gtoreq.1, if D bucket If the value is negative, returning a zero value as the counting result of the overflow bucket of the x-level streaming media data, otherwise, directly returning a difference value D bucket As a result of the count of the overflow buckets of the x-level streaming media data.
Taking the router node R as an example, the process of calculating the damage difference value by adopting the operation mode is as follows:
the 1 st peak data continuously occurs between time t11 and time t12, and the overflow bucket count at the end of time t12 is 9- (1200/1000) assuming that the peak data transmitted per unit time remains 1200 packets at all times, the peak data being the output data greater than the bucket capacity 1*1 7.8 (where the bucket capacity is 1000 packets, the token injection rate is 1000 packets/min, and 800 packets remain in the last token bucket at time t 12); the 2 nd peak data occurs continuously between time t12 and time t13, then the overflow bucket count at the end of time t13 is 8- [ (1200/1000) 1*1 +(1200/1000) 2*2 ]About 4.73, 600 packets remain in the last token bucket at time t 13; the 3 rd peak data occurs continuously between time t13 and time t14, then the overflow bucket count at the end of time t14 is 7- [ (1200/1000) 1*1 +(1200/1000) 2*2 +(1200/1000) 3*3 ]About 1.43, and 400 packets remain in the last token bucket at time t 14; the overflow bucket count is zero for negative values by default. Therefore, the improved flow breaking algorithm can be used for realizing zero clearing of the count of the observation window in only 3 unit time, the priority of the router node R can be quickly adjusted to the lowest, redundant tokens still remain in the token bucket, and the problem of route data congestion caused by slow degradation of the router node R is avoided.
In addition, in order to further improve the utilization efficiency of bandwidth resources, the invention also designs a dynamic adjustment method for the priority of the streaming media data, which specifically comprises the following steps:
when the overflow bucket count in the observation window of the mth-level streaming media data is lower than a set threshold alpha m And the overflow bucket count generated by the last time cell is zero, further judging that the overflow bucket count in the m-1 level streaming media data observation window is higher than a set threshold alpha m-1 When the overflow bucket count generated by the last time cell is not zero, the mth level streaming media data buffer queue is transited to the mth-1 level streaming media data buffer queue, whether the condition meeting the transition is still met or not is repeatedly monitored every interval time period t, if yes, the transition state of the mth level streaming media data buffer queue is kept, otherwise, the mth level streaming media data is transitedAnd recovering the buffer queue, wherein m represents the number of stages of the streaming media data, m is more than or equal to 2, and the higher the number of stages is, the lower the priority of the streaming media data is.
Threshold alpha m And a threshold alpha m-1 The value of (2) can be set to different values according to different levels, and decimal or rounding can be adopted according to the requirement. Assuming that three levels of streaming media data are set in the network, each level of observation window is set to have 10 unit time, and the total duration of each level of observation window is necessarily different because the token bucket capacity and the token injection rate of different levels are not the same. The priority of the 1 st-level streaming media data format is highest, the priority of the 3 rd-level streaming media data format is lowest, and the streaming media data format can be transited to the higher-level streaming media data only with the 2 nd and the 3 rd levels. Here, the level 1-3 observation window count thresholds may be set to 7, 4, and 1, respectively, and then for level 3 streaming media data, the condition that it can meet the priority transition is: and no token overflows in the last time cell, namely all tokens injected into the token bucket in the period are taken away, the count value of the overflow bucket in the level 3 observation window is lower than 1, the count threshold value of the level 2 observation window is higher than 4, and the count value of the overflow bucket generated by the last time cell is not zero, so that the level 3 streaming media data is transited to the level 2. Similarly, the condition for meeting the priority transition for the 2 nd-level streaming media data is as follows: and the overflow bucket count generated by the last time cell is zero, the overflow bucket count value in the 2 nd level observation window is lower than 4, the 1 st level observation window count threshold value is higher than 7, and the overflow bucket count generated by the last time cell is not zero, so that the 2 nd level streaming media data is transited to the 1 st level. When the above transition condition is not satisfied, the streaming media data needs to be degraded, and the minimum is reduced to the original level. The reason for setting the level 1 observation window count threshold at a high level is to keep the level 1 observation window count threshold at a high level, so as to keep the level 1 observation window count threshold always at a high service quality, reserve buffer resources for delivery of bursty traffic, and reduce jitter.
In order to realize the method, the invention also provides a route forwarding control system of the streaming media data, which comprises a plurality of router nodes distributed in a network; the router node supports the RIP routing protocol, as shown in fig. 9, and specifically includes: the system comprises a classifier, a cache, a token controller, an overflow bucket meter, a scheduler, a routing module, a routing data forwarding module, a clock controller and a priority adjustment module;
a classifier: classifying the received streaming media data by adopting IP priority or DSCP classification rules supporting QoS;
a cache memory: temporarily storing stream media data to be forwarded, and distributing independent buffer spaces for the stream media data with different priority classes to form buffer queues corresponding to the classes;
a token controller: setting a token bucket and an overflow bucket for each cache queue, setting an upper limit value of the token bucket which can be filled with tokens and a token injection rate according to the priority of the cache queue, synchronously generating tokens according to the set token injection rate, injecting the tokens overflowed from the token bucket into the overflow bucket, and regenerating an empty overflow bucket after the current overflow bucket is full;
Overflow bucket meter: counting the counts of all levels of stream media data overflow barrels in a router node in a preset period, and broadcasting all levels of count values outwards along with local routing table data to be forwarded;
a scheduler: sequencing the streaming media data of the obtained tokens according to the sequence of the acquisition time, and directing the sequencing queue to the corresponding output port;
and a routing module: in the process of updating a local routing table, a routing path algorithm utilizing a distance vector selects a neighbor node with the highest overflow bucket count value at each level as a strong connection node with corresponding priority for routing table items which are sent by different neighbor nodes and have the same destination address and shortest distance, and updates the routing table items sent by the strong connection node to the local;
a routing data forwarding module: forwarding the streaming media data of the corresponding level according to the path designated by the updated routing table item;
and the clock controller is used for executing clock synchronization on the token controller to which each priority class belongs.
And the priority adjustment module is used for executing transition operation from the low-level streaming media data buffer queue to the high-level streaming media data buffer queue according to the set transition condition, so as to finish dynamic adjustment of the priority of the streaming media data.
Although the subject matter has been described in language specific to structural features and/or methodological acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described above. Rather, the specific features and acts described above are example forms of implementing the claims.
Claims (9)
1. The method for controlling the route forwarding of the streaming media data is characterized by comprising the following steps:
each router node in the network classifies the received streaming media data according to a set priority rule and puts the streaming media data into various corresponding cache queues;
setting a token bucket and an overflow bucket for each cache queue, setting an upper limit value for loading the token bucket into the token and a token injection rate according to the priority of the cache queue, injecting the token overflowed from the token bucket into the overflow bucket, and generating an empty overflow bucket after the previous overflow bucket is full, and continuously injecting the token;
counting the counts of all levels of stream media data overflow barrels in a router node in a preset period, and broadcasting all levels of count values outwards along with local routing table data to be forwarded;
in the process of updating a local routing table, a routing path algorithm utilizing a distance vector selects a neighbor node with the highest overflow bucket count value at each level as a strong connection node with corresponding priority for routing table items which are sent by different neighbor nodes and have the same destination address and shortest distance, and updates the routing table items sent by the strong connection node to the local;
And forwarding the streaming media data of the corresponding level according to the path designated by the updated routing table item.
2. The method according to claim 1, wherein the priority rule uses QoS-supporting IP priority or DSCP classification policy to classify the priority class of the streaming media data with the number of differentiated services field labels in the IP packet header.
3. The method for controlling route forwarding of streaming media data according to claim 1, wherein the counting rule of each stage of overflow buckets of streaming media data is specifically as follows:
setting an observation window consisting of n time cells, each time cell being a time period t, moving the observation window forward by one time cell every time period t,
t x ＝a x /r x
wherein a represents the upper limit value of the token bucket which can be filled with the token, r represents the injection rate of the token bucket, x represents the number of stages of the priority corresponding to the streaming media data, and the upper limit value of the overflow bucket which can be filled with the token is the same as the token bucket of the corresponding priority class;
counting overflow barrel counts generated by each time cell in the observation window in real time, and summing all count values:
wherein c token_i Representing the overflow amount of the token in the period of the ith time cell, C bucket Representing the sum of overflow bucket counts generated by all time cells in the observation window, and adding the count value C to bucket As a result of the count of the overflow buckets of the x-level streaming media data.
4. The method for controlling forwarding of streaming media data according to claim 3, further comprising a traffic impairment process in the counting rule of each stage of overflow buckets of streaming media data:
counting the flow f of the x-level streaming media data forwarded by the local router node in the period of the ith time cell x_i ；
Flow f x_i And the value a x A comparison is made with respect to the number of the cells,if f x_i ≤a x Then calculate the valueAs the flow loss of the ith time cell, the loss difference of the observation window is calculated by the following formula:
if D bucket If the value is negative, returning a zero value as the counting result of the overflow bucket of the x-level streaming media data, otherwise, directly returning a difference value D bucket As the counting result of the x-level stream media data overflow barrel;
if f x_i ＞a x Then calculate the valueAs the flow loss of the ith time cell, the loss difference of the observation window is calculated by the following formula:
wherein k represents f x_i ＞a x The number of consecutive occurrences of the condition before the ith time cell, k.gtoreq.1, if D bucket If the value is negative, returning a zero value as the counting result of the overflow bucket of the x-level streaming media data, otherwise, directly returning a difference value D bucket As a result of the count of the overflow buckets of the x-level streaming media data.
5. The method for controlling routing forwarding of streaming media data according to claim 1, wherein the specific operation procedure of completing updating of the local routing table of the corresponding priority by using each level of strong connection node is as follows:
receiving a routing table sent by a neighbor node X, wherein the distance value from the neighbor node X to a destination address D is D 1 If the destination address D does not exist in the local routing table, adding the routing table entry into the local routing table;
otherwise, judging whether the neighbor node X is the same as the next hop node Y to the destination address D recorded in the local routing table, and if so, updating the routing table entry of the neighbor node X to the local;
if it is different, the distance value d 1 Distance value D to destination address D recorded in local routing table 2 Comparing, if d 1 +1＜d 2 Updating the routing table entry of the neighbor node X to the local;
otherwise, if d 1 +1＝d 2 Extracting a next-hop node Y to a destination address D recorded in a local routing table, comparing the priority of a neighbor node X with the priority of the next-hop node Y, updating the routing table entry of the neighbor node X to the local if the priority of the neighbor node X is greater than the priority of the next-hop node Y, and not updating the local routing table entry if the priority of the neighbor node X is not greater than the priority of the next-hop node Y;
Otherwise, if d 1 +1＞d 2 The local routing table entry is not updated;
wherein the common router node priority < strong connection node priority.
6. The method for controlling forwarding of streaming media data according to claim 3 or 4, further comprising a process of dynamically adjusting priority of streaming media data:
when the overflow bucket count in the observation window of the mth-level streaming media data is lower than a set threshold alpha m And the overflow bucket count generated by the last time cell is zero, further judging that the overflow bucket count in the m-1 level streaming media data observation window is higher than a set threshold alpha m-1 When the overflow bucket count generated by the last time cell is not zero, the mth level streaming media data buffer queue is transited to the mth-1 level streaming media data buffer queue, whether the condition meeting the transition is still met or not is repeatedly monitored at each interval time t, and if yes, the mth level streaming media data buffer is keptAnd storing a queue transition state, otherwise, restoring an m-th level streaming media data buffer queue, wherein m represents the level number identification of streaming media data, m is more than or equal to 2, and the higher the level number is, the lower the priority of the streaming media data is.
7. The route forwarding control system of the streaming media data is characterized by comprising a plurality of router nodes distributed in a network; the router node supports RIP routing protocol, comprising: the system comprises a classifier, a cache, a token controller, an overflow bucket meter, a scheduler, a routing module and a routing data forwarding module;
A classifier: classifying the received streaming media data by adopting IP priority or DSCP classification rules supporting QoS;
a cache memory: temporarily storing stream media data to be forwarded, and distributing independent buffer spaces for the stream media data with different priority classes to form buffer queues corresponding to the classes;
a token controller: setting a token bucket and an overflow bucket for each cache queue, setting an upper limit value of the token bucket which can be filled with tokens and a token injection rate according to the priority of the cache queue, synchronously generating tokens according to the set token injection rate, injecting the tokens overflowed from the token bucket into the overflow bucket, and regenerating an empty overflow bucket after the current overflow bucket is full;
overflow bucket meter: counting the counts of all levels of stream media data overflow barrels in a router node in a preset period, and broadcasting all levels of count values outwards along with local routing table data to be forwarded;
a scheduler: sequencing the streaming media data of the obtained tokens according to the sequence of the acquisition time, and directing the sequencing queue to the corresponding output port;
and a routing module: in the process of updating a local routing table, a routing path algorithm utilizing a distance vector selects a neighbor node with the highest overflow bucket count value at each level as a strong connection node with corresponding priority for routing table items which are sent by different neighbor nodes and have the same destination address and shortest distance, and updates the routing table items sent by the strong connection node to the local;
A routing data forwarding module: and forwarding the streaming media data of the corresponding level according to the path designated by the updated routing table item.
8. The system of claim 7, further comprising a clock controller for performing clock synchronization for the token controller to which each priority class belongs.
9. The system according to claim 7, further comprising a priority adjustment module configured to perform a transition operation from the low-level streaming media data buffer queue to the high-level streaming media data buffer queue according to a set transition condition, thereby completing a dynamic adjustment of the priority of the streaming media data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210154844.6A CN114401226B (en) | 2022-02-21 | 2022-02-21 | Method and system for controlling route flow of stream media data |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210154844.6A CN114401226B (en) | 2022-02-21 | 2022-02-21 | Method and system for controlling route flow of stream media data |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114401226A CN114401226A (en) | 2022-04-26 |
| CN114401226B true CN114401226B (en) | 2024-02-27 |
Family
ID=81234339
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210154844.6A Active CN114401226B (en) | 2022-02-21 | 2022-02-21 | Method and system for controlling route flow of stream media data |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114401226B (en) |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112600761A (en) * | 2020-12-11 | 2021-04-02 | 腾讯科技（深圳）有限公司 | Resource allocation method, device and storage medium |
| CN112804150A (en) * | 2019-11-13 | 2021-05-14 | 烽火通信科技股份有限公司 | Method and system for realizing distributed hierarchical routing forwarding table |
| GB202110569D0 (en) * | 2021-07-22 | 2021-09-08 | Nec Corp | Communication system |
| CN113767601A (en) * | 2019-05-23 | 2021-12-07 | 慧与发展有限责任合伙企业 | Method for distributing software-determined global load information |
Family Cites Families (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100800958B1 (en) * | 2001-10-04 | 2008-02-04 | 주식회사 케이티 | Method for controlling traffic flow using token bucket |
| US7680049B2 (en) * | 2005-02-08 | 2010-03-16 | Cisco Technology, Inc. | Methods and apparatus for allowing promotion in color-based policers |
| US8000235B2 (en) * | 2008-10-05 | 2011-08-16 | Contextream Ltd. | Bandwidth allocation method and apparatus |
| US9231828B2 (en) * | 2009-01-09 | 2016-01-05 | Foundry Networks, Llc | Hierarchical rate color marker |
| CN103763208B (en) * | 2014-01-29 | 2017-08-29 | 华为技术有限公司 | Data traffic method for limiting and device |
-
2022
- 2022-02-21 CN CN202210154844.6A patent/CN114401226B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113767601A (en) * | 2019-05-23 | 2021-12-07 | 慧与发展有限责任合伙企业 | Method for distributing software-determined global load information |
| CN112804150A (en) * | 2019-11-13 | 2021-05-14 | 烽火通信科技股份有限公司 | Method and system for realizing distributed hierarchical routing forwarding table |
| CN112600761A (en) * | 2020-12-11 | 2021-04-02 | 腾讯科技（深圳）有限公司 | Resource allocation method, device and storage medium |
| GB202110569D0 (en) * | 2021-07-22 | 2021-09-08 | Nec Corp | Communication system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114401226A (en) | 2022-04-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US10708192B2 (en) | Method to achieve bounded buffer sizes and quality of service guarantees in the internet network | |
| US20020161914A1 (en) | Method and arrangement for congestion control in packet networks | |
| Loi et al. | Service differentiation in optical burst switched networks | |
| JP2006506845A (en) | How to select a logical link for a packet in a router | |
| US6650618B1 (en) | Method and apparatus for managing communications between nodes in a bi-directional ring network | |
| US8031682B2 (en) | Apparatus and method for aggregating and switching traffic in subscriber network | |
| CN117240345A (en) | Large-scale constellation business management method for optimizing service coverage rate | |
| CN114401226B (en) | Method and system for controlling route flow of stream media data | |
| CN100502370C (en) | Optimizing system and method for medium transmission on different transmission channels | |
| Antoniou et al. | An efficient deadline-credit-based transport scheme for prerecorded semisoft continuous media applications | |
| Rashid et al. | Traffic intensity based efficient packet schedualing | |
| Jeong et al. | QoS support for UDP/TCP based networks | |
| Li et al. | Schedulability criterion and performance analysis of coordinated schedulers | |
| Bakiras et al. | Quality of service support in differentiated services packet networks | |
| Asaduzzaman et al. | Improving Quality of Service in Computer Networks Applying the Eight Classes of Service | |
| Hu et al. | A protocol independent policer and shaper design using virtual scheduling algorithm | |
| Tu et al. | QRP03-3: Bulk Scheduling for Delay Sensitive Streaming Applications | |
| Tu et al. | Enhanced bulk scheduling for supporting end-to-end delay requirements | |
| JP2002124985A (en) | Transmitting device and communication system and communication method | |
| Kuhns et al. | Lightweight flow setup for wirespeed resource reservation | |
| Tu et al. | Enhanced bulk scheduling for supporting delay sensitive streaming applications | |
| Reid et al. | Internet QoS: Past, Present, and Future | |
| Shimabara et al. | Expanded implementation of output buffer mechanism in the Internet router | |
| Antoniou et al. | Efficient end-to-end transport of soft real-time applications | |
| Jahromi et al. | IMPLEMENTATION OF ABR OVER ATM-VCS |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |