[go: up one dir, main page]
More Web Proxy on the site http://driver.im/

CN101969410A - Dynamic queue management method for differentiated service network - Google Patents

Dynamic queue management method for differentiated service network Download PDF

Info

Publication number
CN101969410A
CN101969410A CN2010105329924A CN201010532992A CN101969410A CN 101969410 A CN101969410 A CN 101969410A CN 2010105329924 A CN2010105329924 A CN 2010105329924A CN 201010532992 A CN201010532992 A CN 201010532992A CN 101969410 A CN101969410 A CN 101969410A
Authority
CN
China
Prior art keywords
message
pilot frequency
yellow
green
probability
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.)
Pending
Application number
CN2010105329924A
Other languages
Chinese (zh)
Inventor
周井泉
张顺颐
谈玲
宁向延
张振
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nanjing Post and Telecommunication University
Nanjing University of Posts and Telecommunications
Original Assignee
Nanjing Post and Telecommunication University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Nanjing Post and Telecommunication University filed Critical Nanjing Post and Telecommunication University
Priority to CN2010105329924A priority Critical patent/CN101969410A/en
Publication of CN101969410A publication Critical patent/CN101969410A/en
Pending legal-status Critical Current

Links

Images

Landscapes

  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

The invention relates to a dynamic queue management method for a differentiated service network, disclosing a discontinuous orthogonal frequency division multiplying channel estimation method based on compressed sensing. The method comprises the following steps: designing channel estimation pilot frequency patterns; selecting the pilot frequency patterns; and estimating a channel frequency domain response. In the method, the selection of the pilot frequency patterns utilizes the following two schemes, wherein the scheme 1 comprises the steps of maintaining the traditional uniform pilot frequency patterns and naturally forbidding the pilot frequency in the position of forbidden subcarriers so that the pilot frequency presents natural nonuniformity; the scheme 2 comprises the steps of fixing the pilot frequency quantity and searching a subcarrier transmission pilot frequency signal in available subcarriers by adopting the following optimization problem disclosed in the specification based on the standard for recovering the minimization of matrix cross-correlation, wherein L is the channel length, P is the pilot frequency quantity, and N is the subcarrier sum of an OFDM (Orthogonal Frequency Division Multiplexing) system. By utilizing the method, channel estimation properties and system error rate properties superior to those obtained by the traditional methods can be obtained with less pilot frequency under the scene of various forbidden subcarriers.

Description

Dynamic queue's management method of differentiated services network
Technical field
The present invention relates under a kind of Differentiated Services environment based on cybernatic queue management route management method, how main research is in the queue management technology that realizes on the cybernetics basis under the Differentiated Services environment in the differentiated services network model, and it is multi-field to relate to differentiated services network service system structure, route technology, transfer of data etc.
Background technology
The basic boom of Internet is based upon on the IP basis, and all groupings all are to be subjected to identical processing among the IP, the service model of Here it is traditional " (BE, Best Effort) does one's best ".Along with network by of the development of single transfer of data to the integrated services network that transmits multimedia messagess such as data, voice and video, the BE service structure can not satisfy various application again, and in the network of doing one's best,, also may produce service " differentiation " if adopt suitable controlling mechanism.
Therefore, IETF standardization support several new agreement and the technology of IP QoS, wherein based on the priority technology be to Business Stream classify, shaping, mark.According to the policy in resource management Resources allocation, the business high to qos requirement gives priority treatment.Differentiated Services just is based on this strategy.End-to-end congested Control Study focus is active queue management (AQM, Active Queue Management) in recent years.It can use grouping to abandon and the mark that divides into groups comes notification source end changing down.Mainly be in order to solve following problem:
1) the early detection route is contingent congested, and comes the notification source end to take measures to avoid contingent congested by random drop or mark packets;
2) avoid a plurality of TCP to connect owing to " starting slowly " state that enters that causes is synchronously overflowed in formation;
3) keep less formation, between high-throughput and low time delay, make reasonable balance;
4) fair dealing comprises the various TCP Business Streams of sudden, persistence and intermittence.
Summary of the invention
Technical problem:
The objective of the invention is to study and propose under a kind of Differentiated Services environment based on cybernatic queue management route technology.By Service Management to convergence flow, select corresponding formation to carry out information stores and transmission, and combine with the queue scheduling technology, can reduce information dropout probability and raising resource utilization in the path.
Technical scheme:
The present invention adopts following technical scheme for achieving the above object:
Dynamic queue's management method of differentiated services network of the present invention comprises two parts: total message dropping probability calculates by PIP in first; Second portion is calculated the drop probability of Green, Yellow, Red message by LDP.
Preferably, to calculate total message dropping probability method as follows for described PIP:
When a message arrives:
If be the Green message
Upgrade Green message arrival rate Rate G
If be the Yellow message
Upgrade Yellow message arrival rate Rate Y
Otherwise be the Red message
Upgrade Red message arrival rate Rate R
The PIP module is calculated total losing probability p, the arrival rate sliding window service time method of Green or Yellow or Red message is estimated.
Preferably, the method for the drop probability of described LDP calculating Green, Yellow, Red message is as follows:
When a message arrives:
If be the Green message
With probability p G Abandon this message;
If be the Yellow message
With probability p Y Abandon this message;
If be the Red message
With probability p R Abandon this message;
Otherwise message enters fifo queue;
Figure 2010105329924100002DEST_PATH_IMAGE001
Figure 779562DEST_PATH_IMAGE002
At first p G =0, p Y =0, calculate p R If, p R 1, then p R =1; Secondly p G =0, p R =1, calculate p Y If, p Y 1, then p Y =1; At last p R =1, p Y =1, calculate p G , the arrival rate sliding window service time method of Green or Yellow or Red message is estimated.
Beneficial effect:
By under the Differentiated Services environment based on the queue management route technology of PIPGYR algorithm, we can realize:
1) administration queue makes it to keep less queue length, reduces queuing delay, improves throughput, increases the ability that absorbs burst packet.
2) take into account justice, making can the equality occupied bandwidth between response flow in the link and non-response flow, the friendly stream of TCP.
3) the effectively congested control of abutting end system, along with the increase of the non-response flow in the network, the performance that Active Queue Management Algorithm must obtain when non-response flow exists.
Description of drawings
Fig. 1 is the queue scheduling and management block diagram on the routing node.
Fig. 2 is the DS byte structure.
Fig. 3 is a flow regulator.
Embodiment
Be elaborated below in conjunction with the technical scheme of accompanying drawing to invention:
The present invention proposes under a kind of Differentiated Services environment based on cybernatic queue management method for routing, this method comprises two
Individual part: total message dropping probability calculates by PIP in first, and second portion is by LDP(Loss Probably Divider) calculate the drop probability of Green, Yellow, Red message.In the differentiated services network model, core node is responsible for message is determined next-hop node according to type of service and is transmitted to this node, and neighbor node is carried out the collection and the assessment of degree of belief; The multipath routing algorithm is leading indicator with safety, and Route Selection is only selected in those confidence levels reach node in a certain standard, if a plurality of hop nodes are down arranged, then selects certain next-hop node therein at random according to probabilistic algorithm.
Fig. 1 is the queue scheduling and management block diagram on the routing node.
1 based on the credible tactful route matrix of business intelligence identification with control
Differentiated Services is treated the aggregated flow of different brackets with a certain discrimination based on the polymerization of IP traffic classification, carries according to the DSCP selection in packet header
Transmit service for the scheduling of extra fine quality, its external characteristic be whenever slip a line into.The architecture of Differentiated Services is made up of the plurality of function modules that realizes on network node, comprises PHD, bag grader, bag sign and re-mark, flow regulator, and wherein flow regulator comprises meter, marker, reshaper and packet loss device etc. again.
Flow regulator logically is divided into calculator, marker, reshaper and Dropper again.As shown in Figure 3.
Meter is according to some real-time property of the traffic specification meter among the TCA, as speed etc., and statistical information passed to marker, reshaper and packet loss device.The mark suitable DSCP of marker in the DS in packet header tag field assembles even grouping puts certain behavior under, and the DS byte as shown in Figure 2.Reshaper, packet loss device make become a mandarin (or going out stream) to meet the stream standard of TCA by pressure such as the means that postpone, abandon.
In the differentiated services network environment, PHB is the core of whole differentiated services network, describes the mode that individual node is specific stream aggregation Resources allocation in essence; The whole resource allocation policy of Differentiated Services system just by this one by one the individual node resource allocation realize.A plurality of PHB then just constitute a PHB group because in close relations and must define simultaneously each other when realizing.Some PHB groups can define simultaneously, then claim it to belong to same PHB and organize family.
The Differentiated Services system is a kind ofly to guarantee the system of strategy based on business classification and correlated quality, because its autgmentability and realize has simply preferably become and supported IP QoS preferred manner.
The Queue Management Algorithm PIPGYR route technology computation model of 2 differentiated services networks
When a message arrives:
If be the Green message
Upgrade Green message arrival rate Rate G
If be the Yellow message
Upgrade Yellow message arrival rate Rate Y
Otherwise be the Red message
Upgrade Red message arrival rate Rate R
The PIP module is calculated total losing probability p
The LDP module is calculated the losing probability of Green message, Yellow message and Red message p G , p Y , p R
(if being the Green message)
With probability in p G Abandon this message
(if being the Yellow message)
With probability in p Y Abandon this message
(if being the Red message)
With probability in p R Abandon this message
Otherwise message enters fifo queue.
Variable:
Rate G / Rate Y / Rate R : Green/Yellow/Red message arrival rate;
p: the losing probability of message (Green or Yellow or Red);
p G / p Y / p R :: the losing probability of Green/Yellow/Red message;
In the algorithm, the arrival rate sliding window service time method of Green or Yellow or Red message [10]Estimate.
LDP is asked the losing probability of Green, Yellow, Red message by following equation group:
Order
Figure 108912DEST_PATH_IMAGE003
Then
Figure 307812DEST_PATH_IMAGE001
Figure 789740DEST_PATH_IMAGE002
At first make during calculating p G =0, p Y =0, calculate p R If, p R 1, then p R =1, order again p G =0, p R =1, calculate p Y If, p Y 1, then p Y =1, then order p R =1, p Y =1, calculate p G
3 active queue management Routing Protocols
The system model of Differentiated Services is by to having the application class that different QoS requires, and adopt different priority to handle according to different class, promptly in network node to variety classes professional adopt different whenever slip a line into, realize professional required service quality with this.The different business convergence flow adopts different queue stores, abandons or transmit packet.
3.1 bag grader
The bag grader is divided into different aggregate flows according to the particular provisions among the TCA according to some territory Jiang Bao in IP packet header.Defined BA(Behavior at present) grader and MF(Mutil-Field) grader.The former will wrap classification according to the DSCP in IP packet header; The latter, classifies to the IP bag as IP source address, IP destination address, DSCP source port address, destination interface address, agreement ID etc. according to a plurality of territories in packet header.
3.2 bag Dropper
The Queue Management Algorithm PIPGYR route technology computation model of differentiated services network
When a message arrives:
If be the Green message
Upgrade Green message arrival rate Rate G
If be the Yellow message
Upgrade Yellow message arrival rate Rate Y
Otherwise be the Red message
Upgrade Red message arrival rate Rate R
The PIP module is calculated total losing probability p
The LDP module is calculated the losing probability of Green message, Yellow message and Red message p G , p Y , p R
(if being the Green message)
With probability in p G Abandon this message
(if being the Yellow message)
With probability in p Y Abandon this message
(if being the Red message)
With probability in p R Abandon this message
Otherwise message enters fifo queue.
Variable:
Rate G / Rate Y / Rate R : Green/Yellow/Red message arrival rate;
p: the losing probability of message (Green or Yellow or Red);
p G / p Y / p R :: the losing probability of Green/Yellow/Red message;
In the algorithm, the arrival rate sliding window service time method of Green or Yellow or Red message [10]Estimate.
LDP is asked the losing probability of Green, Yellow, Red message by following equation group:
Order
Figure 783104DEST_PATH_IMAGE003
Then
Figure 599750DEST_PATH_IMAGE001
Figure 336762DEST_PATH_IMAGE002
At first make during calculating p G =0, p Y =0, calculate p R If, p R 1, then p R =1, order again p G =0, p R =1, calculate p Y If, p Y 1, then p Y =1, then order p R =1, p Y =1, calculate p G

Claims (3)

1. dynamic queue's management method of a differentiated services network is characterized in that this method comprises two parts: total message dropping probability calculates by PIP in first; Second portion is calculated the drop probability of Green, Yellow, Red message by LDP.
2. dynamic queue's management method of differentiated services network according to claim 1, it is as follows to it is characterized in that described PIP calculates total message dropping probability method:
When a message arrives:
If be the Green message
Upgrade Green message arrival rate Rate G
If be the Yellow message
Upgrade Yellow message arrival rate Rate Y
Otherwise be the Red message
Upgrade Red message arrival rate Rate R
The PIP module is calculated total losing probability p, the arrival rate sliding window service time method of Green or Yellow or Red message is estimated.
3. dynamic queue's management method of differentiated services network according to claim 1, it is as follows to it is characterized in that described LDP calculates the method for drop probability of Green, Yellow, Red message:
When a message arrives:
If be the Green message
With probability p G Abandon this message;
If be the Yellow message
With probability p Y Abandon this message;
If be the Red message
With probability p R Abandon this message;
Otherwise message enters fifo queue;
Figure 45241DEST_PATH_IMAGE002
At first p G =0, p Y =0, calculate p R If, p R 1, then p R =1; Secondly p G =0, p R =1, calculate p Y If, p Y 1, then p Y =1; At last p R =1, p Y =1, calculate p G , the arrival rate sliding window service time method of Green or Yellow or Red message is estimated.
CN2010105329924A 2010-11-05 2010-11-05 Dynamic queue management method for differentiated service network Pending CN101969410A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2010105329924A CN101969410A (en) 2010-11-05 2010-11-05 Dynamic queue management method for differentiated service network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2010105329924A CN101969410A (en) 2010-11-05 2010-11-05 Dynamic queue management method for differentiated service network

Publications (1)

Publication Number Publication Date
CN101969410A true CN101969410A (en) 2011-02-09

Family

ID=43548515

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2010105329924A Pending CN101969410A (en) 2010-11-05 2010-11-05 Dynamic queue management method for differentiated service network

Country Status (1)

Country Link
CN (1) CN101969410A (en)

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050128947A1 (en) * 2003-12-15 2005-06-16 International Business Machines Corporation System and method for testing differentiated services in a value add network service
US20050135378A1 (en) * 2003-12-22 2005-06-23 Nortel Networks Limited Service aware policer with efficient handling of in-profile traffic
CN101420461A (en) * 2008-12-08 2009-04-29 北京邮电大学 Adaptive token barrel marking method based on service sensing

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20050128947A1 (en) * 2003-12-15 2005-06-16 International Business Machines Corporation System and method for testing differentiated services in a value add network service
US20050135378A1 (en) * 2003-12-22 2005-06-23 Nortel Networks Limited Service aware policer with efficient handling of in-profile traffic
CN101420461A (en) * 2008-12-08 2009-04-29 北京邮电大学 Adaptive token barrel marking method based on service sensing

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
《计算机工程与科学》 20050930 张明杰 等 PIPIO:一个新的面向区分服务确保转发的主动队列管理算法 4-6 第27卷, 第9期 2 *

Similar Documents

Publication Publication Date Title
CN111600754B (en) Industrial heterogeneous network scheduling method for interconnection of TSN (transmission time network) and non-TSN (non-Transmission time network)
CN104272680B (en) Signalling congestion
US7889661B2 (en) Constrained multipath routing method in a multi-protocol label switching (MPLS) network
Yeom et al. Modeling TCP behavior in a differentiated services network
JP4547341B2 (en) Packet relay device with communication quality control function
US7046665B1 (en) Provisional IP-aware virtual paths over networks
WO2017157274A1 (en) Network-traffic control method and network device thereof
CN109714275B (en) SDN controller for access service transmission and control method thereof
CN105897575A (en) Path computing method based on multi-constrained path computing strategy under SDN
US20140293791A1 (en) Ethernet differentiated services conditioning
CN102014075B (en) Traffic policing method and device
US7917648B2 (en) Self-adaptive scheduling method and network element
JP2001358772A (en) Random early demotion and promotion marker
CN104994033A (en) Method for guaranteeing QoS (quality of service) of SDN (software defined network) by means of dynamic resource management
CN102752192B (en) Bandwidth allocation method of forwarding and control element separation (ForCES) transmission mapping layer based on stream control transmission protocol (SCTP)
JP2006506845A (en) How to select a logical link for a packet in a router
CN105007235B (en) A kind of jamming control method in wireless multimedia sensor network
CN101465794B (en) Method and apparatus for scheduling packet
CN105262696B (en) A kind of method and relevant device of multipath shunting
CN101110753A (en) End-to-end quality assurance method and system based on congestion alarm
CN114629847B (en) Coupled multi-stream TCP congestion control method based on available bandwidth allocation
CN101969410A (en) Dynamic queue management method for differentiated service network
Nabeshima et al. Fair queuing with in/out bit in core stateless networks
JP5655484B2 (en) Packet transmission device, signal termination device, communication system, and communication method
CN100490421C (en) Method for realizing flow equitable transmission in MPLS looped network

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C12 Rejection of a patent application after its publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20110209