WO2013151468A1 - Procédé et système de gestion de file d'attente - Google Patents

Procédé et système de gestion de file d'attente Download PDF

Info

Publication number
WO2013151468A1
WO2013151468A1 PCT/SE2012/050362 SE2012050362W WO2013151468A1 WO 2013151468 A1 WO2013151468 A1 WO 2013151468A1 SE 2012050362 W SE2012050362 W SE 2012050362W WO 2013151468 A1 WO2013151468 A1 WO 2013151468A1
Authority
WO
WIPO (PCT)
Prior art keywords
congestion
queues
determining
queue
user
Prior art date
Application number
PCT/SE2012/050362
Other languages
English (en)
Inventor
Ingemar Johansson
Original Assignee
Telefonaktiebolaget L M Ericsson (Publ)
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 Telefonaktiebolaget L M Ericsson (Publ) filed Critical Telefonaktiebolaget L M Ericsson (Publ)
Priority to PCT/SE2012/050362 priority Critical patent/WO2013151468A1/fr
Publication of WO2013151468A1 publication Critical patent/WO2013151468A1/fr

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L43/00—Arrangements for monitoring or testing data switching networks
    • H04L43/08—Monitoring or testing based on specific metrics, e.g. QoS, energy consumption or environmental parameters
    • H04L43/0823—Errors, e.g. transmission errors
    • H04L43/0829—Packet loss
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/50—Queue scheduling
    • H04L47/62—Queue scheduling characterised by scheduling criteria
    • H04L47/6285—Provisions for avoiding starvation of low priority queues
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W28/00—Network traffic management; Network resource management
    • H04W28/02—Traffic management, e.g. flow control or congestion control
    • H04W28/10—Flow control between communication endpoints
    • 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
    • H04L47/31—Flow control; Congestion control by tagging of packets, e.g. using discard eligibility [DE] bits
    • 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
    • H04L47/32—Flow control; Congestion control by discarding or delaying data units, e.g. packets or frames
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/50—Queue scheduling
    • H04L47/62—Queue scheduling characterised by scheduling criteria
    • H04L47/6215—Individual queue per QOS, rate or priority
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L47/00—Traffic control in data switching networks
    • H04L47/50—Queue scheduling
    • H04L47/62—Queue scheduling characterised by scheduling criteria
    • H04L47/625—Queue scheduling characterised by scheduling criteria for service slots or service orders
    • H04L47/6275—Queue scheduling characterised by scheduling criteria for service slots or service orders based on priority

Definitions

  • the present disclosure relates to queue management within ECN (Explicit Congestion Notification) capable wireless communication systems, in particular to active queue management in LTE (Long Term Evolution) or HSPA (High Speed Packet Access) wireless communication systems.
  • ECN Exlicit Congestion Notification
  • LTE Long Term Evolution
  • HSPA High Speed Packet Access
  • So called active queue managers are implemented (in most cases in the PDCP layer) to ensure that users do not use too much bandwidth and to give endpoints a clear signal of congestion. In normal cases these entities measure the queue depth or the packet age at the outgoing link. If the packet depth or the packet age is too high, the queue manager may either drop the packet or mark it.
  • the marking approach is better known as explicit congestion notification (ECN), in which the term marking denotes notifying congestion explicitly, whereas congestion notification more generally denotes notifying congestion either implicitly by dropping a packet or explicitly by marking
  • the above mentioned ECN is a method with which routers can inform the receiver that congestion has been detected in a router in the path.
  • IP Internet Protocol
  • two bits of the header are specifically reserved for this purpose. These two bits are typically used in the following way. as is illustrated below in Table 1 :
  • ⁇ ' ECT(l) May be set by sender if it is ECN capable '1 1' ECN-CE Set by congested router if incoming packet had either ECT(O) or ECT(l)
  • ECN capable Senders that are not ECN capable will declare this by setting the ECN bits to non-ECT. If an ECN capable router detects not-ECT (ECN Capable Transport) it is not allowed to mark the packet with ECN-CE (ECN Congestion Experience). If congestion occurs in the router then it may drop some or many packets.
  • ECN-CE ECN Congestion Experience
  • An ECN capable sender declares that it has "ECN Capable Transport” (ECT) by setting the ECN bits to either ECT(O) or ECT(l).
  • ECT ECN Capable Transport
  • a congested ECN capable router can then indicate congestion by marking the ECN bits with ECN Congestion Experienced (ECN-CE) and then forwarding the packet instead of dropping it.
  • ECN-CE ECN Congestion Experienced
  • An ECN capable receiver that detects an ECN-CE marked packet either informs the sender about the detected congestion or initiates methods to reduce the congestion, typically by requesting a reduced bitrate.
  • the ECN bits are typically part of another field, such as the ToS (Type of Service) octet in IPv4.
  • ToS Type of Service
  • IPv6 the ECN bits are part of the Traffic Class field.
  • Modern 3GPP (3rd Generation Partnership Project) access implement a user context which can contain one or more queues with different scheduling priorities, each queue normally has its own queue management and normally also these queue managers are tuned differently depending on which kind of traffic the queue is destined for (VoIP, Video, Websurf, HTML streaming etc.).
  • the simplest approach is to use strict priority scheduling. This means that the queue with the highest priority is served first and the queues with lower priority are given what is left. This can be expected to work well as long as the highest priority queue does not use up all the available bandwidth.
  • rate throttling limits the rate even when there is no need for it; this has negative impact on e.g. HTTP based streaming or background download both in terms of system capacity and user experience.
  • a first aspect of the present disclosure presents a method of queue management for a user in a wireless communication system, including the steps of determining a congestion metric for the user, and determining a congestion impact parameter for at least one of a plurality of queues associated with the user. Finally, the method includes managing the plurality of queues based on the determined congestion metric and the congestion impact parameter.
  • a second aspect of the present disclosure presents an arrangement for queue management for a user in a wireless communication system, including a congestion metric unit configured for determining a congestion metric for the user, and a congestion impact unit configured for determining a congestion impact parameter for at least one of a plurality of queues associated with the user. Further, the arrangement includes a queue manager configured for managing the plurality of queues based on the determined congestion metric and the congestion impact parameter.
  • the presented disclosure provides means to ensure that lower priority queues are served during congested situations even though strict priority scheduling is used.
  • the disclosure makes rate-throttling functionality unnecessary, something that is beneficial both from a system capacity point of view and from a user experience point of view.
  • Fig. 1 illustrates a flow chart of an embodiment of a method according to the present disclosure
  • Fig. 2 illustrates a flow chart of a further embodiment of a method according to the present disclosure
  • Fig. 3 is a graph illustrating an example of a congestion metric suitable for use in the present disclosure
  • Fig. 4 illustrates a flow chart of an additional embodiment of a method according to the present disclosure
  • Fig. 5 illustrates an implementation in a down link scenario of the present disclosure
  • Fig. 6 illustrates part random byte selection according to an embodiment of the present disclosure
  • Fig. 7 illustrates an implementation in an uplink scenario of the present disclosure
  • Fig. 8 illustrates an embodiment of an arrangement according to the present disclosure
  • Fig. 9 illustrates a further embodiment of an arrangement according to the present disclosure.
  • the present disclosure will be described in the context of an LTE wireless communication system; however, it is equally applicable for a HSPA wireless communication system, with suitable adaptation.
  • the solution presented by the present disclosure enables a balanced use of the available bandwidth despite having a strict priority scheduling regime, by ensuring a high level of service for high priority class traffic but still be able satisfy the needs of lower priority traffic.
  • a problem is otherwise that, in for instance strict priority scheduling regimes, high priority classes may starve or block lower priority traffic.
  • a main point of the present disclosure is to provide an indication that congestion is present to all queues and thereby enabling or forcing the particular queue that is the main cause of the congestion to take action and to reduce its bitrate or drop packets.
  • congestion is measured for all queues for a particular user, a random byte is picked from one of the queues and a congestion metric is determined for the byte. If the congestion metric supersedes a particular threshold (indicative of congestion) the packet to which the random byte was selected is marked with a congestion probability. The originating queue for the random byte (and the thus marked packet) is located and an indication of congestion is sent to that queue, forcing the queue to take action as to reduce its impact on the overall congestion for the user.
  • the functionality is located in an eNodeB in a LTE system, or in a scheduler in the MAC protocol layer.
  • the queues are typically located in the higher RLC protocol layer.
  • RNC Radio Network Controller
  • the functionality is distributed between multiple nodes.
  • the present disclosure is described with downlink transmission in mind. It is loosely built on the byte drop principle [2] with the exception that it does not explicitly take packet size into account, rather it pays attention to how much data specific flows put in the queues or transmit buffers and either ECN-CE marks or drops the packets for the flows based on this information.
  • byte drop principle or byte-mode drop algorithm [2] the probability of dropping a particular packet is dependent on its byte-size.
  • the proposed method is to be seen as a complement to per-queue AQMs.
  • a user is utilizing a plurality of service, such as web- surfing, streaming live video, instant messaging, and possibly downloading media content. These services result in a plurality of queues associated with that particular user.
  • a congestion metric is determined S IO for that particular user.
  • a congestion impact parameter is also determined S20 for at least one of the queues associated with the user. Based on the determined congestion metric and the determined congestion impact parameter, the plurality of queues for the user are managed.
  • a congestion metric is determined for the user.
  • the previously described step of determining S20 a congestion impact parameter for at least one queue associated with the user includes the further steps of picking or selecting S21 a random byte among the bytes that are in queue or scheduled for transmission for the user.
  • a mark probability measure is determined S22 for the selected random byte to provide the congestion impact parameter. If the determined mark probability for the random byte exceeds a predetermined threshold, thus indicating that the byte is contributing to the congestion, the packet to which the byte belongs is marked S23, e.g. either dropped or the originating queue receives an explicit congestion notification. Subsequently, an originating queue for the random byte is located S30 and managed based on the mark probability measure.
  • the determined congestion metric can be any indication that a congested situation is present for the user.
  • a simple metric for the congestion that can be computed in e.g. a scheduler can be formulated as:
  • the marking probability can be implemented as a ramp function where the marking probability is zero when the congestion metric is below a given threshold. Once above this threshold the marking probability increases gradually with higher congestion levels until it saturates at 1.0, as illustrated by the graph in Figure 3, where the mark probability Pmark(C) can be expressed as :
  • C is the determined congestion metric
  • Cth is the threshold at when congestion marking is to occur
  • K is a constant. In a real system, both constants are subject to tuning.
  • the method to pick or select the random byte in step S21 reduces to the problem of picking a random number in the range 1 to N where N is given by:
  • N NQI +NQ2.. +NQK where NQI is the number of bytes in queue 1 , and NQK is the number of bytes in queue K.
  • the originating queue can be traced by the method above.
  • the present disclosure above is intended to fix the issue that a higher priority queue may starve (or block) lower priority queues.
  • An alternative method of the present disclosure is to base the mark probability for each queue on the instantaneous bitrate through the individual queue.
  • a congestion metric is determined S IO for a particular user associated with a plurality of queues.
  • the step of determining S20 a congestion impact parameter includes the further steps of determining S210 a respective bitrate for each of the plurality of queues, and determining S220 a total bitrate for the plurality of queues.
  • the congestion impact parameter is determined based on a comparison between the respective bitrate and the total bitrate for the user. If the determined congestion metric supersedes a predetermined threshold, e.g. congestion is present the step of managing S30 the plurality of queues comprises the further step of selecting and managing S310 a queue based on the comparison between the congestion parameter and the threshold, and a relation between each respective bitrate and said total bitrate.
  • This embodiment may be easier to implement than the previous embodiment.
  • the additional requirement is that it is necessary to measure or provide measurements of the bitrate for each queue, which requires an integration time in order to get a reliable estimate.
  • the probability to mark or drop a packet in any of the queues P m ark (C) can be determined in the same manner as previously presented.
  • the probability that a packet in a queue n is to be marked can be expressed as F(Br n ,Br lol ) where Br n is the bitrate through queue n and Brm is the total bitrate through all queues for the user.
  • the function F(Brn, Bn 0 t) can as an example be realized as
  • F(Br n ,Br lol ) BrjBr lol but also other variants are equally possible.
  • Figure 5 an outline on how the present disclosure can be implemented for downlink processing is illustrated. In the example three queues, a scheduler and the mark/drop control logic described in this disclosure are illustrated. In this case, whole packets are waiting in the different queues.
  • Figure 6 outlines an example how the random byte can be picked and how it can be mapped to the mark/ drop of a packet in the corresponding queue.
  • an uplink direction differs from the downlink direction is illustrated.
  • the number of bytes corresponding to each queue is in this case inferred from information from the terminal.
  • the terminal sends regular buffer status reports (BSR) that informs the base station about how many outstanding bytes the terminal has to transmit in the uplink direction.
  • BSR buffer status reports
  • LTE implements the concept of QCI, which in this context maps directly to the queue concept that is discussed earlier.
  • the BSRs gives reasonably detailed information about how many bytes belong to each QCI. Based on this information, the same algorithm as above can be used to determine which packet (belonging to which queue) should be marked or dropped.
  • marking is to be treated as synonymous to ECN-CE marking, i.e. the two ECN bits in the IP header are set to "11". ECN-CE marking is only allowed if the packet is ECN capable (ECT) With reference to Figure 8, embodiments of an arrangement for providing the previously described functionality according to the present disclosure will be described.
  • an arrangement 1 for queue management for users in a wireless communication system includes a congestion metric unit 10 configured for determining a congestion metric for a user, and a congestion impact unit 20 configured for determining a congestion impact parameter for at least one of a plurality of queues associated with the user. Further, the arrangement 1 includes a queue manager 30 configured for managing the plurality of queues based on the determined congestion metric and the congestion impact parameter.
  • the congestion impact unit 20 further includes a selector 21 configured for picking or selecting a random byte from one of the plurality of queues of the user, and a determining unit 22 configured for determining a mark probability measure for the random byte based on at least the determined congestion metric, to provide the congestion impact parameter, and a marker 23 configured for marking the packet to which the random byte belongs based on the mark probability.
  • the queue manager 30 is configured for locating and managing an originating queue for the random byte based on the determined mark probability measure.
  • the congestion impact unit 20 includes a bitrate determining unit 210 configured for determining a respective bitrate for each of the plurality of queues for the user, and a total bitrate determining unit 220 configured for determining a total bitrate for all the plurality of queues, and a comparator 230 configured for comparing the determined congestion metric to a predetermined threshold.
  • the arrangement includes a queue manager 30 that is configured for selecting and managing a queue based on the comparison of the congestion metric to the predetermined threshold e.g. congestion is present or not, and a relation between each respective bitrate and the total bitrate for each queue.
  • a scheduler or similar entity in a network node, such as an eNodeB or RNC, in a wireless communication system
  • the embodiments of the present disclosure ensure that lower priority queues are served during congested situations even though strict priority scheduling is used. Further, the present disclosure makes rate-throttling functionality unnecessary, something that is beneficial both from a system capacity point of view and from a user experience point of view.
  • the embodiments described above are to be understood as a few illustrative examples of the present disclosure. It will be understood by those skilled in the art that various modifications, combinations and changes may be made to the embodiments without departing from the scope of the present disclosure. In particular, different part solutions in the different embodiments can be combined in other configurations, where technically possible. The scope of the present disclosure is, however, defined by the appended claims.

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Environmental & Geological Engineering (AREA)
  • Data Exchanges In Wide-Area Networks (AREA)

Abstract

L'invention concerne un procédé de gestion de file d'attente pour un utilisateur dans un système de communication sans fil. Ledit procédé consiste à déterminer (S10) une mesure d'encombrement pour l'utilisateur et à déterminer (S20) un paramètre d'impact d'encombrement pour au moins l'une d'une pluralité de files d'attente associées à l'utilisateur, et enfin à gérer (S30) la pluralité de files d'attente sur la base de la mesure d'encombrement déterminée et du paramètre d'impact d'encombrement.
PCT/SE2012/050362 2012-04-03 2012-04-03 Procédé et système de gestion de file d'attente WO2013151468A1 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/SE2012/050362 WO2013151468A1 (fr) 2012-04-03 2012-04-03 Procédé et système de gestion de file d'attente

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/SE2012/050362 WO2013151468A1 (fr) 2012-04-03 2012-04-03 Procédé et système de gestion de file d'attente

Publications (1)

Publication Number Publication Date
WO2013151468A1 true WO2013151468A1 (fr) 2013-10-10

Family

ID=49300829

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/SE2012/050362 WO2013151468A1 (fr) 2012-04-03 2012-04-03 Procédé et système de gestion de file d'attente

Country Status (1)

Country Link
WO (1) WO2013151468A1 (fr)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2869514A1 (fr) * 2013-10-30 2015-05-06 Alcatel Lucent Procédé et système de gestion d'une file d'attente dans un réseau à commutation par paquets
WO2016039673A1 (fr) * 2014-09-10 2016-03-17 Telefonaktiebolaget L M Ericsson (Publ) Marquage de notification explicite de congestion de trafic utilisateurs
WO2018030945A1 (fr) * 2016-08-12 2018-02-15 Telefonaktiebolaget Lm Ericsson (Publ) Procédés de planification de transmissions à l'aide de limites de synchronisation et dispositifs de communication sans fil associés

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2009008817A1 (fr) * 2007-07-06 2009-01-15 Telefonaktiebolaget L M Ericsson (Publ) Contrôle d'encombrement dans un nœud de transmission

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2009008817A1 (fr) * 2007-07-06 2009-01-15 Telefonaktiebolaget L M Ericsson (Publ) Contrôle d'encombrement dans un nœud de transmission

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
FLOYD, S. ET AL.: "Random Early Detection Gateways for Congestion Avoidance", IEEE/ACM TRANSACTIONS ON NETWORKING., vol. 1, no. 4., August 1993 (1993-08-01), pages 400 - 401, Retrieved from the Internet <URL:http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=2518921> *

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP2869514A1 (fr) * 2013-10-30 2015-05-06 Alcatel Lucent Procédé et système de gestion d'une file d'attente dans un réseau à commutation par paquets
WO2015063018A1 (fr) * 2013-10-30 2015-05-07 Alcatel Lucent Procédé et système de gestion de files d'attente dans un réseau à commutation de paquets
US10116579B2 (en) 2013-10-30 2018-10-30 Alcatel Lucent Method and system for queue management in a packet-switched network
WO2016039673A1 (fr) * 2014-09-10 2016-03-17 Telefonaktiebolaget L M Ericsson (Publ) Marquage de notification explicite de congestion de trafic utilisateurs
WO2018030945A1 (fr) * 2016-08-12 2018-02-15 Telefonaktiebolaget Lm Ericsson (Publ) Procédés de planification de transmissions à l'aide de limites de synchronisation et dispositifs de communication sans fil associés
US10893533B2 (en) 2016-08-12 2021-01-12 Telefonaktiebolaget Lm Ericsson (Publ) Methods of scheduling transmissions using timing limits and related wireless communication devices

Similar Documents

Publication Publication Date Title
JP7238933B2 (ja) 受信した体感品質情報に基づくリソース割り当ての最適化
RU2515997C2 (ru) Активное управление очередью для восходящей линии связи в сети беспроводной связи
EP1632059B1 (fr) Transmission de paquets de supervision destines a commander l&#39;encombrement et l&#39;etablissement d&#39;appel dans des reseaux de paquets a largeur de bande limitee
JP4847541B2 (ja) データパケットトラフィック輻輳を解決する方法及び装置
US9386597B2 (en) QoE-aware traffic delivery in cellular networks
US8817615B2 (en) Method for operating a wireless network and a wireless network
CN108738080B (zh) 数据传输的方法和装置
EP2789184B1 (fr) Contrôle de flux orienté application dans un réseau radio
US20080225725A1 (en) Method and apparatus for supporting uplink starvation avoidance in a long term evolution system
US20090067335A1 (en) Congestion control in a transmission node
US8130709B2 (en) Apparatus and method for transmitting and receiving control information for uplink data in a mobile communication system and system thereof
CN102239666A (zh) 电信网络中能够实现拥塞指示的方法和装置
AU8043701A (en) Domain based congestion management
IL200892A (en) METHOD AND SYSTEM FOR SUPPORTING PREVIOUS QUALITY RELATIONSHIP WITH A SATELLITE IN A DEVELOPING LONG TERM SYSTEM
CN105471763B (zh) 控制报文传输方法及装置
WO2013082789A1 (fr) Procédé et dispositif de contrôle de congestion de réseau
WO2008154786A1 (fr) Contrôle du débit binaire maximum de liaison montante
US20100046424A1 (en) Method for controlling load matching in a radio communication system
WO2013151468A1 (fr) Procédé et système de gestion de file d&#39;attente
Szilágyi et al. Efficient LTE PDCP buffer management
WO2014128239A1 (fr) Procédé, entité de gestion et entité d&#39;agent destinés à une attribution de bande passante homogène
Yeom et al. Providing consistent delay differentiation
Deiß et al. QoS
Korossy et al. QoS Aware HSDPA Congestion Control Algorithm

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 12873586

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 12873586

Country of ref document: EP

Kind code of ref document: A1