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
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04Q—SELECTING
  • H04Q11/00—Selecting arrangements for multiplex systems
  • H04Q11/0001—Selecting arrangements for multiplex systems using optical switching
  • H04Q11/0062—Network aspects
  • H04Q11/0067—Provisions for optical access or distribution networks, e.g. Gigabit Ethernet Passive Optical Network (GE-PON), ATM-based Passive Optical Network (A-PON), PON-Ring
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04B—TRANSMISSION
  • H04B10/00—Transmission systems employing electromagnetic waves other than radio-waves, e.g. infrared, visible or ultraviolet light, or employing corpuscular radiation, e.g. quantum communication
  • H04B10/25—Arrangements specific to fibre transmission
  • H04B10/2581—Multimode transmission
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
  • H04L12/00—Data switching networks
  • H04L12/28—Data switching networks characterised by path configuration, e.g. LAN [Local Area Networks] or WAN [Wide Area Networks]
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04Q—SELECTING
  • H04Q11/00—Selecting arrangements for multiplex systems
  • H04Q11/0001—Selecting arrangements for multiplex systems using optical switching
  • H04Q11/0062—Network aspects
  • H04Q2011/0064—Arbitration, scheduling or medium access control aspects
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04Q—SELECTING
  • H04Q11/00—Selecting arrangements for multiplex systems
  • H04Q11/0001—Selecting arrangements for multiplex systems using optical switching
  • H04Q11/0062—Network aspects
  • H04Q2011/0084—Quality of service aspects
• • H—ELECTRICITY
  • H04—ELECTRIC COMMUNICATION TECHNIQUE
  • H04Q—SELECTING
  • H04Q11/00—Selecting arrangements for multiplex systems
  • H04Q11/0001—Selecting arrangements for multiplex systems using optical switching
  • H04Q11/0062—Network aspects
  • H04Q2011/0086—Network resource allocation, dimensioning or optimisation

Definitions

• the present invention relates to congestion avoidance in a communication network and more particularly, to providing a centralized and dynamically-adjusting- congestion avoidance in a Passive Optical Network.
• Network congestion often occurs when there is an overcrowding of traffic in the network.
• traffic refers to a packet, a message, streams, or other suitable form(s) of data, voice or combinations thereof.
• Symptoms of network congestion can include, for example, network delay, degradation of quality of service (QOS), and an extreme underutilization of network capacity. For this reason, communications systems typically employ techniques to avoid network congestion when oversubscription is applied.
• oversubscription refers to when the total amount of bandwidth that is assigned to subscribers is more than actual capacity.
• An aspect of the present invention involves a method for congestion avoidance in a passive optical network having an optical line terminator communicatively connected to a plurality of optical network termination devices.
• the method comprising periodically requesting a dynamic bandwidth allocation information from an optical network termination device by the optical line terminator, receiving the dynamic bandwidth allocation information having a queue fill level from the optical network termination device in response to the request, determining an adjusted bandwidth allocation for the optical network termination device by the optical line terminator, determining by the optical line terminator a packet-drop command to be taken at the optical network termination device, and sending to the optical network termination device the adjusted bandwidth allocation and the packet-drop command.
• dynamic congestion avoidance is provided centrally at the optical line terminator for the plurality of devices.
• Another aspect of the present invention involves a method for congestion avoidance in an optical network termination device of a passive optical network.
• the method comprising sending upon request, a dynamic-bandwidth-allocation information to an optical line terminator, receiving an adjusted bandwidth allocation and a packet-drop command from the optical line terminator, and executing the packet-drop command.
• a dynamic-congestion avoidance is provided remotely for the optical line terminator.
• optical line terminator providing a dynamic congestion avoidance in a passive optical network.
• the optical line terminator comprising a communication connection to a plurality of optical network termination devices each having a priority queue, a plurality of network avoidance parameters, a dynamic bandwidth mechanism that allocates bandwidth for the optical network termination devices and that dynamically updates the network avoidance parameters, a feedback mechanism that analyzes upstream traffic from the optical network termination device and periodically provides information based on the analysis to the dynamic bandwidth mechanism to aid in the update of the parameters.
• a dynamic-congestion avoidance is provided centrally at the optical line terminator for the plurality of devices.
• Figure 1 illustrates an exemplary prior art schematic diagram of a communication system having a standalone congestion handling in a G-PON system
• Figure 2 illustrates another embodiment of an exemplary prior art schematic diagram of a communication system having a standalone congestion handling in a G- PON system
• Figure 3 illustrates an exemplary schematic diagram of a communication system having a centralized congestion avoidance mechanism in a G-PON system in accordance to the present invention
• Figure 4 illustrates an exemplary message flow diagram of a communication system having a centralized-congestion avoidance in a G-PON in accordance with the present invention
• Figure 5 illustrates another exemplary message flow diagram of a communication system having a centralized-congestion avoidance in a G-PON in accordance with the present invention.
• Figure 6 illustrates an exemplary flow diagram for a method of a communication system in accordance with the present invention
• the invention described herein may employ one or more of the following concepts.
• one concept relates to a centralized-congestion avoidance located in an Optical Line Terminator (OLT).
• OLT Optical Line Terminator
• PON Passive Optical Network
• Another concept relates to collocating the congestion avoidance with a Dynamic Bandwidth Allocation (DBA) mechanism in the OLT.
• DBA Dynamic Bandwidth Allocation
• Another concept relates to a remotely located congestion avoidance as viewed by an Optical Network Terminator (ONT).
• Yet another concept relates to discarding a packet from a priority queue in the ONT for congestion avoidance.
• Still another concept relates to a feedback mechanism periodically providing information to facilitate a dynamic update of a congestion-avoidance parameter.
• the present invention is disclosed in context of use of a Gigabit Passive Optical Network (G-PON).
• G-PON Gigabit Passive Optical Network
• the principles of the present invention are not limited to use within a G-PON but may be applied to other PONs such as Broadband PON (BPON) or Ethernet PON (EPON).
• BPON Broadband PON
• EPON Ethernet PON
• ONT Optical Network Unit
• TCP Transmission Control Protocol
• OSI Open System Interconnection
• other transport protocols that provide a reliable transmission and a traffic congestion mechanism to slow down traffic during congestion may be used.
• the present invention is further disclosed in context of use of a modified Random Early Detection (RED) for providing centralized and dynamically-adjustable-congestion avoidance.
• RED Random Early Detection
• One skilled in the art would recognize other congestion avoidance schemes such as Weighted Random Early Detection could be modified to provide centralized and dynamically-adjustable- congestion avoidance.
• the illustration and description of the present invention in context of a G-PON having a modified RED for providing centralized and dynamically- adjustable-congestion- avoidance is merely one possible embodiment of the present invention.
• FIG. 1 an exemplary schematic diagram of a prior art communication system 10 having standalone congestion handling in a G-PON is shown.
• the G-PON includes an ONT 12 having a First-In-First-Out (FIFO) priority queue 14, a bidirectional communication link 18, and an OLT 22.
• the bidirectional communication link 18 facilitates communication between the ONT 12 and the OLT 22.
• a bidirectional communication link 16 allows subscribers of the ONT 12 to communicate with the G- PON via the ONT 12.
• the priority queue 14 stores packets from the subscriber that are to be sent to the OLT 22.
• the priority queue 14 is full having packets B, C, G, and H where B is at the head of the priority queue 14 and H is at the tail of the priority queue 14.
• Traffic 26 is sent from the ONT 12 to the OLT 22 and includes packet A that was previously in the priority queue 14.
• Traffic 24 is sent from the subscriber to the ONT 12 and includes packet I.
• FIG. 2 another exemplary schematic diagram of a prior art communication system 10 having standalone congestion handling in a G-PON is shown.
• the priority queue 14 With the priority queue 14 is full, congestion is handled by dropping incoming traffic from the subscriber.
• the traffic 24 ( Figure 1) is dropped via the ONT 12 prior to entering the priority queue 14.
• Dropping traffic due to the priority queue 14 being full also known as tail dropping, is an easy and inexpensive approach to handling congestion. However, this approach is weak and reacts when congestion occurs and is not a method to avoid congestion.
• RED/WRED uses various criteria such as average queue size to determine when to drop packets prior to entering the priority queue 14.
• including a standalone RED/WRED to the ONT 12 is costly and has limited effectiveness.
• the G-PON includes a plurality of ONTs 112 each having a Transmission Container (T-CONT) 115, a bidirectional communication link 118, a passive splitter 130, and an Optical Line Terminator (OLT) 122 having a feedback mechanism 128 and a DBA mechanism 130.
• T-CONT Transmission Container
• OLT Optical Line Terminator
• Communication between the ONTs 112 and the OLT 122 is facilitated via the bidirectional communication link 118 and the splitter 130. Subscribers of the ONT 112 are connected to the ONT 112 via a bidirectional communication link 116. Communication in the direction of the subscriber towards the OLT 122 is an upstream communication, whereas communication in the direction of the OLT 122 towards the subscriber is a downstream communication.
• the ONT 112 is shown including a single T-CONT 115.
• the ONT 112 may include a plurality of T-CONTs 115.
• a T-CONT 115 is used by the ONT 112 to transport multiplexed subscriber packets toward the OLT 122 using the bandwidth granted specifically to the T-CONT 115 by the OLT 122.
• the type of the T-CONT 115 determines the types of upstream bandwidth assigned by the OLT 122 including fixed, assured, non-assured and best-effort bandwidth.
• the T-CONT 115 shown includes a single FIFO single priority queue 114.
• the T-CONT 115 may include a plurality of priority queues 114 wherein each priority queue accommodates a subset of the priorities for the traffic from the subscriber. It would be recognized by those skilled in the art that the congestion handling for a plurality of priority queues 114 would need to consider the priority type of each priority queue 114.
• a packet from the subscriber is stored in the priority queue 114 and the packet may later be sent to the OLT 122 via the bidirectional communication link 118.
• the ONT 112(a) includes the priority queue 114(a) having packets B and C where B is at the head of the priority queue 114(a) and packet C is at the tail of the priority queue 114(a).
• the ONT 112(b) includes the priority queue 114(b) having the packets N, O, and P where N is at the head of the priority queue 114(b) and P is at the tail of the priority queue 114(b).
• the G-PON employs the DBA mechanism 130 in the OLT 122 to handle dynamically assigning bandwidth due to oversubscription as described in further detail below.
• the DBA mechanism 130 may be extended to include congestion- avoidance parameters.
• the exemplary embodiment uses RED congestion-avoidance parameter such as a minimum threshold, a maximum threshold and a maximum drop probability.
• the feedback mechanism 128 allows incremental adjustments to the congestion-avoidance parameters by analyzing the overall upstream bandwidth utilization from the ONTs 112 following each adjustment.
• the overall upstream bandwidth utilization may be determined from a queue fill level provided by the ONTs 112.
• the congestion-avoidance parameters are dynamically adjusted.
• the RED congestion-avoidance parameters are dynamically adjusted per T- CONT 115. It would be understood by those skilled in the art that if WRED were used the congestion-avoidance parameters would be dynamically adjusted per priority queue 114.
• the congestion-avoidance parameters in the standalone approach are static or at most administratively adjustable via a network operator. Furthermore, in the standalone approach a dynamic adjustment is not possible due to lack of visibility to bandwidth utilization.
• the feedback mechanism 128 should be periodically executed. 128. For example, the feedback mechanism 128 should be executed periodically every 2 milliseconds or less. Preferable, the feedback mechanism 128 is at the same cycle as the DBA mechanism 130.
• the DBA mechanism 130 and the feedback mechanism 128 provide for a centralized-congestion avoidance in the OLT 122. Therefore, the congestion avoidance according to the ONT 112 is remotely handled by the OLT 122.
• the ONT 112(a) will have information concerning the T-CONT 115 (a) to send to the OLT 122.
• the T-CONT information 156 is included in a Dynamic Bandwidth Report upstream (DBRu) portion 154 of a transmission 152 to the OLT 122.
• the T- CONT information 156 may include, for example, the queue fill level and an identifier of the T-CONT 115(a).
• the queue fill level is the amount of data currently stored in the priority queue 114(a). In the case that T-CONT 115(a) has a plurality of priority queues, it would be understood that the amount of data stored may be individually reported for each priority queue.
• the DBA mechanism 130 uses the T-CONT information 156 to determine an adjusted bandwidth allocation for the T-CONT 115(a). Furthermore, the feedback mechanism 128 analyzes upstream bandwidth utilization for each of the bandwidth type. The information in both the DBA mechanism 130 and the feedback mechanism 128 are used to dynamically update the parameters that are used in determining if a packet should be discarded.
• a transmission 160 from the OLT 122 to the ONT 112 includes an upstream bandwidth map having T-CONT 115 specific allocation structures including a packet- drop command 164.
• the packet-drop command 164 may include, for example the following:
• the exemplary transmission 160(a) includes the packet-drop command 164(a) to discard a packet in the priority queue 114(a) and to start dropping all incoming packets to the queue 114(a).
• the ONT 112(a) may discard the packet at the head of the priority queue 114(a).
• the ONT 112(a) may discard the packet at the tail of the priority queue 114(a).
• Figure 5 shows an update of the ONT 112(a) after executing the packet-drop command 164(a) from Figure 4 by the ONT 112(a). In this case, the packet at the head of the queue 114(a) was discarded and the head of the priority queue 114(a) was updated.
• the ONT 112(a) After receiving a command to drop incoming packets, the ONT 112(a) continues to drop incoming packets until the ONT 112(a) receives a transmission 160(b) having a subsequent packet-drop command 164(b). At which point the subsequent packet-drop command 164(b) is executed.
• the exemplary transmission 160(b) includes the packet-drop command 164(a) to stop dropping all incoming packets.
• the OLT receives the available priority queue size from each ONT 200.
• the available queue size is a size that is provisioned at the ONT.
• the OLT then initializes the congestion-avoidance parameters for each priority queue 202. The initialization is based on the received priority queue sizes and may further be based on network operator input. [0030] Next, the OLT determines a target queue size for each priority queue 204. The determination is based on the initial values of the congestion-avoidance parameters. A preferable target queue size places the queue in equilibrium so that the available amount of bandwidth is proportional to the regulated amount of upstream traffic. Thus, congestion is avoided.
• a T-CONT traffic waiting at an ONT is detected by the OLT 206.
• the OLT allocates bandwidth to the T-CONT 208 when available. This information is sent to the OLT.
• the OLT in turn sends a DBRu request for the queue fill level.
• the OLT receives the DBRu status report and updates the bandwidth allocation for the T-CONT.
• the OLT updates the RED congestion-avoidance parameter and the target queue size 212. Additionally, the OLT calculates the average queue size and determines a packet drop command 214. The average queue size is used to smooth out transient bursts so the drop command is issued using on a drop probability that corresponds to the long-term traffic amount. The OLT sends the updated bandwidth allocation and drop commands to the ONT 216. The ONT executes the drop command and updates the bandwidth 118.
• elements of the method may be repeated. That is elements may be looped through multiple times. For example, if the OLT receives a plurality of T CONTs then elements 206-218 may be looped though for each received T_CONT.
• the present invention has been described with the ONT 112 having a single priority queue 114 it would be understood by those skilled in the art that the ONT 112 may include plurality of priority queues 114, in which case a modified WRED mechanism could be used. Furthermore, different modes for a DBRu report may be used to support a report containing a fill level for each queue in the case of multiple priority queues. Also, additional drop commands may need to be defined to encode the queue types in the commands.

Landscapes

• Engineering & Computer Science (AREA)
• Computer Networks & Wireless Communication (AREA)
• Signal Processing (AREA)
• Physics & Mathematics (AREA)
• Electromagnetism (AREA)
• Small-Scale Networks (AREA)
• Data Exchanges In Wide-Area Networks (AREA)

Priority Applications (5)

Applications Claiming Priority (2)

Publications (1)

Family

ID=37562821

Family Applications (1)

Country Status (6)

Cited By (1)

Families Citing this family (15)

Citations (2)

Family Cites Families (10)

• 2006
  • 2006-09-28 EP EP06806865A patent/EP1989914B1/en not_active Not-in-force
  • 2006-09-28 KR KR1020087023085A patent/KR101314053B1/ko not_active Expired - Fee Related
  • 2006-09-28 WO PCT/EP2006/066838 patent/WO2007096006A1/en not_active Ceased
  • 2006-09-28 CN CN2006800532473A patent/CN101385383B/zh not_active Expired - Fee Related
  • 2006-09-28 US US12/280,311 patent/US8078057B2/en active Active
  • 2006-09-28 JP JP2008555643A patent/JP4996629B2/ja not_active Expired - Fee Related

Patent Citations (2)

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events