Method to achieve bounded buffer sizes and quality of service guarantees in the internet network
Methods to achieve bounded router buffer sizes and Quality of Service guarantees for traffic flows in a packet-switched network are described. The network can be an Internet Protocol (IP) network, a Differentiated Services network, an MPLS network, wireless mesh network or an optical network. The routers can use input queueing, possibly in combination with crosspoint queueing and/or output queueing. Routers may schedule QoS-enabled traffic flows to ensure a bounded normalized service lead/lag. Each QoS-enabled traffic flow will buffer O(K) packets per router, where K is an integer bound on the normalized service lead/lag. Three flow-scheduling methods are analysed. Non-work-conserving flow-scheduling methods can guarantee a bound on the normalized service lead/lag, while work-conserving flow-scheduling methods typically cannot guarantee the same small bound. The amount of buffering required in a router can be reduced significantly, the network links can operate near peak capacity, and strict QoS guarantees can be achieved.
1. A switch for switching a plurality of guaranteed-rate (GR) traffic flows over a set of output ports, given a scheduling frame comprising F time-slots for integer F, comprising:
N input ports for receiving packets and M output ports for transmitting packets, for integers N and M;
a plurality of queues, wherein each queue buffers packets which arrive at a common one of said N input ports and which depart on a common one of said M output ports;
wherein each GR traffic flow is associated with a guaranteed data-rate requirement, wherein each GR traffic flow is associated with one queue, and wherein packets associated with a GR traffic flow are buffered in its associated queue;
memory for storing a flow-schedule,
wherein the flow-schedule determines which of the GR traffic flows associated with said queues, if any, have a reservation to transmit in each time-slot in said scheduling frame;
wherein the flow-schedule provides each of said GR traffic flows with a guaranteed number of time-slot reservations per scheduling frame sufficient to meet its guaranteed data-rate requirement,
and wherein for each GR traffic flow with a guaranteed data-rate requirement which exceeds one packet transmission per scheduling frame, said flow-schedule provides said GR traffic flow with a substantially equal number of time-slot reservations for transmission in each half of the scheduling frame with a substantially equal size.
2. The switch of claim 1 ,
wherein the flow-schedule provides each of said GR traffic flows with a substantially pro-rated number of time-slot reservations for transmission in a subset of the scheduling frame comprising time-slots 1 . . . E for integer E, where 1<=E<=F.
3. The switch of claim 1 , wherein the memory is distributed throughout the switch.
4. The switch of claim 1 , wherein the GR traffic flows belong to a traffic class in the Differentiated Services service model.
5. The switch of claim 1 , further comprising a traffic-shaper module,
comprising a controller, memory to buffer packets, memory to buffer tokens, and a token-generator,
wherein said traffic-shaper module is operable to buffer incoming packets associated with one of the GR traffic flows, and release said packets in a manner to reduce the burstiness of said GR traffic flow.
6. The switch of claim 1 , wherein each of said queues is partitioned into a plurality of flow-queues,
and wherein each flow-queue buffers packets associated with one GR traffic flow.
7. The switch of claim 2 , wherein each of said queues is partitioned into a plurality of flow-queues,
and wherein each flow-queue buffers data associated with one GR traffic flow.
8. The switch of claim 1 ,
further comprising memory for storing a queue-schedule,
wherein each of said plurality of queues is associated with a guaranteed data-rate requirement,
wherein said queue-schedule determines which of said queues, if any, have a reservation to transmit in each time-slot of said scheduling frame,
and wherein said queue-schedule provides each of said queues with a guaranteed number of time-slot reservations per scheduling frame sufficient to satisfy its guaranteed data-rate requirement.
9. The switch of claim 8 ,
wherein for each queue with a guaranteed data-rate requirement which exceeds one packet transmission per scheduling frame, said queue-schedule provides that queue with a substantially equal number of time-slot reservations for transmission in each half of the scheduling frame with a substantially equal size.
10. The switch of claim 9 ,
wherein the queue-schedule provides each queue with a substantially pro-rated number of time-slot reservations for transmission in a subset of said scheduling frame comprising time-slots 1 . . . E for integer E, where 1<=E <=F.
11. The switch of claim 1 which also switches a plurality of class-based (CB) traffic flows over a set of output ports, comprising
a plurality of class-queues, wherein each class-queue is associated with a guaranteed data-rate requirement,
wherein each class-queue buffers packets which arrive at a common one of said N input ports and depart on a common one of said M output ports, wherein each CB traffic flow is associated with one class-queue, and wherein packets associated with a CB traffic flow are buffered in its associated class-queue;
wherein a class-queue can transmit in the time-slots when a corresponding queue is not transmitting.
12. The switch of claim 11 , further comprising
memory to store a class-schedule,
wherein each class-queue is associated with a guaranteed data-rate requirement,
wherein said class-schedule determines which of said class-queues, if any, has a reservation to transmit in each time-slot in said scheduling frame;
and wherein said class-schedule provides each of said class-queues with a guaranteed number of time-slot reservations in a scheduling frame sufficient to meet its guaranteed data-rate requirement.
13. The switch of claim 12 ,
wherein for each class-queue with a guaranteed data-rate requirement which exceeds one packet transmission per scheduling frame, said class-schedule provides each of said class-queues with a substantially equal number of time-slot reservations for transmission in each half of the scheduling frame with a substantially equal size.
14. The switch of claim 13 ,
wherein said class-schedule provides each of said class-queues with a substantially pro-rated number of time-slot reservations for transmission in a subset of the scheduling frame comprising time-slots 1 . . . E for integer E, where 1<=E <=F.
15. The switch of claim 12 ,
wherein the CB traffic flows belong to a traffic class in the Differentiated Services service model.
16. The switch of claim 10 , further associated with a scheduling-processor,
wherein said scheduling-processor will process the contents of a class-queue, to determine which of the packets buffered within said class-queue is selected to transmit next.
17. The switch of claim 16 ,
wherein said scheduling-processor selects an oldest packet in the class-queue to transmit next.
18. The switch of claim 1 which also switches a plurality of Best-Effort (BE) traffic flows over a set of output ports, comprising
a plurality of BE-queues, wherein each BE-queue buffers packets which arrive at a common one of said N input ports and depart on a common one of said M output ports, wherein a plurality of one or more BE traffic flows is associated with each BE-queue, and wherein packets associated with a BE traffic flow are buffered in its associated BE-queue;
wherein a BE-queue can transmit in the time-slots when a corresponding queue is not transmitting.
19. The switch of claim 11 which also switches a plurality of Best-Effort (BE) traffic flows over a set of output ports, comprising
a plurality of BE-queues, wherein each BE-queue buffers packets which arrive at a common one of said N input ports and depart on a common one of said M output ports, wherein a plurality of one or more BE traffic flows is associated with each BE-queue, and wherein packets associated with a BE traffic flow are buffered in its associated BE-queue;
wherein a BE-queue can transmit in the time-slots when a corresponding queue is not transmitting, and when a corresponding class-queue is not transmitting.
20. A packet-switched network to deliver a plurality of guaranteed rate (GR) traffic flows through a plurality of switches,
wherein each of said plurality of GR traffic flows originates at a source node, terminates at a destination node, and traverses a fixed path of intermediate switches between its source node and its destination node, and wherein each GR traffic flow is associated with a guaranteed data-rate requirement;
said network comprising a plurality of J switches for integer J, wherein said switches satisfy the requirements of claim 1 .
21. The network of claim 20 ,
wherein for each switch, the schedule for said switch provides each flow-queue within said switch with a substantially pro-rated number of time-slot reservations for transmission in a subset of the scheduling frame comprising time-slots 1 . . . E for integer E, where 1<=E<=F, wherein F is the length of the scheduling-frame for said switch.
22. The network of claim 20 ,
wherein the memory within each switch is distributed throughout the switch.
23. The network of claim 20 , wherein the GR traffic flows belong to a traffic class in the Differentiated Services service model.
24. The network of claim 20 ,
wherein each switch further satisfies the requirements of claim 9 .
25. The network of claim 24 ,
wherein each switch further satisfies the requirements of claim 10 .
26. The network of claim 20 ,
wherein each switch further satisfies the requirements of claim 13 .
27. The network of claim 26 ,
wherein each switch further satisfies the requirements of claim 14 .
28. The network of claim 20 ,
wherein at least one of said plurality of GR traffic flows is associated with a plurality of fixed paths between its source node and its destination node,
wherein each fixed path comprises a plurality of intermediate switches,
and wherein each fixed path in said plurality of fixed paths is associated with a guaranteed data-rate requirement.
29. The network of claim 20 ,
wherein a source node of at least one of said GR traffic flows comprises a traffic shaper module,
wherein said traffic-shaper module comprises a controller, memory to buffer packets, memory to buffer tokens, and a token-generator,
wherein said traffic-shaper module is operable to buffer packets associated with said GR traffic flow, and release said packets for transmission along a fixed path with a relatively constant rate of transmission.