Virtual burst-switching networks

• US 7,889,723 B2
• Filed: 06/09/2008
• Issued: 02/15/2011
• Est. Priority Date: 10/09/2003
• Status: Expired due to Fees

• Current Assignee: RPX Clearinghouse LLC (RPX Corporation)
• Original Assignee: Nortel Networks Limited (Nortel Networks Corporation)
• Inventors: Jamoussi, Bilel N., Beshai, Maged E.
• Primary Examiner(s): Ton, Dang T
• Assistant Examiner(s): O CONNOR, BRIAN T
• Application Number:
  US12/135,526
• Publication Number:
  US 20080232809A1
• Time in Patent Office: 981 Days
• Field of Search: 370/352, 370/355, 370/360, 370/389, 370/390, 370/391, 370/400, 370/401, 370/404, 370/477, 398/45, 398/47, 398/48, 398/50, 398/51, 398/52, 398/54, 398/59
• US Class Current: 370/355
• CPC Class Codes: H04L 12/56 Packet switching systems
• CPC Class Codes: H04L 45/122 by minimising distances, e.g. by selecting a route with minimum of number of hops
• CPC Class Codes: H04L 45/124 using a combination of metrics
• CPC Class Codes: H04L 45/20 Hop count for routing purposes, e.g. TTL
• CPC Class Codes: H04L 45/302 Route determination based on requested QoS
• CPC Class Codes: H04L 45/40 Wormhole routing
• CPC Class Codes: H04L 45/62 Wavelength based optical switching H04Q11/0062
• CPC Class Codes: H04Q 11/0066 Provisions for optical burst or packet networks
• CPC Class Codes: H04Q 2011/0045 Synchronisation
• CPC Class Codes: H04Q 2011/0064 Arbitration, scheduling or medium access control aspects

Abstract

A time-shared network comprising edge nodes and optical core nodes may be dynamically divided into several embedded networks, each of which covering selected edge nodes. At least one of the edge nodes may host an embedded-network controller operable to form multiple-source flow-rate allocation requests each of the requests specifying flow-rate allocations to a plurality of paths from several source nodes to several sink nodes. A core node may also host an embedded-network controller or several embedded-network controllers. The time-shared network may use both time-division multiplexing and burst switching.

Claims

1. An edge node comprising:

a plurality of ingress ports for receiving data packets from subtending data sources;

a plurality of inbound ports for receiving inbound data bursts from a plurality of bufferless core nodes and disassembling said inbound data bursts into constituent data packets;

a switch fabric;

an edge controller;

a plurality of egress ports for transmitting data packets to subtending data sinks; and

a plurality of outbound ports, where each outbound port is coupled to a burst-formation device, said burst-formation device;

forms outbound data bursts each containing a number of data packets received from ingress ports and from inbound ports through said switch fabric; and

transmits said outbound data bursts to a selected core node among said plurality of bufferless core nodes;

where said burst-formation device forms said outbound data bursts according to at least one of two modes;

in a first mode;

said each burst-formation device independently forms said bursts from already waiting data packets; and

sends scheduling requests to a corresponding core node;

in a second mode;

said each burst-formation device sends flow-rate allocation requests to a respective core node;

receives burst-transfer permits containing burst-formation instructions from said respective core nodes; and

forms constrained bursts according to said burst-formation instructions.

2. The edge node of claim 1 wherein said scheduling requests include burst descriptors, each burst descriptor specifying a burst size.

3. The edge node of claim 2 wherein at least one of said inbound ports receives, from a core node among said plurality of core nodes:

burst-transfer schedules for waiting bursts in accordance with said first mode; and

burst-transfer permits for forthcoming bursts in accordance with said second mode.

4. The edge node of claim 3 wherein said edge controller:

receives incoming control signals from said ingress ports and from said plurality of core nodes through said inbound ports and said switch fabric; and

transmits outgoing control signals to said egress ports and to said plurality of core nodes through said switch fabric and said outbound ports.

5. The edge node of claim 1 further comprising a virtual-network controller for:

receiving traffic-data information from a set of selected edge nodes connecting to said plurality of bufferless core nodes;

forming a matrix of flow-rate-allocation requirements from each of said selected edge nodes to at least one other selected edge node within said set of selected edge node; and

communicating said matrix to at least one of said bufferless core nodes.

6. The edge node of claim 1 wherein said edge controller:

determines a round-trip propagation delay τ

between the edge node and said selected core node;

determines aggregate flow rate ρ

of burst streams already being transferred from a respective outbound port among said plurality of outbound ports to said selected core node;

determines a proximity index as a ratio of said round-trip propagation delay to a prescribed delay tolerance; and

determines a storage index as ρ

×

τ

/Λ

where Λ

is a size of an outbound buffer of said respective outbound port connecting to said selected core node; and

selects said first mode, for a new burst stream directed to said selected core node, subject to a condition of said proximity index not exceeding a first predefined threshold and said storage index not exceeding a second predefined threshold.

7. A core node interconnecting a plurality of edge nodes, said core node comprising:

a plurality of input ports for receiving optical signals from said plurality of edge nodes;

an optical switching fabric for switching said optical signals to a plurality of output ports, said output ports transmitting said optical signals to respective edge nodes; and

a core controller including a burst-descriptor generator and a burst scheduler;

wherein said core controller receives;

burst-admission requests; and

multiple-source connection requests each specifying at least two source edge nodes, and for each of the at least two source edge nodes;

at least one sink edge node; and

a specified flow-rate allocation;

wherein each of said burst-admission requests contains;

a request identifier “

00”

;

an identifier of a requesting edge node;

a number χ

.>

0 of destinations for which burst-scheduling is requested;

multiple-destination burst-description records, one for each of said χ

destination, each record including;

a cyclic burst number relevant to said requesting edge node;

an identifier of a sink edge-node to which a corresponding burst is destined; and

a size of said corresponding burst.

8. The core node of claim 7 further comprising a burst-transfer-request processor for:

submitting an indication of said size of said corresponding burst to said burst scheduler; and

directing said specified flow-rate allocation to said burst-descriptor generator;

wherein said burst-descriptor generator;

determines burst-transfer permits corresponding to said specified flow-rate, each burst-transfer permit stating a nominal burst size and inter-burst time interval for a burst-stream of anticipated bursts; and

submits said burst-transfer permits to said burst scheduler;

and wherein said burst scheduler determines a requisite arrival time for said corresponding burst and requisite arrival times for said anticipated bursts.

9. The core node of claim 7 wherein each said multiple-source connection requests includes:

a request identifier “

11”

;

an identifier of a requesting edge node;

identifiers of said at least two source edge nodes; and

a super record for each of said at least two source edge nodes, said super record indicating said at least one sink edge node and said specified flow-rate allocation.

10. A network comprising:

a plurality of edge nodes, each edge node having an edge controller;

a plurality of bufferless core nodes, each core node having a core controller, and

links connecting said edge nodes to said core nodes;

wherein said each edge node has a link to said each core node;

wherein said each edge node is time-locked to said each core node; and

wherein at least one of said edge nodes hosts a virtual-network controller, said virtual-network controller;

forms multiple-source flow-rate allocation requests where each of said multiple-source flow-rate allocation requests specifies flow-rate allocations for each directed pair of selected edge nodes among said plurality of edge nodes; and

sends each multiple-source flow-rate allocation request to a selected core node among said plurality of core nodes.

11. The network of claim 10 wherein said virtual-network controller directs said each multiple-source flow-rate allocation request to at least one other core node among said core nodes.

12. The network of claim 10 further comprising a burst-transfer regulator coupled to said selected core node, the burst transfer regulator:

parses said each multiple-source flow-rate allocation request into constituent data streams, each constituent data stream defined according to a source edge node and a destination edge node;

determines a time interval between successive bursts and an upper bound of burst size for said each constituent data stream as a function of a respective flow-rate allocation; and

generates time-regulated burst descriptors for said each constituent data stream, each burst descriptor specifying said upper bound and said time interval.

13. A parent network comprising:

a plurality of independent bufferless core nodes;

a plurality of edge nodes each edge node having a time-locked path to each core node of said plurality of independent bufferless core nodes; and

a plurality of virtual-network controllers, each virtual-network controller coupled to a respective edge node among said plurality of edge nodes;

said parent network being shared by a plurality of embedded networks, each embedded network comprising a respective set of selected edge nodes among said plurality of edge nodes and a respective virtual-network controller, among said plurality of virtual-network controllers, configured to;

determine flow-rate requirements for each directed pair of edge nodes within said respective set of edge nodes;

determine capacity requirements of paths between each edge node of said respective set of selected edge nodes and said each core node; and

communicate said capacity requirements to at least one of said plurality of independent bufferless core nodes.

14. The parent network of claim 13 wherein said respective virtual-network controller performs automated network-sizing for updating said capacity requirements to follow spatial and temporal traffic-level fluctuations.

15. The parent network of claim 14 wherein said each core node has a core controller for scheduling capacity requirements and communicating specifics of scheduled capacity to respective virtual-network controllers.

16. The parent network of claim 14 wherein a selected core node among said plurality of independent bufferless core nodes hosts a virtual-network controller for each of at least one embedded network.

17. The parent network of claim 14 wherein said automated network-sizing determines said capacity requirements for upstream paths from said respective set of selected edge nodes to the independent bufferless core nodes and downstream paths from the independent bufferless core nodes to said respective set of selected edge nodes.