SYSTEM AND METHOD FOR EFFICIENT TRANSPORT OF LARGE DATA FILES

• US 20150236957A1
• Filed: 02/20/2015
• Published: 08/20/2015
• Est. Priority Date: 02/20/2014
• Status: Active Grant

• Current Assignee: UVic Industry Partnerships Inc. (University of Victoria)
• Original Assignee: UVic Industry Partnerships Inc. (University of Victoria)
• Inventors: Darcie, Thomas E., Ganti, Sudhakar, Neville, Stephen W., Albanese, Ilijc, Yazir, Yagiz O.
• Granted Patent: US 9,800,509 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: H04L 12/6418 Hybrid transport
• CPC Class Codes: H04L 47/193 at the transport layer, e.g. TCP related
• CPC Class Codes: H04L 47/22 Traffic shaping
• CPC Class Codes: H04L 47/30 in combination with information about buffer occupancy at either end or at transit nodes
• CPC Class Codes: H04L 47/36 by determining packet size, e.g. maximum transfer unit [MTU]

Abstract

Telecommunication networks are under stress due to rapid traffic increase cause mostly by large file transfers. Disclosed herein is a cross-layer transport protocol specifically designed to efficiently handle large transactions. Traffic generated from large transactions is shaped into a periodic succession of fixed-size data frames. Each transaction can then be scheduled for transmission using a two-way reservation protocol. Exemplary results show that the proposed approach is capable of significantly improving goodput and end-to-end delay relative to TCP, improving efficiency of bandwidth utilization by over 40%.

Claims

1. A method for transporting a data file, comprising:

dividing the data file into a plurality of data frames;

concatenating the plurality of data frames into at least one chain with a predefined transparency degree;

associating the at least one chain with a control packet comprising information regarding at least a chain length of the at least one chain and the predefined transparency degree;

scheduling transmission of the at least one chain along a predetermined path of a network based on the control packet; and

transmitting each data frame along the predetermined path at a time derived from an estimated time-to-transmit parameter.

2. The method of claim 1, further comprising concatenating the plurality of data frames into the at least one chain with one or more void frames.

3. The method of claim 1, wherein dividing the data file into a plurality of data frames comprises selecting a data frame size matching that of a transport layer.

4. The method of claim 1, further comprising using a buffer having a buffer size proportional to a data frame size and a transparency degree to delay the at least one chain so as to interleave the at least one chain with a previously received chain or an incoming chain.

5. The method of claim 4, wherein a maximum buffer size is selected based on a maximum data frame size and a maximum transparency degree of the at least one chain.

6. The method of claim 1, wherein the data file has a size of at least 100 MB.

7. The method of claim 1, wherein the predefined transparency degree of the at least one chain is associated with a number of periods of the at least one chain having at least one timeslot occupied by at least one data frame.

8. The method of claim 1, wherein the predefined transparency degree is associated with a number of void frames in the at least one chain.

9. The method of claim 1, wherein the estimated time-to-transmit parameter is included in the control packet and is communicated to at least one node of the predetermined path and is associated with a delay between control packet arrival and chain reception commencement.

10. A system for transporting a data file, comprising at least one source server comprising a software module configured to:

divide the data file into a plurality of data frames;

concatenate the plurality of data frames into at least one chain with a predefined transparency degree; and

associate the at least one chain with a control packet comprising information regarding at least a chain length of the at least one chain and the predefined transparency degree.

11. The system of claim 10, wherein the software module is further configured to:

schedule transmission of the at least one chain along a predetermined path of a network based on the control packet; and

transmit each data frame along the predetermined path at a time derived from an estimated time-to-transmit parameter.

12. The system of claim 10, wherein the at least one chain further comprises one or more void frames.

13. The system of claim 10, wherein the plurality of data frames comprises data frames of different sizes.

14. The system of claim 10, wherein the predefined transparency degree of the at least one chain is associated with a number of periods of the at least one chain comprising at least one timeslot occupied by at least one data frame.

15. The system of claim 10, wherein the predefined transparency degree is associated with a number of void frames in the at least one chain.

16. The system of claim 10, wherein the estimated time-to-transmit parameter is included in the control packet and is communicated to at least one node of the predetermined path and is associated with a delay between control packet arrival and chain reception commencement.

17. The system of claim 10, wherein the system includes a network interface that communicates the at least one chain and wherein the network interface is an optical transport network or an Ethernet-based network.

18. A network node, comprising a network interface configured to receive at least one chain that includes a plurality of data frames and void frames, or a portion of the at least one chain, wherein the network node transmits an updated estimated time-to-transmit parameter to a destination network.

19. The network node of claim 18, wherein the network interface is configured to receive a control packet associated with the at least one chain and configured to extract an estimated time-to-transmit parameter from the control packet.

20. The network node of claim 18, wherein the updated estimated time-to-transmit parameter is determined by adding an original estimated time-to-transmit parameter of the network node to an original buffering time of the network node and subtracting an estimated control packet processing time for the network node.