System and method for recovery of packets in overlay networks
AssigneeLTN Global Inc.
InventorsYair Amir, Jonathan Stanton, John William Lane, John Lane Schultz
Life of this patent
Estimating the term…
Abstract
A packet recovery transmits N requests for retransmission of the transmitted packet to the sender node at N scheduled times upon determining satisfaction of a packet recovery condition associated with a transmitted packet from a sender node at a receiver node. At least one of the N scheduled times includes a receiver waiting time. M retransmissions of the transmitted packet from sender node at M scheduled times is performed upon the first receipt of a request for retransmission of the transmitted packet. At least one of the M scheduled times includes a sender waiting time. N and M are non-zero positive integers, and at least one of N or M is greater than one.
Claims40
A packet recovery method, comprising:
- upon determining satisfaction of a packet recovery condition associated with a transmitted packet from a sender node at a receiver node, transmitting N requests for retransmission of the transmitted packet to the sender node at N scheduled times, wherein at least one of the N scheduled times includes a receiver waiting time, wherein the receiver waiting time is a non-zero delay less than a round trip latency between the sender and receiver nodes;
- and
- upon the first receipt of a request for retransmission of the transmitted packet, performing M retransmissions of the transmitted packet from sender node at M scheduled times, wherein at least one of the M scheduled times includes a sender waiting time, wherein the sender waiting time is a non-zero delay between the end of one retransmission and the start of a subsequent retransmission that is also less than the round trip latency, wherein N and M are non-zero positive integers, and wherein at least one of N or M is greater than one.
A method for recovering packets, comprising:
- a) upon determining satisfaction of a packet recovery condition for a transmitted packet from a sender node to a receiver node:
- associating N scheduled times with N requests for retransmission of the transmitted packet to the sender node, wherein at least one of the N scheduled times includes a receiver waiting time, wherein the receiver waiting time includes a non-zero delay that is less than a round trip latency between the sender and receiver nodes;
- and
- at each scheduled time, transmitting a request for retransmission of the transmitted packet;
- and
- b) upon the first receipt of a request for retransmission of the transmitted packet at the sender node:
- associating M scheduled times with M retransmissions of the transmitted packet to the receiver node, wherein at least one of the M scheduled times includes a sender waiting time, wherein the sender waiting time includes a non-zero delay between the end of one retransmission and the start of a subsequent retransmission that is less than the round trip latency between the sender and receiver nodes;
- and
- at each scheduled time, retransmitting the transmitted packet, wherein the N and M are non-zero positive integers and wherein at least one of N or M is greater than one.
A packet recovery method, comprising:
- upon determining satisfaction of a packet recovery condition associated with a transmitted packet from a sender node at a receiver node, transmitting N requests for retransmission of the transmitted packet to the sender node at N scheduled times, wherein at least one of the N scheduled times includes a receiver waiting time, wherein the receiver waiting time includes a non-zero delay that is less than a round trip latency between the sender and receiver nodes;
- and
- upon the first receipt of a request for retransmission of the transmitted packet, using an erasure coding function, with integer parameter k greater than 1 and less than M, to generate M encoded blocks such that any k out of M encoded blocks are sufficient to recreate the transmitted packet at the receiver node,
- transmitting the M encoded blocks to the receiver node at M scheduled times, wherein at least one of the M scheduled times includes a sender waiting time, wherein N and M are non-zero positive integers, and wherein M is greater than two.
A method for recovering packets, comprising:
- a) upon determining satisfaction of a packet recovery condition for a transmitted packet from a sender node to a receiver node:
- associating N scheduled times with N requests for retransmission of the transmitted packet to the sender node, wherein at least one of the N scheduled times includes a receiver waiting time, wherein the receiver waiting time includes a non-zero delay that is less than a round trip latency between the sender and receiver nodes;
- and
- at each scheduled time, transmitting a request for retransmission of the transmitted packet;
- and
- b) upon the first receipt of a request for retransmission of the transmitted packet at the sender node:
- using an erasure coding function, with integer parameter k greater than 1 and less than M, to generate M encoded blocks such that any k out of M encoded blocks are sufficient to recreate the transmitted packet at the receiver node;
- and
- associating M scheduled times, each of them with one of the M encoded blocks, wherein at least one of the M scheduled times includes a sender waiting time;
- and
- at each scheduled time, transmitting the corresponding encoded block to the receiver node, wherein the N and M are non-zero positive integers and M is greater than two.
A packet recovery method at a receiver node, comprising:
- only upon the first determination of satisfaction of a packet recovery condition associated with a transmitted packet to the receive node, transmitting N requests for retransmission of the transmitted packet at N scheduled times, wherein at least one of the N scheduled times includes a receiver waiting time, wherein the receiver waiting time is a non-zero delay less than a round trip latency between the sender and receiver nodes;
- wherein N is an integer greater than one;
- and
- receiving M retransmissions of the transmitted packet, wherein M is a non-zero positive integer.
A packet retransmission method at a sender node, comprising:
- receiving N requests for retransmission of a transmitted packet by the sender node;
- and
- only upon the first receipt of a request for retransmission of the transmitted packet, performing M retransmissions of the transmitted packet at M scheduled time, wherein at least one of the M scheduled time includes a sender waiting time, wherein the sender waiting time is a non-zero delay between the end of one retransmission and the start of a subsequent retransmission that is also less than the round trip latency;
- wherein N and M are non-zero positive integers, and wherein N and M are greater than one.
A receiver node, comprising:
- a processor configured to make a determination that a packet recovery condition associated with a transmitted packet to the receiver node is satisfied;
- a transmitter that only upon the first determination of satisfaction of the packet recovery condition transmits N requests for retransmission of the transmitted packet at N scheduled times, wherein at least one of the N scheduled times includes a receiver waiting time, wherein the receiver waiting time is a non-zero delay less than a round trip latency between the sender and receiver nodes;
- and wherein N is an integer greater than one;
- and
- a receiver configured to receive M retransmissions of the transmitted packet, wherein M is a non-zero positive integer.
A sender node, comprising:
- a receiver configured to receive N requests for retransmission of a transmitted packet by the sender node;
- a processor configured to make a determination of the first receipt of a request for retransmission of the transmitted packet;
- and
- a transmitter that only upon the first receipt of a request for retransmission of the transmitted packet, performs M retransmissions of the transmitted packet at respective M scheduled times, including a sender waiting time that is a non-zero delay between the end of one retransmission and the start of a subsequent retransmission that is also less than the round trip latency, wherein N and M are non-zero positive integers, and wherein N and M are greater than one.
Dependencies are inferred from the claim text.