System and method for peer-to-peer multi-party voice-over-IP services

• US 20070211703A1
• Filed: 03/10/2006
• Published: 09/13/2007
• Est. Priority Date: 03/10/2006
• Status: Active Grant

• Current Assignee: Meta Platforms, Inc. (f/k/a Facebook, Inc.)
• Original Assignee: International Business Machines Corporation
• Inventors: Shae, Zon-Yin, Gu, Xiaohui, Wen, Zhen, Yu, Philip
• Granted Patent: US 7,379,450 B2
• Time in Patent Office: Days
• US Class Current: 370/356
• CPC Class Codes: H04L 12/1818 Conference organisation arrangements, e.g. handling schedules, setting up parameters needed by nodes to attend a conference, booking network resources, notifying involved parties
• CPC Class Codes: H04L 45/16 Multipoint routing
• CPC Class Codes: H04L 45/48 Routing tree calculation

Abstract

A system, method, and computer program product for establishing multi-party VoIP conference audio calls in a distributed, peer-to-peer network where any number of nodes are able to arbitrarily and asynchronously start or stop producing audio output to be mixed into a single composite audio stream that is distributed to all nodes. A single distribution tree is used that has optimal communications characteristics to distribute the composite audio signal to all nodes. An audio mixing tree is established and maintained by adaptively and dynamically adding and merging intermediate mixing nodes operating between user nodes and the root of the single distribution tree. The intermediate mixing nodes and the root of the single distribution tree are all hosted, in an exemplary embodiment, on user nodes that are endpoints of the distribution tree.

Claims

1. A method for providing multi-party voice-over-IP services, the method on an information processing system comprising:

establishing a peer-to-peer distribution tree to broadcast a composite audio signal from a root mixing node to a plurality of participating nodes, wherein a subset of the participating nodes are originating nodes that provide a respective audio stream output received by the root mixing node to produce the composite audio signal; and

identifying, independently of the peer-to-peer distribution tree, a first intermediate mixing node from within the plurality of participating nodes, the first intermediate mixing node providing to the root mixing node a mixing of respective audio stream outputs provided by a first subset of originating nodes to the first intermediate mixing node.

2. The method of claim 1, wherein the root mixing node directly receives at least one of audio stream output from a second subset of originating nodes within the participating nodes, the method further comprising:

monitoring a processing load on the root mixing node;

identifying, in response to an increase in the processing load on the root mixing node and independently of the distribution tree, a second intermediate mixing node from the at least one neighbor; and

assigning, in response to the identifying the second intermediate mixing node, the second subset of originating nodes to the second intermediate mixing node and configuring the second intermediate mixing node to provide a mixing of the at least one audio stream outputs from the second subset to the root mixing node.

3. The method of claim 1, further comprising:

identifying, independently of the peer-to-peer distribution tree, a backup intermediate mixing node from within the plurality of participating nodes;

monitoring, at the backup intermediate mixing node, a first status of the first intermediate mixing node; and

assigning, in response to the first status, the first subset of originating nodes to the backup intermediate mixing node, thereby causing the backup intermediate mixing node to provide a mixing of respective audio stream outputs provided by the first subset of originating nodes to the root mixing node.

4. The method of claim 3, wherein the monitoring the first status comprises receiving a repetitive heartbeat message from the first mixing node, and wherein the assigning is performed in response to a failure to receive the repetitive heartbeat message.

5. The method of claim 3, wherein the first intermediate mixing node provides the mixing of respective audio stream outputs to the root mixing node by data communications through at least one higher level mixing nodes, the method further comprising:

identifying a number of additional backup intermediate mixing nodes, the number of additional backup mixing nodes being dependent upon a number of the at least one higher level mixing nodes through which the mixing is communicated from the first intermediate mixing node to reach the root mixing node.

6. The method of claim 1, further comprising:

monitoring a processing load on the first intermediate mixing node and on at least one neighbor node within the participating nodes;

identifying, in response to an increase in the processing load on the first intermediate mixing node and independently of the distribution tree, a second intermediate mixing node from the at least one neighbor; and

assigning, in response to the identifying, a portion of the first subset of originating nodes to the second intermediate mixing node and configuring the second intermediate mixing node to provide a mixing of the portion of the first subset of originating nodes to a parent mixing node of the first intermediate mixing node.

7. The method of claim 6, further comprising determining an initiation, at a non-originating node within the participating nodes, of a production of an additional audio stream to be delivered to the first intermediate mixing node for inclusion in the mixing, and wherein the increase in the processing load is a result of the initiating the production of the additional audio stream.

8. The method of claim 6, further comprising:

determining that the combined processing load on the first intermediate mixing node and the second intermediate mixing node each are below a pre-determined threshold; and

assigning, in response to the determining that the combined processing load on the first intermediate mixing node and the second intermediate mixing node are below a pre-determined threshold, all originating nodes that are assigned to the second intermediate mixing node to the first intermediate mixing node.

9. The method of claim 1, wherein the establishing the peer-to-peer distribution tree comprises:

establishing overlay mesh interconnections among the participating nodes;

constructing, at each node within the participating nodes, a respective broadcast tree from the respective node to every other node within the participating nodes;

distributing a respective performance characteristic for each respective broadcast tree to each node within the participating nodes; and

performing, at each node within the participating node based upon the respective performance characteristic, an election protocol to determine an optimal root mixing node to be the root mixing node of the peer-to-peer distribution tree.

10. The method of claim 9, further comprising:

accepting a request from a new processing node to become a participating node, wherein the new processing node has been added to the overlay mesh connections and the request is broadcast to at least a portion of the participating nodes;

adding, in response to the request, the new processing node to the peer-to-peer distribution tree;

transmitting, by a first responder from within the participating nodes, a response to the request; and

assigning the destination of a new audio stream produced by the new processing node to be a parent mixing node of the first responder.

11. A peer-to-peer processing node for participating in multi-party voice-over-IP services, the peer-to-peer processing node comprising:

a distribution tree processor that establishes a peer-to-peer distribution tree to broadcast a composite audio signal from a root mixing node to a plurality of participating nodes, wherein a subset of the participating nodes are originating nodes that provide a respective audio stream output received by the root mixing node to produce the composite audio signal; and

a mixing tree processor that identifies, independently of the distribution tree, a first intermediate mixing node from within the plurality of participating nodes;

a mixer that provides to the root mixing node a mixing of respective audio stream outputs provided by a first subset of originating nodes to the mixer.

12. The peer-to-peer processing node of claim 11, wherein the distribution tree processor further:

establishes overlay mesh interconnections among the participating nodes;

constructs a respective distribution tree from the respective node to every other node within the participating nodes;

distributes a respective performance characteristic for each respective distribution tree to each node within the participating nodes; and

performs an election protocol to determine an optimal root mixing node.

13. The peer-to-peer processing node of claim 11, further comprising:

a responder that transmits a response to a request, received from a remote peer-to-peer processing node, to become a participating node, and

wherein the mixing tree processor further;

accepts a received request from a new processing node to become a participating node, wherein the new processing node has been added to the overlay mesh connections and the request is broadcast to at least a portion of the participating nodes, and

assigns the destination of a new audio stream produced by the new processing node to be a parent mixing node of a remote peer-to-peer processing node that first responds to the request, and

wherein the distribution tree processor adds, in response to the mixing tree processor accepting the request, the new processing node to the peer-to-peer distribution tree.

14. The peer-to-peer processing node of claim 11, further comprising:

a mixing load monitor that monitors a mixing processing load of the mixer and a mixing processing load of at least one neighbor node within the participating nodes, and

wherein the mixing tree processor further;

identifies, in response to an increase in the processing load monitored by the mixing load monitor and independently of the distribution tree, a second intermediate mixing node from the at least one neighbor, and

assigns, in response to the identification of the second intermediate mixing node, a portion of the first subset of originating nodes to the second intermediate mixing node and configures the second intermediate mixing node to provide a mixing of the portion of the first subset of originating nodes to a parent mixing node of the first intermediate mixing node.

15. The peer-to-peer processing node of claim 14, wherein

the mixing load monitor further determines that the combined processing load on the first intermediate mixing node and the second intermediate mixing node are below a pre-determined threshold, and

wherein the mixing tree processor further assigns, in response to the determining by the mixing load monitor that the combined processing load on the first intermediate mixing node and the second intermediate mixing node are below a pre-determined threshold, all originating nodes that are assigned to the second intermediate mixing node to the first intermediate mixing node.

16. The peer-to-peer processing node of claim 11, wherein the mixing tree processor further identifies, independently of the distribution tree, a backup intermediate mixing node from within the plurality of participating nodes, the peer-to-peer processing node further comprising:

a primary mixing monitor that monitors, at the backup intermediate mixing node, a first status of the first intermediate mixing node; and

a child node controller that assigns, in response to the first status, the first subset of originating nodes to the backup intermediate mixing node, thereby causing the backup intermediate mixing node to provide a mixing of respective audio stream outputs provided by the first subset of originating nodes to the root mixing node.

17. The peer-to-peer processing node of claim 16, wherein the primary mixing monitor comprises a heartbeat monitor that monitors the first status by receiving a repetitive heartbeat message from the first mixing node, and

wherein the child node controller assigns in response to a failure to receive the repetitive heartbeat message.

18. A computer program product for providing multi-party voice-over-IP services, the computer program product comprising instructions for:

establishing a peer-to-peer distribution tree to broadcast a composite audio signal from a root mixing node to a plurality of participating nodes, wherein a subset of the participating nodes are originating nodes that provide a respective audio stream output received by the root mixing node to produce the composite audio signal; and

identifying, independently of the peer-to-peer distribution tree, a first intermediate mixing node and a backup intermediate mixing node from within the plurality of participating nodes, the first intermediate mixing node providing to the root mixing node a mixing of respective audio stream outputs provided by a first subset of originating nodes to the first intermediate mixing node.

19. The computer program product of claim 18, further comprising instructions for:

monitoring a processing load on the first intermediate mixing node and on at least one neighbor node within the participating nodes;

identifying, in response to an increase in the processing load on the first intermediate mixing node and independently of the distribution tree, a second intermediate mixing node from the at least one neighbor; and

assigning, in response to the identifying, a portion of the first subset of originating nodes to the second intermediate mixing node and configuring the second intermediate mixing node to provide a mixing of the portion of the first subset of originating nodes to a parent mixing node of the first intermediate mixing node.