System and method for dynamically changing error algorithm redundancy levels
First Claim
1. A system architecture for an internet telephone gateway server, comprising:
- hardware for interfacing with the internet and a public switched telephone network; and
software for connecting telephone calls between transmitters and receivers, said software having the capability of dynamically changing a level of redundancy of a forward error correction algorithm from packet-to-packet in a data stream so as to accommodate data dropouts,whereby aural data in a packet is entirely duplicated to maintain the voice quality present prior to the data dropout.
3 Assignments
0 Petitions
Accused Products
Abstract
The invention is concerned with improvements in full duplex Internet telephone systems with a system architecture having low latency and permitting voice communication with telephone to telephone or PC to telephone connections. The architecture permits dynamic packet-to-packet change in codec to adjust for Internet conditions. The voice port creates self-describing packet conditions so that the higher-level software of the system is independent of codec selection. In addition to adjusting the codec, the voice port has the capability of dynamically and concurrently selecting other factors such as the level of error correction redundancy, the packet size and packet bundling on a packet-to-packet basis. The invention further includes a technique to eliminate dead air spaces in the voice data transmission stream by speeding up or slowing down the data rate in the buffer while maintaining a constant pitch of speech.
40 Citations
5 Claims
-
1. A system architecture for an internet telephone gateway server, comprising:
-
hardware for interfacing with the internet and a public switched telephone network; and software for connecting telephone calls between transmitters and receivers, said software having the capability of dynamically changing a level of redundancy of a forward error correction algorithm from packet-to-packet in a data stream so as to accommodate data dropouts, whereby aural data in a packet is entirely duplicated to maintain the voice quality present prior to the data dropout. - View Dependent Claims (2, 3, 4, 5)
-
Specification