Echo canceller for voip networks
US-2015043361-A1 · Feb 12, 2015 · US
US10439673B2 · US · B2
| Field | Value |
|---|---|
| Publication number | US-10439673-B2 |
| Application number | US-201715837118-A |
| Country | US |
| Kind code | B2 |
| Filing date | Dec 11, 2017 |
| Priority date | Dec 11, 2017 |
| Publication date | Oct 8, 2019 |
| Grant date | Oct 8, 2019 |
A practical reading order for non-experts. Skip the full description unless you need deep technical detail.
What the patent document calls the invention.
A short plain-language summary of the technical disclosure.
Who owns or filed the patent and who is credited as inventor.
Filing, priority, publication, and grant dates set the timeline.
The legal scope of protection — read this for what is actually claimed.
Technology tags used to group this patent with similar filings.
Prior art links and similar publications in this corpus.
Official abstract text for this publication.
A cloud based echo canceller is set forth for recreating an estimate of a lost packet or data at a server without requiring redundant data over the network or freezing operation of the echo canceller. In an exemplary embodiment, the echo cancelling function is not located in a single device, but is shared between the end-point and a cloud service, where the function of the end-point is to provide a time synchronized copy of the signal from the end-point loudspeaker and the signal received by the end-point microphone. Consequently, the high CPU intensive operations can be offloaded to a server such as a cloud server. In addition, several users can share the echo canceller, thereby reducing the cost of the overall function. According to an additional aspect, a further synchronization block is provided, in the form of a packet estimator, to compensate for packet or data loss in the send direction.
Opening claim text (preview).
What is claimed is: 1. A cloud based echo canceller, comprising: a system for receiving a signal (Rin) and in response applying a signal (Ro) to a network, and for receiving a signal (Si+Ro′) from said network, splitting said signal (Si+Ro′) into signals (Si) and (Ro′), the signal (Rd) being a modified version of the signal (Ro) and the signal (Si) containing an echo of the signal (Ro′), and canceling echoes from said signal (Si) using the signal (Ro′); at least one device that is remote from the system for (i) receiving and outputting said signal (Ro′), where said signal (Ro′) is a corrupted version of the signal (Ro) resulting from transmission over the network, (ii) receiving the signal (Si) containing an echo of said signal (Ro′), and (iii) combining and transmitting said signals (Si) and (Ro′) over the network to said system for echo cancellation of said signal (Si) using said signal (Ro′); and wherein said system further includes a packet estimator for creating an estimate (Ro″) of the signal (Ro′) from the signal (Rin) in the event a packet of the signal (Si+Ro′) is lost, and replacing the signal (Ro′) in step (iii) with the estimate (Ro″) for echo cancellation of said signal (Si) using said estimate (Ro″). 2. The cloud based echo canceller of claim 1 , wherein the system further comprises a splitter, a subtractor, and an echo estimator and control. 3. The cloud based echo canceller of claim 2 , wherein said at least one device is a handsfree device. 4. A method of distributed echo cancelling, comprising: receiving a signal (Rin) at a system and in response transmitting a first reference signal (Ro) from the system to a network; receiving a second reference signal (Ro′) by at least one device that is remote from said system over said network, wherein the second reference signal comprises the first reference signal as modified by network effects due to transmission over the network; receiving within said at least one device a further signal (Si) containing an echo of said second reference signal (Ro′); combining said further signal and said second reference signal at said at least one device; transmitting said combined signal (Si+Ro′) from said at least one device to said network; receiving within said system the combined signal (Si+Ro′); splitting said second reference signal (Ro′) from said further signal; canceling the echoes from said further signal (Si) using the second reference signal (Ro′); and creating an estimate of the second reference signal (Ro″) in the event a packet of the second reference signal (Ro′) is lost from the combined signal (Si+Ro′), and replacing the second reference signal (Ro′) with the estimate (Ro″) for canceling the echoes from said further signal (Si). 5. The method of claim 4 , wherein creating the estimate of the second reference signal (Ro″) further comprises: (a) performing a correlation of the second reference signal (Ro′) with the signal (Rin); (b1) if the second reference signal (Ro′) and the signal (Rin) are correlated then using a relative shift offset of the second reference signal (Ro′) to the signal (Rin) to read the estimate (Ro″) from a buffer for the signal (Rin), inserting the estimate (Ro″) into an echo canceller history buffer, and canceling the echoes from the further signal (Si) using the second using the estimate (Ro″); or (b2) if the second reference signal (Ro′) and the signal (Rin) are not correlated then ceasing canceling of the echoes from the further signal (Si) and invoking a packet loss compensation algorithm until effect of the lost reference signal packet Ro′(n) is flushed from the echo canceller history buffer.
using echo cancellers (echo cancellers per se H04B3/23) · CPC title
the noise being echo, reverberation of the speech · CPC title
using two adaptive filters, e.g. for near end and for end echo cancelling · CPC title
Processing in the frequency domain · CPC title
Related publications grouped by family.
Answers are generated from the same data shown on this page.