Advanced multi-band excitation
Advanced Multi-Band Excitation ( AMBE ) is a proprietary and patented algorithm for parametric audio coding , also known as a vocoder , for very low bit rates of 2 kBit / s to 9.6 kBit / s. The name is a registered trademark of Digital Voice Systems .
features
Its performance exceeds that of MELP at the same data rate (although AMBE aims at a slightly higher data rate range). At the same time, AMBE has a low level of complexity - less than that of the CELP process.
The audio data can be combined with forward error correction data of up to 7,200 bit / s. Block losses that occur anyway can be masked by using the parameters of the previous block to fill the gap.
The draft of the AMBE Codebook provides for certain (interference) noises of non-human origin such as mains hum , wind noise and the like to be radically rejected. The implementation in DVSI's AMBE 2020 chip offers speech pause detection and comfort noise .
Since it manages with such low data rates, it can be used to communicate over narrow-band channels like single-sideband (SSB) -modulated voice signals , even though the signals are digital .
technology
AMBE works with codebooks at a sampling rate of 8 kHz in blocks of 20 ms.
history
DVSI improved the Massachusetts Institute of Technology (MIT) 's 1980 Multi-Band Excitation (MBE) technology , which is used in many communication systems around the world. The Improved Multi-Band Excitation (IMBE) voice codec was the resulting product. AMBE represents the further improved successor.
Usage examples
It is used in digital satellite telephone systems (e.g. Inmarsat , Iridium , ACeS and Optus systems), on certain (voice) channels in XM Satellite Radio , the D-STAR transmission standard and in the amateur radio protocol by Charles Brain ( callsign G4GUO ) on the shortwave used and is the voice codec at OpenSky - trunked radio systems .
In Digital Mobile Radio (DMR), AMBE is an option for voice data. At APCO P25 , AMBE is introduced with phase 2 devices.
literature
- Peter Vary, Ulrich Heute, Wolfgang Hess: Digital speech signal processing. Teubner, Stuttgart 1998, ISBN 3-519-06165-1 , p. 585.