An Open-Source Speech Codec at 450 bit/s with Pseudo-Wideband Mode

被引：0

作者：

Erhardt, Stefan ^{[1
]}

Kurin, Thomas ^{[1
]}

Lurz, Fabian ^{[1
]}

Weigel, Robert ^{[1
]}

Koelpin, Alexander ^{[2
]}

机构：

[1] Friedrich Alexander Univ Erlangen Nuremberg, Inst Elect Engn, Erlangen, Germany

[2] Brandenburg Tech Univ Cottbus, Chair Elect & Sensor Syst, Cottbus, Germany

来源：

2019 49TH EUROPEAN MICROWAVE CONFERENCE (EUMC) | 2019年

关键词：

Speech coding; vocoders; open source software; radio transceivers;

D O I：

10.23919/eumc.2019.8910691

中图分类号：

TM [电工技术]; TN [电子技术、通信技术];

学科分类号：

0808 ; 0809 ;

摘要：

Most modern digital speech codecs are proprietary and do not give insights into their implementations. "Codec 2" is an open-source codec and opens up the possibility of application-tailored adaptations. In this publication a new 450 bit/s mode is introduced. It includes a pseudo-wideband mode that can be optionally enabled at the receiver without resulting in a higher bit rate. By sorting the codebook vectors a bit error rate of 10(-1) is still intelligible. An encoded audio transmission was tested with two low-power Sub-GHz transceivers in the 70 cm amateur radio band achieving a sensitivity of -130 dBm. For proofing the concept, a field test was conducted and compared to a propagation simulation.

引用

页码：1048 / 1051

页数：4