
dsPIC33F
DS70155C-page 52 Preliminary © 2005 Microchip Technology Inc.
11.7 Soft Modem Library
The Microchip data modem library is composed of
ITU-T compliant algorithms for V.21, V.22, V.22bis,
V.23, V.32 and V.32bis modem recommendations. Bell
standard 103 is also included in this library.
V.21, V.23 and Bell 103 are Frequency Shift Keying
(FSK) modems. V.32, V.32bis and V.22bis are
Quadrature Amplitude Modulation (QAM) modems. V.22
is a Quadrature Phase Shift Keying (QPSK) modem.
V.21, V.22, V.22bis, V.32 and V.32bis are all 2-wire, full-
duplex modems. V.23 is a full-duplex modem when it
operates with a 75 bps backwards channel.
V.22bis includes fallback to V.22, V.23 and V.21
standards. V.32bis optionally falls back to V.22bis, V.22
and V.21 standards.
The dsPIC DSC Soft Modem is well-suited for small
transaction oriented applications, such as, but not
limited to:
•POS Terminals
• Set Top Boxes
•Drop Boxes
• Fire Panels
• Internet Enabled Home Security Systems
• Internet Connected Power, Gas and Water Meters
• Internet Connected Vending Machines
• Smart Appliances
• Industrial Monitoring
Functions supporting ITU-T Recommendation V.42 are
provided with each library. V.42 contains a High-Level
Data Link Control (HDLC) protocol, referred to as Link
Access Procedure for Modems (LAPM) and defines
error correcting protocols for modems.
All data pump modulation and demodulation functions
are written in ASM30 assembly code yielding optimal
code size and execution time. The AT, V.42 and data
pump APIs are written in C30 C Compiler language.
Electronic documentation accompanies the modem
library to help you become familiar with and implement
the library functions. A comprehensive
“dsPIC30F Soft
Modem Library User’s Guide”
describes the required
APIs for the AT, V.42 and data pump layers.
11.8 Speech Recognition Library
The dsPIC Speech Recognition Library provides voice
control of embedded applications that require an
alternative user interface. With a vocabulary of up to
100 words, the Speech Recognition Library allows
users to control their application using spoken
commands. The Speech Recognition Library is an ideal
front end for hands-free products, such as modem
appliances, security panels and cell phones. The
Speech Recognition Library has very modest memory
and processing requirements and is targeted for the
dsPIC30F5011, dsPIC30F5013, dsPIC30F6012 and
dsPIC30F6014 processors.
Key features of the dsPIC DSC Speech Recognition
Library include:
• US English language support
• Speaker independent recognition of isolated
words
• No speaker training is required
• Hidden Markov Modem-based recognition system
• Recognition time < 500 msec
• Master library of 100 common words (listed in the
“dsPIC30F Speech Recognition Library User’s
Guide”
)
• Windows operating system-based utility allows
the user to create a custom word library from the
master library
• Additional words can be added to the master
library (fee based)
• Data tables can be stored in external memory
• Optional keyword activation and silence detection
• Optional system self-test using a predefined
keyword
• Flexible API
• Full compliance with Microchip MPLAB C30 C
Compiler Language Tools
•
“dsPIC30F Speech Recognition Library User’s
Guide”
and
“dsPIC30F Word Library Builder
User’s Guide”
• Designed to run on dsPICDEM™ 1.1 General
Purpose Development Board (DM300014)
11.8.1 RESOURCE REQUIREMENTS
• Sampling Interface: Si-3000 Audio Codec
operating at 12.0 kHz
• System Operating Frequency: 12.288, 18.432 or
24.576 MHz
• Computational Power: 8 MIPS
• Program Flash Memory: 18 KB + 1.5 KB for each
library word
• RAM: < 3.0 KB
Comentários a estes Manuais