AudioCodes Media Server (IPmedia) is a VoP infrastructure unit providing value-added media processing services to Next Generation IP/ATM networks. It can be applied to a wide range of applications and services, including:
- Conference servers
- Announcements servers
- IVRs
- Prepaid and postpaid calling cards
- Call centers
- Voicemail and unified messaging
- IP Centrex/Hosted PBX
- Wireless 2G VAD services
- Wireless 3GPP – IMS Fixed Mobile Convergence server (MRF)
- Transcoding (including mandatory wireless coders support, AMR, EVRC, etc.)
- Tapping/replication – legal interception, push-to-talk, music-on-hold
- Recording servers
- Speech-enabled services – command & control, speech portals and voice dialing
- Video – announcements, transcoding, IVR/info portals, video-mail, streaming and conferencing (Roadmap).
The IPmedia Media Server is controlled by a control agent such as a softswitch or application server. The IPmedia communicates with these agents via standard control protocols such as MEGACO (H.248), MGCP and SIP.
Following is an example for a Media Server application in an IP Centrex application:

AudioCodes IPmedia product line consists of the following media server platforms:
- IPmedia 2000 – 60-240 channels, 1U ,entry level
- IPmedia 3000 – 480-2016 channels, 2U small-sized , high availability , carrier-grade system
- IPmedia 5000 – 2,000-8000 channels, 5U medium-sized, HA, carrier-grade system, planned for 2Q2005
- IPmedia 8000 – 2,000-18,000 channels, 12U large sized, HA, carrier-grade system, planned for 2nd half of 2005
Each of the above systems provides the following functionality:
Voice Processing:
- Packet streaming per RTP/RTCP (RFC 1889/1890).
- Voice Compression:
- VoIP: G.711, G.726, G.727, G.729A, G.723.1, G.728 , NetCoder®
- GSM/UMTS: GSM-FR, GSM-EFR, AMR (all 8 rates),
- CDMA: EVRC, QCELP 8k, QCELP 13k
- Echo Cancellation: G.168-2000 compliant with tail of 30/64/128 msec.
- Voice & Tone generation and signaling discrimination
- VAD, CNG and PLC for all coders
- IPmedia Detectors: Pattern Detector, Answer Detector (AD), Answering Machine Detection, AGC and Energy Detector
- Voice, Data and Fax Discrimination
- Fax Relay per T.38 (Group 3 real time fax relay up to 14.4 Kbps with auto fallback; tolerant to delays up to 4 seconds)
- RTP Multicast for Legal Intercept
- RTP<->RTP Transcoding
- RTP Redundancy support (RFC 2198)
- Field proven dynamic jitter buffer
- Sophisticated packet interpolation
- VoATM (3000/5000/8000 platforms)
- AAL1 (PVC, SVC)
- AAL2 (PVC, SVC)
IVR
- DTMF detection and generation (either inband or using RFC 2833)
- Prompts playback locally or from remote repository (via HTTP streaming)
- Play-Collect DTMF advanced audio packaged (play sequences of messages, collect digits, screening, repeat announcements, variables, handle errors and loop indefinitely)
- Barge-in (DTMF and/or Speech detected)
- Complex “Menu” logic: VXML
- Field proven solutions with 3rd party 3rd party Application Server partnerships
- SIP, MEGACO, MGCP or C++ library interfaces to control the announcement server features
Announcements
- Announcements stored locally and/or on an external HTTP server
- Abstract Audio:
- Language abstraction: Support for multiple languages (23 languages)
- Support variables: date, currencies, numbers, etc.
- Allow user defined abstractions (e.g., by gender, time-of-day, day-of-week, customer, etc.)
- Aggregated functions to play sequences of messages, collect digits, repeat announcements , variables, handle errors and loop indefinitely
- Audio Package widely accepted (IETF, ITU, CableLabs, ISC)
- “Advanced Audio Package” H.248.9 package
- PacketCable’s ‘Basic Audio Package’
- MGCP RFC available: RFC 2897
- Advance Audio Announcements (NetAnn, MSCML, MOML) - SIP support
- Tools to convert existing announcement libraries to WAV format
- SIP, MEGACO/H.248, MGCP or AudioCodes Networked API (TPNCP)
- Powerful Audio Provisioning Server (APS) – provides users the capability to add, delete, and replace audio to the media servers in the network.
Conferencing
- Mixing and automatic transcoding between codec types
- Loudest N and preferred speakers
- Control: Moderator, Active, Passive listener, whisper mode(coaching), mute, “Bip” control (on entry/exit)
- Independent gain control on all ports (Volume up/down)
- Automatic gain control
- DTMF detection and muting
- Dynamic size configuration
- SIP, MEGACO, MGCP or TPNCP/API


