With the advent of smart phones and innovative applications, recent years have seen a tremendous upsurge of the data demands from users worldwide. Operators around the world, in the current day, are facing multiple challenges in terms of maintaining the profitability, providing superior user satisfaction, deploying innovative services etc. Though data usage trends are on inclining trend, voice revenues still constitute significant portion of mobile operator's revenues, according to a study. Given this scenario, any new technology or feature that differentiates the voice service is certainly a welcome from the operator's point of view and would be a chance to ship more devices from the device vendor's perspective.
High Definition Voice (HD Voice) is one such technology which provides superior voice quality to end users and would certainly be a differentiating service for a mobile operator. First commercial deployment of HD Voice service is carried out by Orange in its network in Moldova in September 2009. Since then the deployments of HD Voice service has seen a steady growth and, today, there exist as many as 36 mobile networks with HD Voice capability.
When initial HD Voice capable networks and devices started to appear, different operators and vendors used different logos to market the HD capability. Operator community joined together to solve this problem and a common logo was proposed as shown in the picture. From the end users perspective, HD Voice provides superior voice quality which means the voice sounds more natural and it would be easy to recognize the caller. The speech would be easier to comprehend even in noisy surroundings.
The technology behind the HD Voice in mobile networks is called Adaptive Multi Rate – Wide Band (AMR – WB). AMR WB is not new and has been in 3GPP standards for a while. AMR WB is a speech coding technology that preserves more frequency components of the original speech in the encoded speech signal. Original AMR has a maximum encoded bit rate of 12.2kbps and it is possible to run the encoder in different other speech coding output rates depending on the quality of the air interface (for example the speech coder can be configured to output a bit rate of 12.2kbps under best radio conditions and when the radio conditions deteriorate it is possible to turn to lower speech coding bit rates). The original analog speech signal picked up by the microphone is sampled at 8 KHz (according to Nyquist criteria, the sampling frequency needs to be at least twice the highest frequency of the signal undergoing sampling. In the original AMR technology the frequency range of the input signal component is 300 Hz – 3400 Hz). In the WB-AMR technology, the frequency component (range) that is transmitted is increased from (300 Hz-3000 Hz) to (100 Hz-7000 Hz) and hence the sampling frequency increases to 16 KHz from 8 KHz. Apart from this, the speech coder is modified accordingly and outputs a maximum data rate of 12.65 kbps. The advantages are: increased voice quality while preserving the voice capacity (ie, no additional air interface capacity is needed to Support AMR-WB). The following picture depicts, as an example, changes needed in the UMTS mobile network in order to support HD Voice (AMR-WB). Note that even though the picture depicts UMTS example, AMR-WB can also be supported in the GSM network.
Handset Support for HD Voice
As indicated, handsets should support AMR-WB speech CoDeC, AMR-WB bearer establishment, must manage transitions between different AMR-WB rates and between AMR-WB and AMR-NB rates (would be useful when the destination network is not supporting the AMR-WB data delivery).
Access Network Support for HD Voice
On the access network side, in the UMTS case, the node B will usually be transparent to the AMR-WB changes; however there will be some revalidation of the Node B software needed in order to verify the support of AMR-WB data rates. RNC shall be capable to admit the AMR-WB Radio Access Bearer (RAB) and its combinations with PS bearers (in case a Packet Switch call is initiated in parallel or during the ongoing AMR-WB speech call) and shall be able to manage transitions among AMR-WB (radio quality adaptation)
Core Network (CN) Support for HD Voice
In the Core Network which is based on the 64 kbps circuit switch links, the speech data (AMR or any other speech encoded data) is trans-coded in to 64 kbps PCM data. The 64 kbps PCM data is again converted back to original speech encoded data rate in the destination network before the data is finally decoded in the mobile phone using appropriate speech decoder. This kind of trans-coding on to 64 kbps introduces additional complexity to network nodes and also contributes to additional end to end speech delay. With the introduction of Bearer Independent Core Network (BCIN), links between RNC and CN and also the CN back bone are based on the ATM or IP links. This makes it possible to transport the speech data without additional trans-coding performed at CN. This further leads to the possibility to deactivate the transcoders when BCIN functionality is available. However, it is important to note that this is applicable to the case in which the entire call travels only through UMTS network. This is called Transcoding Free Operation (TrFO). In case when either of the users is in GSM network, transcoding is mandatory since the interface between BSC and the CN (A interface) is always PCM based. So, in case of a GSM network, transcoders automatically come in to picture due to the inherent PCM link presence. However, these transcoders exchange in-band signalling, once call is established, and can eliminate the tandom operation by tunnelling the AMR-WB data in the PCM links (this wastes some BW but the speech quality is preserved). This sort of operation is called Tandom Free Operation (TFO). If the call is landing on a PSTN network then one transcoding in to 64 kbps is needed at the media gateway which connects the core network to PSTN network.
No comments:
Post a Comment