## Statistical Multiplexing For Digital Headend System

As we all knows first we shifted from Analog Headend System to Digital Headend System to increase the number of channels.In analog headend we could run only 1 channel per frequency bandwidth where as now in digital headend system we can run 8 channels easily per frequency bandwidth.

But now in digital headend system where we can run 8 channels easily per frequency we need to increase more number of channels in the same frequency bandwidth that is why the term came Statistical Multiplexing. *Statistical Multiplexing can increase 30% more digital channels in same bandwidth.*

## Statistical Multiplexing Working

Basic meaning of statistical multiplexing is that it removes all the additional information available in the video. In other words it remove frame which are repeating or having same information.Thus it delivers different frames only. In news channel, if a news reader reading the news then it is containing less information because background is same so it needs less bandwidth space.

On the other hand if there is visual of outside news room in public then the information is changing rapidly.It contain more information that is why it needs more bandwidth. In statistical multiplexing there is a pool which contains number of channels.We gives particular bandwidth to pool only and then statistical multiplexing decides how much bandwidth needs to allocate to a particular service that is available in pool. Statistical multiplexer monitor the channel 50 times in one second to check bandwidth requirement.

Statistical multiplexing is having two main factor that is “Rate Sensing” and “Control Algorithm” where control algorithm is a mathematical equation.Thus these two factors works simultaneously for statical multiplexing. The static multiplexing system block diagram is shown below, As mentioned above in diagram this is static multiplexing system.In this system there are three encoders,three buffers,one multiplexer,one channel buffer and one joint remote control. Encoders are used to encode the signal then output goes into the buffer and then into the multiplxer.

These different stream produced by the encoders are multiplexed by the multiplexer as MPTS (Multi program transport stream). In this system encoders are giving two output one is bitstream and second one is statistics. When the video is encoded it generated Bitstream as well as the statistics these statistics goes into the “Joint Remote Control.

The information that Joint Rate Controller receives it decides what data rate should be for this particular channel.This controlled signal goes in the encoder to change the data rate for this particular channel.

Thus this channel is encoded accordingly data rate sent by JRC. While remaining data rates are allocated to other channels. So whenever encoder receive the rapidly changing information or high data rate signal from the source then it requests to increase the data rate accordingly.

So if there is free data rate space available then it is allocated to this encoder. If the data rate is not available from all the channels then output comes with poor quality.

If more data is available then buffers temporarily store it.This data is delayed for few mili seconds in buffer. The mathematical algorithms used for MPEG 2 and MPEG 4 system are little bit different. This was all about the statistical multiplexing for digital headend system.