This is a very basic question ,but basics important, : How is a stream of data bits (zeros and ones that express a certain information) and at a certain rate being generated? what is involved in terms of processing and possibly hardware? a clock? a quantizer? a sampler? or what exactly? I really would like to get a brief idea

  • $\begingroup$ UART. universal asynchronous receiver transceiver. $\endgroup$ – user28715 Oct 9 '18 at 18:13
  • $\begingroup$ Can you inform us in which context you're asking this? Basically, the central theorem on which all information theory is based, is that you can represent any finite amount of information as bits. So, anything that yields a constant rate of information is a fixed-rate data source of bits. That can be something as simple as a coin flipped every 10s, or a camera producing uncompressed video, or a hardware random number generator, or a geiger counter, or … $\endgroup$ – Marcus Müller Oct 9 '18 at 18:48
  • $\begingroup$ I basically want to know what is involved, so take for example an uncompressed video, what is involved (very briefly) in converting the video information into bits that are generated at a certain rate? $\endgroup$ – Hatem Tawfik Oct 9 '18 at 19:17

Multimedia signals such as video,picture, music, speech have bandwidths associated with their baseband analog source nature.

For digital processing this bandlimited analog source is passed through a analog to digital converter (into a sequence of samples), which operates at a clock-frequency determined by the signal sampling rate, $F_s$, which is also limited by the source bandwidth due to Nyquist-Shannon sampling theorem.

Furthermore for each sample, a number of bits are used to quantize (numerically resolve) it. The number of bits is dictated by the signal amplitude accuracy requirement. For digital video, it's mostly 8-bits per RGB channel. Or for audio it can be 16-bits per stereo L/R channel.

Hence, the multiplication of samlping rate $F_s$ and bit-per-sample gives you the raw source bit rate $R_b$. You may prefer to apply some statistical or perceptual data compression to reduce $R_b$ to a minimum.

Then you choose a digital modulation scheme such as FSK,PSK,GMSK,OFDM... and transmit those bits through your favorite medium. Before doing this, you may increse $R_b$ (by adding redundancies) for channel transmission error correction purposes. The result is your bit-stream (not necessarily logic signal levels of 0s and 1s though) that's traveling accross the universe...

I hope I clarified it.

| improve this answer | |

Not the answer you're looking for? Browse other questions tagged or ask your own question.