Bit Stuffing

Bit Stuffing o    For Example, device A uses one time slot, while the faster device B uses two slots o    The Time slot length is FIXED o    Therefore data rates should be integer multiples of each other o    For example, we can accommodate a device that is 5 times faster than the other device by giving it five slots to one for each of the other devices o    We  cannot  accommodate  a  device  which  is  five  and  a  half  times  faster  using  this method because we cannot introduce half a time slot into a frame o    When the speeds are not integer multiples of each other, they can be made to behave as if they were o    This is done by a technique known as BIT STUFFING o    In bit stuffing, MUX adds extra bits to a device o    For Example, if we have one device with a bit rate of 2.75 times that of other devices, we can add enough bits to raise this rate to 3 times that of others o    The extra bits are then discarded by the Demultiplexer Asynchronous TDM o    Synchronous TDM does not guarantee full utilization of the timeslots o    Because the time slots are fixed and pre assigned, whenever a connected device is not transmitting,  the  corresponding  slot  is  empty  and  much  of  the  channel  capacity  is wasted o    For  Example,  imagine  that  we  have  multiplexed  the  o/p  of  20  identical  computers onto a single line o    Using synchronous TDM, the speed of that line must be at least 20 times the speed of each i/p line o    But what if only 10 computers are in use at a time? o    Half of the capacity of the line is wasted o    Asynchronous TDM or Statistical TDM is designed to avoid this type of waste o    Asynchronous means flexible or Not fixed o    In  an  asynchronous  system,  if  we  have  'n'  input  lines,  the  frame  contains  no  more than 'm' slots, where m is less than n o    In this way asynchronous TDM supports the same number of I/p lines as synchronous TDM with a lower capacity link o    A slot is available to any device that wants to send data o    MUX  scans  I/p  lines,  accepts  data  until  a  frame  is  filled  and  then  sends  the  frame across the link ™   Advantages of Asynchronous TDM ŠTwo major advantages: -Ability to allocate time slots dynamically -Lower ration of time slots to I/p lines Above two factors greatly reduce the likelihood of a waste

o    Fig. shows a system with 5 I/p lines sharing a link using Asynchronous TDM o    Frame size is 3 slots per frame o    Fig shows how MUX handles 3 levels of traffic o    In the first case, only 3 of the 5 computers have data to send o    In the second case, 4 lines are sending data o    In the third case, all devices are sending data o    In  each  case,  MUX  scans  the  devices  in  order  from  1  to  5  filling  time  slots  as  it encounters data to be sent    Asynchronous TDM Figure 1 o    In the first case, the 3 active i/p lines correspond to the 3 slots in each frame o    For the first 4 frames, the I/p is symmetrically distributed among all the devices. th o    By the 5    frame however, devices 3 and 5 have completed their transmission but device 1 still has two characters to go    Asynchronous TDM Figure 2 o    The MUX picks up the A from device 1, scans down the line without finding another transmission and returns to device 1 to pick up the last A o    There being no data to fill the final slot, the MUX then fills the 5    frame with only 2 slots filled o    Compare with Synchronous TX: 6 frames of 5 slots each would be required=30 slots, 14 slots used only o    In second case, there is one more I/p line than there are slots in each frame o    This  time  MUX  scans  from  1  to  5  and  fills  up  a  frame  before  each  of  the  lines  are checked o    The first frame contains data from device 1, 3,and 4 th    Asynchronous TDM Figure 3 o    MUX continues the scan and puts first portion of 5    device into the first slot of next frame and so on o    When the number of active senders does not equal the number of slots in a frame, the time slots are not filled symmetrically o    Device 1 occupies the first slot in the first frame , 2    slot in second frame and so on o    In the third case, frames are filled as shown above o    All 5 I/p lines are active o    In this case device 1 occupies the 1   slot in the first frame, the 3    slot in the second frame and so on Aspects of Asynchronous TDM    Addressing and Overhead o    Case 2 & 3 above show a major weakness of Asynchronous TDM o    How does the DEMUX know which slot belongs to which output line? o    As opposed to Synchronous TDM, in this case, data from a given device might be in the first slot of one frame and in the third of the next o    Therefore, each time slot must carry an address telling the DEMUX how to direct data o    This address is for local use only attached by the MUX and detached by the DEMUX o    In the figure above address is specified by a digit o    Adding  address  bits  to  each  time  slot  increase  the  overhead  of  an  Asynchronous system and limits its efficiency o    Addresses usually consist of only a small number of bits o    Need   for   Addressing   makes   Asynchronous   TDM   inefficient   for   bit   or   byte interleaving o    Imagine bit interleaving with each bit carrying an address o    One bit of data plus 3 bits of address o    Asynchronous TDM is efficient only when the size of the time slot is kept relatively large Inverse Multiplexing o    Opposite of Multiplexing o    Takes data from one high speed line and breaks it into portions that can be sent over several lower speed lines simultaneously Why do we need Inverse Multiplexing? o    An organization wants to send data, voice and video each of which requires a different data rate o    To send voice it needs 64Kbps, o    To send data, it needs 128 Kbps link o    To send video it may need 1.544 Mbps link o    It can lease a 1.544 Mbps line from a common carrier and only use it fully for sometime o    Or it can lease several separate channels of lower data rates o    Voice can be sent over any of these channels o    Data & Video can be broken into smaller portions using Inverse Multiplexing and TX