0.8 Stuff happens  (Page 8/14)

Another idealized assumption made in idsys.m is that the receiver knows the start of each frame; that is, it knows where each four-symbol group begins. This is a kind of“frame synchronization” problem and was absorbed into the specification ofa parameter $l$ which appears in the code as 0.5*fl+M . With the default settings, $l$ is 125. This problem poses the question. “What if this isnot known, and how can it be fixed?”

1. Verify, using idsys.m , that the message becomes scrambled if the receiver is mistakenabout the start of each group of four. Add a random number of 4-PAM symbols before the messagesequence, but do not “tell” the receiver that you have done so (i.e., do not change $l$ ). What value of $l$ would fix the problem? Can $l$ really be known before hand?
2. [link] proposed the insertion of a marker sequence as way to synchronize the frame.Add a seven-symbol marker sequence just prior to the first character of the text.In the receiver, implement a correlator that searches for the known marker.Demonstrate the success of this modification by adding random symbols at the start of the transmission.Where in the receiver have you chosen to put the correlation procedure? Why?
3. One quirk of the system (observed in the eye diagram in [link] ) is that each group of four begins with a negative number. Use this feature (rather than aseparate marker sequence) to create a correlator in the receiver that can be used to find the start of the frames.
4. The previous two exercises showed two possible solutions to the frame synchronization problem. Explain the prosand cons of each method, and argue which is a “better” solution.