EE535 - PROJECT 2
Ethernet Collision Simulation Applet


Jorge Guajardo, guajardo@ece
Pedro Soria-Rodríguez, sorrodp@ece



  1. Introduction.
  2. The detection of packet collisions on a Ethernet cable can be determined by measuring the local average of the data pulses on the wire permanently and comparing it against a fixed threshold to find whether a collision occured or not. In practice the average is found using a low pass filter. However, we will use a close approximation to measuring the average by calculating the arithmetic average of the pulse amplitudes.

    One of the factors that need to be accounted for when making the collision decision is the attenuation that signals undergo in the Ethernet cable. In general, this attenuation is a function of the frequency but this study will be limited to the case of a distortionless cable, in which the atenuation is constant and the phase shift is linear for all frequencies. Under this assumption, the only transmission line effect to consider is the line attenuation.

    This project presents a simulation of the collision detection algorithm. The line parameters can be adjusted to simulate the waveforms along a transmission line when two stations are transmitting at the same time, and determine whether a collision occured or not.

  3. Sinusoidal Steady-state Analysis of Transmission Lines
  4. Let be the sinusoidal source that excites the transmission line as shown in Figure 1.


    Figure 1.

    From the Telegrapher's equations, the voltage and current at any point along the line are given by

    For a sinusoidal excitation, v(z,t) and i(z,t) may be expressed as

    Substituting equations (3) and (4) into equations (1) and (2), and canceling out the terms from both sides of the equations, one obtains

    Here we have assumed a source of time variation of the form because it simplifies the equations by getting rid of the derivative with respect to time.

    Differentiating equation (5), once more with respect to z, and substituting equation (6) into the result, we obtain

    Similarly, one obtains

    Equations (7) and (8) are second order differential equations which have solutions

      

    where and the amplitude of the forward and backward voltage and current waves are related as follows:

    Finally, it is important to point out that in the case of a distortionless transmission line, and where the attenuation factor and the velocity factor.

    It is important to point out that in this project we deal with transmission lines of characteristic impedance that are terminated with resistors with the same impedance. In our case, we assume ohms. This implies that we don't have any reflections in the line. Therefore, Equations (9) and (10), become for the case of a forward traveling wave:

      

    and for a backward traveling wave:

      



  5. Application to a Square Wave
  6. It is known from Fourier analysis, that we can represent a periodic wave as a sum of complex exponentials, as indicated in Equation (17)

     

    where the coefficients are complex in general and are given by:

     

    Now consider, sending one sinusoidal of frequency down the line. Then Equations (13), (14), (15), and (16) still are applicable and the only thing that changes is which becomes . It is easy to see that the same thing will hold true for . Thus, we obtain that for a periodic wave, the previous equations become:

    Notice that , where is entirely independent of the frequency (under our assumptions). Therefore, we can factor out or and have it multiply the expression inside of the summation sign. Then, since we assumed a distortionless line (one in which is linear with ) we have the same pulse train that we had in the beginning attenuated over the length of the line.


Fri Feb 21 14:59:25 EST 1997

Back to the Ethernet Collision Simulation Applet
Project Homepage. HW2 - Spring 97.

Back to the WPI EE535
Telecommunications and Transmission Technologies
Course Homepage