next up previous
Next: Maximum Likelihood Sequence Estimation Up: Collected Problems Previous: Modulation

Multi-Path Channel and Tapped Delay-Line Model

A mobile communication system employs binary phase shift keying with rectangular pulses

\begin{displaymath}
u(t) =
\left\{
\begin{array}{cl}
1 & \mbox{for $0\leq t \leq T$}\\
0 & \mbox{else}
\end{array}\right.
\end{displaymath}

to transmit information at a rate of one bit per $T$ seconds over a channel with (time-invariant) baseband-equivalent impulse response


Furthermore, the channel adds white Gaussian noise $N_t$ of spectral height $\frac{N_0}{2}$. The receiver is designed under the assumption that $c(t)=\delta(t)$ and hence is simply a filter matched to the transmitted pulses $u(t)$. In summary, the communication system can be described by the following block diagram.

\begin{figure*}\begin{center}
\par\leavevmode
\setlength{\epsfxsize}{5.5in} \epsfbox{/home/pparis/courses/ece732/P2.eps}\end{center}\end{figure*}

The objective of this problem is to determine the parameters of the equivalent tapped delay-line model between the input sequence $I_n$ and the output sequence $R_n$. A generic block diagram for a tapped delay-line model is given below.

\begin{figure*}\begin{center}
\par\leavevmode
\setlength{\epsfysize}{3in} \epsfbox{/home/pparis/courses/ece732/P2a.eps} \end{center}\end{figure*}

  1. Compute the convolution of the pulse shaping filter $u(t)$ with the channel impulse response $c(t)$.
  2. How many memory elements $L$ are there in the tapped delay-line?
  3. What are values for the coefficients in the tapped delay-line model?
  4. Find the distribution (type, mean, variance, etc.) of the additive noise $N_n$ samples in the tapped delay line model.


next up previous
Next: Maximum Likelihood Sequence Estimation Up: Collected Problems Previous: Modulation
Dr. Bernd-Peter Paris
2003-12-08