Exercise 4.8Z: What does the AWGN Channel Capacity Curve say?

From LNTwww

Channel capacity as a function of  $10 \cdot \lg (E_{\rm B}/{N_0})$

We consider the channel capacity of the AWGN channel as in  Exercise 4.8:

$$C_{\rm Gaussian}( E_{\rm B}/{N_0}) = {1}/{2} \cdot {\rm log}_2 \hspace{0.1cm} ( 1 + \frac { 2 \cdot R \cdot E_{\rm B}}{N_0}) . $$
  • The curve is shown on the right with logarithmic abscissa between  $-2 \ \rm dB$  and  $+6 \ \rm dB$  dargestellt.
  • The addition of  "Gaussian"  indicates that a Gaussian distribution was assumed for this curve at the AWGN input.


Three system variants are indicated by dots in the graph:

  • System $X$:    with  $10 \cdot \lg (E_{\rm B}/{N_0}) = 4 \ \rm dB$  and  $R = 1$,
  • System $Y$:    with  $10 \cdot \lg (E_{\rm B}/{N_0}) = 0 \ \rm dB$  and  $R = 2$,
  • System $Z$:    with  $10 \cdot \lg (E_{\rm B}/{N_0}) = 6 \ \rm dB$  and  $R = 1.5$.


In the questions for this exercise, we still use the following terms:

  • Digital system:   Symbol set size  $M_X = |X|$  beliebig,
  • Binary system:   Symbol set size  $M_X = 2$,
  • Quaternary system:   Symbol set size  $M_X = 4$.



Hints:


Questions

1

What statement does the point  $X$  provide for digital signal transmission?

For  $10 \cdot \lg (E_{\rm B}/{N_0}) = 4 \ \rm dB$  a digital system with rate  $R = 1$  and error probability zero can be imagined.
Such a system does not require channel coding.
Such a system uses an infinitely long code.
A binary system can also meet the requirements.

2

What statement does the point  $Y$  provide for digital signal transmission?

For  $10 \cdot \lg (E_{\rm B}/{N_0}) = 0 \ \rm dB$  a digital system with rate  $R = 2$ and error probability zero can be imagined.
For  $10 \cdot \lg (E_{\rm B}/{N_0}) = 0 \ \rm dB$   ⇒   $R = 0.5$  would be sufficient.
For rate  $R = 2$   ⇒   $10 \cdot \lg (E_{\rm B}/{N_0}) = 5 \ \rm dB$  would be sufficient.

3

What statement does point  $Z$  provide for binary transmission?

A binary system does not meet the requirements in any case.
The curve  $C_\text{Gaussian}(E_{\rm B}/{N_0})$  is not sufficient for this evaluation.

4

Which statement does the point  $Z$  provide for the quaternary transmission?

A quaternary system does not meet the requirements in any case.
The curve  $C_\text{Gaussian}(E_{\rm B}/{N_0})$  is not sufficient for this evaluation.


Solution

(1)  Proposed solutions 1 and 3  are correct:

  • Since the point  $X$  lies to the right of the channel capacity curve  $C_\text{Gaussian}(E_{\rm B}/{N_0})$,  there is (at least) one transmission system of rate  $R = 1$  that provides  "quasi–error–free"  transmission with  $10 \cdot \lg (E_{\rm B}/{N_0}) = 4 \ \rm dB$.
  • Despite the code rate  $R = 1$,  this system includes a channel coding with an infinitely long code,  but unfortunately this code is unknown.
  • However,  a binary system of rate  $R = 1$  does not allow channel coding.



(2)  Only the  proposed solution 2  is correct.  Here the following statements are valid:

  • The required  $E_{\rm B}/{N_0}$  for the rate  $R = 2$  results in
$$(E_{\rm B}/{N_0})_{\rm min} = \frac{2^{2R} - 1} { 2 \cdot R} = \frac{2^4 - 1} { 4 } = 3.75 \hspace{0.3cm}\Rightarrow \hspace{0.3cm} 10\cdot {\rm lg} \hspace{0.1cm}(E_{\rm B}/{N_0})_{\rm min} = 15.74\,{\rm dB} \hspace{0.05cm}. $$
  • The maximum code rate  $R_{\rm max}$  for  $10 \cdot \lg (E_{\rm B}/{N_0}) = 0 \ \rm dB$    ⇒   $E_{\rm B}/{N_0} = 1$  is calculated as follows:
$$C = R = \frac{1}{2} \cdot {\rm log}_2 \hspace{0.1cm} ( 1 + \frac { 2 \cdot R \cdot E_{\rm B}}{N_0}) \hspace{0.3cm}\Rightarrow \hspace{0.3cm} 2^{2R} - 1 \stackrel{!}{=} 2 R \hspace{0.3cm}\Rightarrow \hspace{0.3cm} R_{\rm max} = 0.5 \hspace{0.05cm}. $$
  • Both calculations show that the point  $Y$ with characteristics  $10 \cdot \lg (E_{\rm B}/{N_0}) = 0 \ \rm dB$  and  $R = 1$  does not satisfy the channel coding theorem.



(3)  With a binary system,  the rate  $R = 1.5$  can never be realized   ⇒   proposed solution 1.



(4)  Correct is the  proposed solution 2:

  • The point  $Z$  lies to the right of the boundary curve and for the code rate of a quaternary system holds  $R \le 2$. 
  • So the code rate  $R =1.5$  would be quite realizable with  $M_X = 4$.
  • The proposed solution 1 is wrong.  On the other hand,holds the second solution suggestion is correct:
  • The given curve  $C_\text{Gaussian}(E_{\rm B}/{N_0})$  always assumes a Gaussian distributed input.
  • For a binary system,  a different boundary curve results,  namely  $C_\text{BPSK} ≤ 1 \ \rm bit/channel \ use$.  $C_\text{Gaussian}$  and  $C_\text{BPSK}$  are significantly different.
  • For the quaternary system  $(M_X = 4)$  one would have to calculate and analyze the curve  $C_{M=4}$ .  Again,  $C_{M=4} ≤ C_\text{Gaussian}$ .
  • For small  $E_{\rm B}/{N_0}$:  $C_{M=4} \approx C_\text{Gaussian}$  holds,  after which the curve diverges significantly and ends in a horizontal at  $C_{M=4} = 2 \ \rm bit/channel \ use$.


The point  $Z$    ⇒   $10 \cdot \lg (E_{\rm B}/{N_0}) = 6 \ \rm dB, \ \ R = 1.5$  lies below  $C_{M=4}$.

  • Such a quaternary system would thus be feasible,  as will be shown in  Exercise 4.10 .
  • But only from knowledge of  $C_\text{Gaussian}$  the question cannot be answered  (proposed solution 2).