Exercise 4.3: Iterative Decoding at the BSC

From LNTwww

BSC model and  (above)
possible received values

We consider in this exercise two codes:

$$\underline{x} = \big (\hspace{0.05cm}(0, 0, 0), \hspace{0.1cm} (0, 1, 1), \hspace{0.1cm} (1, 0, 1), \hspace{0.1cm} (1, 1, 0) \hspace{0.05cm} \big ) \hspace{0.05cm}, $$
$$\underline{x} = \big (\hspace{0.05cm}(0, 0, 0), \hspace{0.1cm} (1, 1, 1) \hspace{0.05cm} \big ) \hspace{0.05cm}.$$

The channel is described at bit level by the  $\text{BSC model}$.  According to the graphic,  the following applies:

$${\rm Pr}(y_i \ne x_i) =\varepsilon = 0.269\hspace{0.05cm},$$
$${\rm Pr}(y_i = x_i) =1-\varepsilon = 0.731\hspace{0.05cm}.$$

Here,  $\varepsilon$  denotes the falsification probability of the BSC model.

Except for the last subtask,  the following received value is always assumed:

$$\underline{y} = (0, 1, 0) =\underline{y}_2 \hspace{0.05cm}. $$

The chosen indexing of all possible received vectors can be taken from the graphic.

  • The most considered vector  $\underline{y}_2$  is highlighted in red.
  • Only for subtask  (6)  applies:
$$\underline{y} = (1, 1, 0) =\underline{y}_6 \hspace{0.05cm}. $$

For decoding purposes, the exercise will examine:

  • the  "syndrome decoding",  which follows the concept  "hard decision maximum likelihood detection"  $\rm (HD-ML)$  for the codes under consideration.
    $($because soft values are not available at the BSC$)$;




Hints:

  • Reference is made in particular to the sections
  • The code word selected by the decoder is denoted in the questions by  $\underline{z}$.



Questions

1

Which statements are valid for decoding the  $\text{SPC (3, 2, 2)}$?

The hard decision syndrome decoding yields the result  $\underline{z} = (0, \, 1, \, 0)$.
The hard decision syndrome decoding returns the result  $\underline{z} = (0, \, 0, \, 0)$.
The hard decision syndrome decoding fails here.

2

Which statements are valid for the  $\text{ RC (3, 1, 3)}$?

The hard decision syndrome decoding returns the result  $\underline{z} = (0, \, 1, \, 0)$.
The hard decision syndrome decoding returns the result  $\underline{z} = (0, \, 0, \, 0)$.
The hard decision syndrome decoding fails here.

3

How certain is this decision if we define reliability  $($"security"$)$  $S$  as the quotient of the probabilities for a correct or an incorrect decision?
Set the falsification probability of the BSC model to  $\varepsilon = 26.9\%$.

$S \ = \ $

$\hspace{0.75cm} \ln {(S)} \ = \ $

4

What are the intrinsic log likelihood ratios for iterative bit-wise decoding of the  $\text{RC (3, 1)}$  received word  $\underline{y}_2 = (0, \, 1, \, 0)$?

$L_{\rm K}(1) \ = \ $

$L_{\rm K}(2) \ = \ $

$L_{\rm K}(3) \ = \ $

5

Which statements are true for decoding the received word  $\underline{y}_2 = (0, \, 1, \, 0)$?  Continue to assume the  $\text{RC (3, 1, 3)}$.

From the first iteration all signs of  $L_{\rm APP}(i)$  are positive.
Already after the second iteration  ${\rm Pr}(\underline{x}_0\hspace{0.05cm} |\hspace{0.05cm} \underline{y}_2)$  is greater than  $99\%$.
With each iteration the absolute values  $|L_{\rm APP}(i)|$  become larger.

6

Which statements are true for decoding the received word  $\underline{y}_6 = (1, 1, 0)$,  when  $\underline{x}_0 = (0, 0, 0)$  was sent?

The iterative decoder decides correctly.
The iterative decoder decides wrong.
The  "reliability"  for  $\underline{y}_6 \Rightarrow \underline{x}_0$  increases with increasing  $I$.


Solution

(1)  Correct is the  proposed solution 3:

  • The received word  $\underline{y}_2 = (0, 1, 0)$  is not a valid code word of the single parity–check code  $\text{SPC (3, 2)}$.  Thus,  the first statement is false.
  • In addition,  since the  $\text{SPC (3, 2)}$  has only the minimum distance  $d_{\rm min} = 2$,  no error can be corrected.


(2)  Correct is the  proposed solution 2:

  • The possible code words at the  $\text{RP (3, 1)}$  are  $\underline{x}_0 = (0, 0, 0)$  and  $\underline{x}_1 = (1, 1, 1)$.
  • The minimum distance of this code is  $d_{\rm min} = 3$,  so  $t = (d_{\rm min} \, - 1)/2 = 1$  error can be corrected.
  • In addition to  $\underline{y}_0 = (0, 0, 0)$,  the received words  $\underline{y}_1 = (0, 0, 1), \ \underline{y}_2 = (0, 1, 0)$,  and  $\underline{y}_4 = (1, 0, 0)$  are also assigned to the decoding result  $\underline{x}_0 = (0, 0, 0)$.


(3)  According to the BSC model,  the conditional probability that  $\underline{y}_2 = (0, 1, 0)$  is received,  given that  $\underline{x}_0 = (0, 0, 0)$  was sent:

$${\rm Pr}(\underline{y} = \underline{y}_2 \hspace{0.1cm}| \hspace{0.1cm}\underline{x} = \underline{x}_0 ) = (1-\varepsilon)^2 \cdot \varepsilon\hspace{0.05cm}.$$
  • The first term  $(1 \, –\varepsilon)^2$  indicates the probability that the first and third bit were transmitted correctly.  $\varepsilon$ describes the falsification probability for the second bit.
  • Correspondingly,  for the second possible code word  $\underline{x}_1 = (1, 1, 1)$:
$${\rm Pr}(\underline{y} = \underline{y}_2 \hspace{0.1cm}| \hspace{0.1cm}\underline{x} = \underline{x}_1 ) = \varepsilon^2 \cdot (1-\varepsilon) \hspace{0.05cm}.$$
  • According to Bayes' theorem,  the inference probabilities are then:
$${\rm Pr}(\hspace{0.1cm}\underline{x} = \underline{x}_0 \hspace{0.1cm}| \hspace{0.1cm}\underline{y} = \underline{y}_2 ) \hspace{-0.15cm} \ = \ \hspace{-0.15cm} {\rm Pr}(\underline{y} = \underline{y}_2 \hspace{0.1cm}| \hspace{0.1cm}\underline{x} = \underline{x}_0 ) \cdot \frac{{\rm Pr}(\underline{x} = \underline{x}_0)} {{\rm Pr}(\underline{y} = \underline{y}_2)} \hspace{0.05cm},$$
$${\rm Pr}(\hspace{0.1cm}\underline{x} = \underline{x}_1 \hspace{0.1cm}| \hspace{0.1cm}\underline{y} = \underline{y}_2 ) \hspace{-0.15cm} \ = \ \hspace{-0.15cm} {\rm Pr}(\underline{y} = \underline{y}_2 \hspace{0.1cm}| \hspace{0.1cm}\underline{x} = \underline{x}_1 ) \cdot \frac{{\rm Pr}(\underline{x} = \underline{x}_1)} {{\rm Pr}(\underline{y} = \underline{y}_2)} $$
$$\Rightarrow \hspace{0.3cm} S = \frac{{\rm Pr(correct \hspace{0.15cm}decision)}} {{\rm Pr(wrong \hspace{0.15cm}decision) }} = \frac{(1-\varepsilon)^2 \cdot \varepsilon}{\varepsilon^2 \cdot (1-\varepsilon)}= \frac{(1-\varepsilon)}{\varepsilon}\hspace{0.05cm}.$$
  • With  $\varepsilon = 0.269$  we get the following numerical values:
$$S = {0.731}/{0.269}\hspace{0.15cm}\underline {= 2.717}\hspace{0.3cm}\Rightarrow \hspace{0.3cm}{\rm ln}\hspace{0.15cm}(S)\hspace{0.15cm} \underline {= 1}\hspace{0.05cm}.$$


Iterative decoding of  $(+1, -1, +1)$

(4)  The sign of the channel log likelihood ratios  $L_{\rm K}(i)$  is

  • positive if  $y_i = 0$,  and
  • negative for  $y_i = 1$.


The absolute value   $|L_{\rm K}(i)|$  indicates the reliability of  $y_i$. 

  • In the BSC model,  $|L_{\rm K}(i)| = \ln {(1 \, – \varepsilon)/\varepsilon} = 1$  for all  $i$.  Thus:
$$\underline {L_{\rm K}}(1)\hspace{0.15cm} \underline {= +1}\hspace{0.05cm},\hspace{0.5cm} \underline {L_{\rm K}}(2)\hspace{0.15cm} \underline {= -1}\hspace{0.05cm},\hspace{0.5cm} \underline {L_{\rm K}}(3)\hspace{0.15cm} \underline {= +1}\hspace{0.05cm}.$$


(5)  The adjacent table illustrates the iterative bit-wise decoding starting from  $\underline{y}_2 = (0, \, 1, \, 0)$.

These results can be interpreted as follows:

  • The preassignment  $($iteration  $I = 0)$  happens according to  $\underline{L}_{\rm APP} = \underline{L}_{\rm K}$.  A hard decision  ⇒  "$\sign \ {\underline{L}_{\rm APP}(i)}$"  would lead to the decoding result  $(0, \, 1, \, 0)$.
  • The reliability of this obviously incorrect result is given as  $|{\it \Sigma}_{\rm APP}| = 1$.  This value agrees with  $\ln (S)$  calculated in subtask  (3).
  • After the first iteration  $(I = 1)$  all a-posteriori log likelihood ratios are  $L_{\rm APP}(i) = +1$.  A hard decision here would yield the  $($expected$)$  correct result  $\underline{x}_{\rm APP} = (0, \, 0, \, 0)$. 
  • The probability that this outcome is correct is quantified by  $|{\it \Sigma}_{\rm APP}| = 3$:
$${\rm ln}\hspace{0.25cm}\frac{{\rm Pr}(\hspace{0.1cm}\underline{x} = \underline{x}_0 \hspace{0.1cm}| \hspace{0.1cm}\underline{y}=\underline{y}_2)}{1-{\rm Pr}(\hspace{0.1cm}\underline{x} = \underline{x}_0 \hspace{0.1cm}| \hspace{0.1cm}\underline{y}=\underline{y}_2)} = 3 \hspace{0.3cm}\Rightarrow \hspace{0.3cm} \frac{{\rm Pr}(\hspace{0.1cm}\underline{x} = \underline{x}_0 \hspace{0.1cm}| \hspace{0.1cm}\underline{y}=\underline{y}_2)}{1-{\rm Pr}(\hspace{0.1cm}\underline{x} = \underline{x}_0 \hspace{0.1cm}| \hspace{0.1cm}\underline{y}=\underline{y}_2)} = {\rm e}^3 \approx 20$$
$$\hspace{0.3cm}\Rightarrow \hspace{0.3cm}{\rm Pr}(\hspace{0.1cm}\underline{x} = \underline{x}_0 \hspace{0.1cm}| \hspace{0.1cm}\underline{y}=\underline{y}_2) = {20}/{21} {\approx 95.39\%}\hspace{0.05cm}.$$
  • The second iteration confirms the decoding result of the first iteration.  The reliability is even quantified here with  $9$.  This value can be interpreted as follows:
$$\frac{{\rm Pr}(\hspace{0.1cm}\underline{x} = \underline{x}_0 \hspace{0.1cm}| \hspace{0.1cm}\underline{y}=\underline{y}_2)}{1-{\rm Pr}(\hspace{0.1cm}\underline{x} = \underline{x}_0 \hspace{0.1cm}| \hspace{0.1cm}\underline{y}=\underline{y}_2)} = {\rm e}^9 \hspace{0.3cm}\Rightarrow \hspace{0.3cm} {\rm Pr}(\hspace{0.1cm}\underline{x} = \underline{x}_0 \hspace{0.1cm}| \hspace{0.1cm}\underline{y}=\underline{y}_2) = {{\rm e}^9}/{({\rm e}^9+1)} \approx 99.99\% \hspace{0.05cm}.$$
  • With each further iteration the reliability value and thus the probability  ${\rm Pr}(\underline{x}_0 | \underline{y}_2)$  increases drastically   ⇒   All proposed solutions  are correct.


Iterative decoding of  $(–1, –1, +1)$

(6)  Correct are  the proposed solutions 2 and 3:

  • For the received vector  $\underline{y}_6 = (1, \, 1, \, 0)$,  the second table applies.
  • The decoder now decides for the sequence  $\underline{x}_1 = (1, \, 1, \, 1)$.
  • The case  »$\underline{y}_6 = (1, \, 1, \, 0)$  received under the condition  $\underline{x}_1 = (1, \, 1, \, 1)$  sent«  would correspond exactly to the constellation  »$\underline{y}_2 = (0, \, 1, \, 0)$  received and  $\underline{x}_0 = (0, \, 0, \, 0)$ sent«  considered in the last subtask.
  • But since  $\underline{x}_0 = (0, \, 0, \, 0)$  was sent,  there are now two bit errors with the following consequence:
  • The iterative decoder decides incorrectly.
  • With each further iteration the wrong decision is declared as more reliable.