Loading [MathJax]/jax/output/HTML-CSS/fonts/TeX/fontdata.js

Difference between revisions of "Aufgaben:Exercise 3.10Z: Maximum Likelihood Decoding of Convolutional Codes"

From LNTwww
 
(2 intermediate revisions by the same user not shown)
Line 1: Line 1:
 
{{quiz-Header|Buchseite=Channel_Coding/Decoding_of_Convolutional_Codes}}
 
{{quiz-Header|Buchseite=Channel_Coding/Decoding_of_Convolutional_Codes}}
  
[[File:P_ID2678__KC_Z_3_10.png|right|frame|Overall system model,  given for this exercise]]
+
[[File:EN_KC_Z_3_10.png|right|frame|Overall system model,  given for this exercise]]
 
The Viterbi algorithm represents the best known realization form for the maximum likelihood decoding of a convolutional code.  We assume here the following model:
 
The Viterbi algorithm represents the best known realization form for the maximum likelihood decoding of a convolutional code.  We assume here the following model:
 
* The information sequence  u_  is converted into the code sequence  x_  by a convolutional code.  
 
* The information sequence  u_  is converted into the code sequence  x_  by a convolutional code.  
Line 44: Line 44:
  
 
* For more information on this topic,  see the following sections in this book:
 
* For more information on this topic,  see the following sections in this book:
** [[Channel_Coding/Channel_Models_and_Decision_Structures#Criteria_.C2.BBMaximum-a-posteriori.C2.AB_and_.C2.BBMaximum-Likelihood.C2.AB| "MAP and ML criterion"]],
+
:* [[Channel_Coding/Channel_Models_and_Decision_Structures#Criteria_.C2.BBMaximum-a-posteriori.C2.AB_and_.C2.BBMaximum-Likelihood.C2.AB| "MAP and ML criterion"]],
  
** [[Channel_Coding/Channel_Models_and_Decision_Structures#Maximum-likelihood_decision_at_the_BSC_channel| "ML decision at the BSC channel"]],
+
:* [[Channel_Coding/Channel_Models_and_Decision_Structures#Maximum-likelihood_decision_at_the_BSC_channel| "ML decision at the BSC channel"]],
  
** [[Channel_Coding/Channel_Models_and_Decision_Structures#Maximum-likelihood_decision_at_the_AWGN_channel| "ML decision at the AWGN channel"]],
+
:* [[Channel_Coding/Channel_Models_and_Decision_Structures#Maximum-likelihood_decision_at_the_AWGN_channel| "ML decision at the AWGN channel"]],
  
** [[Channel_Coding/Decoding_of_Linear_Block_Codes#Block_diagram_and_requirements| "Decoding linear block codes"]].
+
:* [[Channel_Coding/Decoding_of_Linear_Block_Codes#Block_diagram_and_requirements| "Decoding linear block codes"]].
  
  
Line 58: Line 58:
 
{How are   dH(x_,y_)   and   dE(x_,y_)   related in the BSC model?
 
{How are   dH(x_,y_)   and   dE(x_,y_)   related in the BSC model?
 
|type="()"}
 
|type="()"}
-   dH(x_,y_)=dE(x_,y_) is valid.
+
-   dH(x_,y_)=dE(x_,y_)  is valid.
-   dH(x_,y_)=d2E(x_,y_) is valid.
+
-   dH(x_,y_)=d2E(x_,y_)  is valid.
+   dH(x_,y_)=d2E(x_,y_)/4 is valid.
+
+   dH(x_,y_)=d2E(x_,y_)/4  is valid.
  
 
{Which of the equations describe the maximum likelihood decoding in the BSC model?  The minimization/maximization refers alwaysto all  \underline{x} ∈\mathcal{ C}.
 
{Which of the equations describe the maximum likelihood decoding in the BSC model?  The minimization/maximization refers alwaysto all  \underline{x} ∈\mathcal{ C}.
Line 82: Line 82:
 
===Solution===
 
===Solution===
 
{{ML-Kopf}}
 
{{ML-Kopf}}
'''(1)'''&nbsp; Correct is the <u>proposed solution 3</u>:
+
'''(1)'''&nbsp; Correct is the&nbsp; <u>proposed solution 3</u>:
*Let the two binary sequences be x_ and y_ with x_i &#8712; \{-1, \, +1\}, \ y_i &#8712; \{-1, \, +1\}. Let the sequence length be L in each case.
+
*Let the two binary sequences be &nbsp; x_ &nbsp; and &nbsp; y_ &nbsp; with &nbsp; x_i &#8712; \{-1, \, +1\}, \ y_i &#8712; \{-1, \, +1\}.&nbsp; Let the sequence length be&nbsp; L&nbsp; in each case.
*The Hamming distance dH(x_,y_) gives the number of bits in which x_ and y_ differ, for which thus x_i \, - y_i = &plusmn;2 &nbsp; &#8658; &nbsp; (xiyi)2=4 holds.  
+
 
*Equal symbols (xi=yi) do not contribute to the Hamming&ndash;distance and give (x_i \, &ndash; y_i)^2 = 0. According to the <u>solution 3</u>, we can therefore write:
+
*The Hamming distance &nbsp; dH(x_,y_) &nbsp; gives the number of bits in which&nbsp; x_&nbsp; and&nbsp; y_&nbsp; differ,&nbsp; for which thus&nbsp; x_i \, - y_i = &plusmn;2 &nbsp; &#8658; &nbsp; (xiyi)2=4&nbsp; holds.
 +
 +
*Equal symbols&nbsp; (xi=yi)&nbsp; do not contribute to the Hamming distance and give&nbsp; (x_i \, &ndash; y_i)^2 = 0.&nbsp; According to the&nbsp; <u>solution 3</u>,&nbsp; we can therefore write:
 
:$$ d_{\rm H}(\underline{x}  \hspace{0.05cm}, \hspace{0.1cm}\underline{y}) =
 
:$$ d_{\rm H}(\underline{x}  \hspace{0.05cm}, \hspace{0.1cm}\underline{y}) =
 
\frac{1}{4} \cdot \sum_{i=1}^{L} \hspace{0.2cm}(x_i - y_i)^2= \frac{1}{4} \cdot d_{\rm E}^2(\underline{x}  \hspace{0.05cm}, \hspace{0.1cm}\underline{y})\hspace{0.05cm}.$$
 
\frac{1}{4} \cdot \sum_{i=1}^{L} \hspace{0.2cm}(x_i - y_i)^2= \frac{1}{4} \cdot d_{\rm E}^2(\underline{x}  \hspace{0.05cm}, \hspace{0.1cm}\underline{y})\hspace{0.05cm}.$$
  
  
'''(2)'''&nbsp; <u>All proposed solutions</u> are correct:
+
'''(2)'''&nbsp; <u>All proposed solutions</u>&nbsp; are correct:
*In the BSC model, it is common practice to select the code word x_ with the smallest Hamming distance dH(x_,y_) for the given received vector y_:
+
*In the BSC model,&nbsp; it is common practice to select the code word&nbsp; x_&nbsp; with the smallest Hamming distance&nbsp; dH(x_,y_)&nbsp; for the given received vector&nbsp; y_:
 
:$$\underline{z} = {\rm arg} \min_{\underline{x} \hspace{0.05cm} \in \hspace{0.05cm} \mathcal{C}} \hspace{0.1cm}
 
:$$\underline{z} = {\rm arg} \min_{\underline{x} \hspace{0.05cm} \in \hspace{0.05cm} \mathcal{C}} \hspace{0.1cm}
 
d_{\rm H}(\underline{x}  \hspace{0.05cm}, \hspace{0.1cm}\underline{y})\hspace{0.05cm}.$$
 
d_{\rm H}(\underline{x}  \hspace{0.05cm}, \hspace{0.1cm}\underline{y})\hspace{0.05cm}.$$
*But according to the subtask '''(1)''' also applies:
+
*But according to the subtask&nbsp; '''(1)'''&nbsp; also applies:
 
:$$\underline{z} = {\rm arg} \min_{\underline{x} \hspace{0.05cm} \in \hspace{0.05cm} \mathcal{C}} \hspace{0.1cm}
 
:$$\underline{z} = {\rm arg} \min_{\underline{x} \hspace{0.05cm} \in \hspace{0.05cm} \mathcal{C}} \hspace{0.1cm}
 
d_{\rm E}^{\hspace{0.15cm}2}(\underline{x}  \hspace{0.05cm}, \hspace{0.1cm}\underline{y})/4
 
d_{\rm E}^{\hspace{0.15cm}2}(\underline{x}  \hspace{0.05cm}, \hspace{0.1cm}\underline{y})/4
Line 103: Line 105:
 
\hspace{0.05cm}.$$
 
\hspace{0.05cm}.$$
  
*The factor 1/4 does not matter for the minimization. Since d_{\rm E}(\underline{x}, \, \underline{y}) &#8805; 0, it does not matter whether the minimization is done with respect to dE(x_,y_) or d2E(x_,y_).  
+
*The factor&nbsp; 1/4&nbsp; does not matter for the minimization.&nbsp; Since&nbsp; d_{\rm E}(\underline{x}, \, \underline{y}) &#8805; 0,&nbsp; it does not matter whether the minimization is done with respect to &nbsp; dE(x_,y_) &nbsp; or &nbsp; d2E(x_,y_).  
  
  
  
'''(3)'''&nbsp; Correct is the <u>proposed solution 2</u>:
+
'''(3)'''&nbsp; Correct is the&nbsp; <u>proposed solution 2</u>:
 
*The square of the Euclidean distance can be expressed as follows:
 
*The square of the Euclidean distance can be expressed as follows:
 
:$$d_{\rm E}^{\hspace{0.15cm}2}(\underline{x}  \hspace{0.05cm}, \hspace{0.1cm}\underline{y}) =
 
:$$d_{\rm E}^{\hspace{0.15cm}2}(\underline{x}  \hspace{0.05cm}, \hspace{0.1cm}\underline{y}) =
Line 115: Line 117:
 
\hspace{0.05cm}.$$
 
\hspace{0.05cm}.$$
  
*The first two summands are each equal to L and need not be considered for minimization.  
+
*The first two summands are each equal to&nbsp; L&nbsp; and need not be considered for minimization.
*For the last expression in this equation, &ndash;2 \cdot &#9001; \underline{x}, \, \underline{y} &#9002; can be written.  
+
*Due to the negative sign, minimization becomes maximization &nbsp; &#8658; &nbsp; <u>answer 2</u>.
+
*For the last expression in this equation,&nbsp; &ndash;2 \cdot &#9001; \underline{x}, \, \underline{y} &#9002;&nbsp; can be written.
 +
 +
*Due to the negative sign,&nbsp; minimization becomes maximization &nbsp; &#8658; &nbsp; <u>answer 2</u>.
  
  
  
'''(4)'''&nbsp; Correct are <u>proposed solutions 2 and 3</u>:
+
'''(4)'''&nbsp; Correct are the&nbsp; <u>proposed solutions 2 and 3</u>:
*For the AWGN channel, unlike the BSC, no Hamming distance can be specified.  
+
*For the AWGN channel,&nbsp; unlike the BSC,&nbsp; no Hamming distance can be specified.
 +
 
*Based on the equation
 
*Based on the equation
 
:$$d_{\rm E}^{\hspace{0.15cm}2}(\underline{x}  \hspace{0.05cm}, \hspace{0.1cm}\underline{y}) =  
 
:$$d_{\rm E}^{\hspace{0.15cm}2}(\underline{x}  \hspace{0.05cm}, \hspace{0.1cm}\underline{y}) =  
Line 128: Line 133:
 
\hspace{0.1cm}-2 \cdot \sum_{i=1}^{L} \hspace{0.1cm} x_i \cdot y_i$$
 
\hspace{0.1cm}-2 \cdot \sum_{i=1}^{L} \hspace{0.1cm} x_i \cdot y_i$$
  
:the same statements apply for the first and last summands as for the BSC model &ndash; see subtask (3).  
+
:the same statements apply for the first and last summands as for the BSC model &ndash; see subtask&nbsp; '''(3)'''.  
*For the middle summand, yi=xi+ni and x_i &#8712; \{&ndash;1, \, +1\} hold:  
+
 
 +
*For the middle summand,&nbsp; yi=xi+ni&nbsp; and&nbsp; x_i &#8712; \{&ndash;1, \, +1\}&nbsp; hold:  
 
:$$\sum_{i=1}^{L} \hspace{0.1cm} y_i^{\hspace{0.15cm}2} =  
 
:$$\sum_{i=1}^{L} \hspace{0.1cm} y_i^{\hspace{0.15cm}2} =  
 
\hspace{0.1cm}\sum_{i=1}^{L} \hspace{0.1cm} x_i^{\hspace{0.15cm}2} \hspace{0.1cm}+ \hspace{0.1cm}\sum_{i=1}^{L} \hspace{0.1cm} n_i^{\hspace{0.15cm}2}
 
\hspace{0.1cm}\sum_{i=1}^{L} \hspace{0.1cm} x_i^{\hspace{0.15cm}2} \hspace{0.1cm}+ \hspace{0.1cm}\sum_{i=1}^{L} \hspace{0.1cm} n_i^{\hspace{0.15cm}2}
 
\hspace{0.1cm}+2 \cdot \sum_{i=1}^{L} \hspace{0.1cm} x_i \cdot n_i \hspace{0.05cm}.$$
 
\hspace{0.1cm}+2 \cdot \sum_{i=1}^{L} \hspace{0.1cm} x_i \cdot n_i \hspace{0.05cm}.$$
  
*The first summand again gives L, the second is proportional to the noise power, and the last term vanishes since x_ and n_ are uncorrelated.  
+
*The first summand gives again&nbsp; L,&nbsp; the second is proportional to the noise power,&nbsp; and the last term vanishes since&nbsp; x_&nbsp; and&nbsp; n_&nbsp; are uncorrelated.
*So for minimizing dE(x_,y_), the sum over y2i need not be considered since there is no relation to the code sequences x_.
+
 +
*So for minimizing&nbsp; dE(x_,y_),&nbsp; the sum over&nbsp; y2i&nbsp; need not be considered since there is no relation to the code sequences&nbsp; x_.
 
{{ML-Fuß}}
 
{{ML-Fuß}}
  

Latest revision as of 16:57, 21 November 2022

Overall system model,  given for this exercise

The Viterbi algorithm represents the best known realization form for the maximum likelihood decoding of a convolutional code.  We assume here the following model:

  • The information sequence  u_  is converted into the code sequence  x_  by a convolutional code.
  • It is valid  ui{0,1}.  In contrast,  the code symbols are represented bipolar   ⇒   x_i ∈ \{–1, \, +1\}.
  • Given a received sequence  \underline{y}  the Viterbi algorithm decides for the sequence  \underline{z}  according to
\underline{z} = {\rm arg} \max_{\underline{x}_{\hspace{0.03cm}i} \hspace{0.03cm} \in \hspace{0.05cm} \mathcal{C}} \hspace{0.1cm} {\rm Pr}( \underline{x}_{\hspace{0.03cm}i} |\hspace{0.05cm} \underline{y} ) \hspace{0.05cm}.
\underline{z} = {\rm arg} \max_{\underline{x}_{\hspace{0.03cm}i} \hspace{0.05cm} \in \hspace{0.05cm} \mathcal{C}} \hspace{0.1cm} {\rm Pr}( \underline{y} \hspace{0.05cm}|\hspace{0.05cm} \underline{x}_{\hspace{0.03cm}i} ) \hspace{0.05cm}.
  • As a further result,  the Viterbi algorithm additionally outputs the sequence  \underline{v}  as an estimate for the information sequence  \underline{u}.


In this exercise,  you should determinethe relationship between the  \text{Hamming distance}  d_{\rm H}(\underline{x}, \, \underline{y})  and the  \text{Euclidean distance}

d_{\rm E}(\underline{x} \hspace{0.05cm}, \hspace{0.1cm}\underline{y}) = \sqrt{\sum_{i=1}^{L} \hspace{0.2cm}(x_i - y_i)^2}\hspace{0.05cm}.

Then,  the above maximum likelihood criterion is to be formulated with

  • the Hamming distance  d_{\rm H}(\underline{x}, \, \underline{y}),
  • the Euclidean distance  d_{\rm E}(\underline{x}, \, \underline{y}),  and





Hints:

  • For simplicity,  "tilde"  and  "apostrophe"  are omitted.
  • For more information on this topic,  see the following sections in this book:


Questions

1

How are   d_{\rm H}(\underline{x}, \, \underline{y})   and   d_{\rm E}(\underline{x}, \, \underline{y})   related in the BSC model?

  d_{\rm H}(\underline{x}, \, \underline{y}) = d_{\rm E}(\underline{x}, \, \underline{y})  is valid.
  d_{\rm H}(\underline{x}, \, \underline{y}) = d_{\rm E}^2(\underline{x}, \, \underline{y})  is valid.
  d_{\rm H}(\underline{x}, \, \underline{y}) = d_{\rm E}^2(\underline{x}, \, \underline{y})/4  is valid.

2

Which of the equations describe the maximum likelihood decoding in the BSC model?  The minimization/maximization refers alwaysto all  \underline{x} ∈\mathcal{ C}.

\underline{z} = \arg \min {d_{\rm H}(\underline{x}, \, \underline{y})},
\underline{z} = \arg \min {d_{\rm E}(\underline{x}, \, \underline{y})},
\underline{z} = \arg \min {d_{\rm E}^2(\underline{x}, \, \underline{y})},

3

Which equation describes the maximum likelihood decision in the BSC model?

\underline{z} = \arg \min 〈 \underline{x} \cdot \underline{y} 〉,
\underline{z} = \arg \max 〈 \underline{x} \cdot \underline{y} 〉.

4

What equations apply to the maximum likelihood decision in the AWGN model?

\underline{z} = \arg \min {d_{\rm H}(\underline{x}, \, \underline{y})},
\underline{z} = \arg \min {d_{\rm E}(\underline{x}, \, \underline{y})},
\underline{z} = \arg \max 〈 \underline{x} \cdot \underline{y} 〉.


Solution

(1)  Correct is the  proposed solution 3:

  • Let the two binary sequences be   \underline{x}   and   \underline{y}   with   x_i ∈ \{-1, \, +1\}, \ y_i ∈ \{-1, \, +1\}.  Let the sequence length be  L  in each case.
  • The Hamming distance   d_{\rm H}(\underline{x}, \, \underline{y})   gives the number of bits in which  \underline{x}  and  \underline{y}  differ,  for which thus  x_i \, - y_i = ±2   ⇒   (x_i \, - y_i)^2 = 4  holds.
  • Equal symbols  (x_i = y_i)  do not contribute to the Hamming distance and give  (x_i \, – y_i)^2 = 0.  According to the  solution 3,  we can therefore write:
d_{\rm H}(\underline{x} \hspace{0.05cm}, \hspace{0.1cm}\underline{y}) = \frac{1}{4} \cdot \sum_{i=1}^{L} \hspace{0.2cm}(x_i - y_i)^2= \frac{1}{4} \cdot d_{\rm E}^2(\underline{x} \hspace{0.05cm}, \hspace{0.1cm}\underline{y})\hspace{0.05cm}.


(2)  All proposed solutions  are correct:

  • In the BSC model,  it is common practice to select the code word  \underline{x}  with the smallest Hamming distance  d_{\rm H}(\underline{x}, \, \underline{y})  for the given received vector  \underline{y}:
\underline{z} = {\rm arg} \min_{\underline{x} \hspace{0.05cm} \in \hspace{0.05cm} \mathcal{C}} \hspace{0.1cm} d_{\rm H}(\underline{x} \hspace{0.05cm}, \hspace{0.1cm}\underline{y})\hspace{0.05cm}.
  • But according to the subtask  (1)  also applies:
\underline{z} = {\rm arg} \min_{\underline{x} \hspace{0.05cm} \in \hspace{0.05cm} \mathcal{C}} \hspace{0.1cm} d_{\rm E}^{\hspace{0.15cm}2}(\underline{x} \hspace{0.05cm}, \hspace{0.1cm}\underline{y})/4 \hspace{0.2cm}\Rightarrow \hspace{0.2cm} \underline{z} = {\rm arg} \min_{\underline{x} \hspace{0.05cm} \in \hspace{0.05cm} \mathcal{C}} \hspace{0.1cm} d_{\rm E}^{\hspace{0.15cm}2}(\underline{x} \hspace{0.05cm}, \hspace{0.1cm}\underline{y}) \hspace{0.2cm}\Rightarrow \hspace{0.2cm} \underline{z} = {\rm arg} \min_{\underline{x} \hspace{0.05cm} \in \hspace{0.05cm} \mathcal{C}} \hspace{0.1cm} d_{\rm E}(\underline{x} \hspace{0.05cm}, \hspace{0.1cm}\underline{y}) \hspace{0.05cm}.
  • The factor  1/4  does not matter for the minimization.  Since  d_{\rm E}(\underline{x}, \, \underline{y}) ≥ 0,  it does not matter whether the minimization is done with respect to   d_{\rm E}(\underline{x}, \, \underline{y})   or   d_{\rm E}^2(\underline{x}, \, \underline{y}).


(3)  Correct is the  proposed solution 2:

  • The square of the Euclidean distance can be expressed as follows:
d_{\rm E}^{\hspace{0.15cm}2}(\underline{x} \hspace{0.05cm}, \hspace{0.1cm}\underline{y}) = \sum_{i=1}^{L} \hspace{0.2cm}(x_i - y_i)^2 = \hspace{0.1cm}\sum_{i=1}^{L} \hspace{0.1cm} x_i^{\hspace{0.15cm}2} \hspace{0.1cm}+ \hspace{0.1cm}\sum_{i=1}^{L} \hspace{0.1cm} y_i^{\hspace{0.15cm}2} \hspace{0.1cm}-2 \cdot \sum_{i=1}^{L} \hspace{0.1cm} x_i \cdot y_i \hspace{0.05cm}.
  • The first two summands are each equal to  L  and need not be considered for minimization.
  • For the last expression in this equation,  –2 \cdot 〈 \underline{x}, \, \underline{y} 〉  can be written.
  • Due to the negative sign,  minimization becomes maximization   ⇒   answer 2.


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

  • For the AWGN channel,  unlike the BSC,  no Hamming distance can be specified.
  • Based on the equation
d_{\rm E}^{\hspace{0.15cm}2}(\underline{x} \hspace{0.05cm}, \hspace{0.1cm}\underline{y}) = \hspace{0.1cm}\sum_{i=1}^{L} \hspace{0.1cm} x_i^{\hspace{0.15cm}2} \hspace{0.1cm}+ \hspace{0.1cm}\sum_{i=1}^{L} \hspace{0.1cm} y_i^{\hspace{0.15cm}2} \hspace{0.1cm}-2 \cdot \sum_{i=1}^{L} \hspace{0.1cm} x_i \cdot y_i
the same statements apply for the first and last summands as for the BSC model – see subtask  (3).
  • For the middle summand,  y_i = x_i + n_i  and  x_i ∈ \{–1, \, +1\}  hold:
\sum_{i=1}^{L} \hspace{0.1cm} y_i^{\hspace{0.15cm}2} = \hspace{0.1cm}\sum_{i=1}^{L} \hspace{0.1cm} x_i^{\hspace{0.15cm}2} \hspace{0.1cm}+ \hspace{0.1cm}\sum_{i=1}^{L} \hspace{0.1cm} n_i^{\hspace{0.15cm}2} \hspace{0.1cm}+2 \cdot \sum_{i=1}^{L} \hspace{0.1cm} x_i \cdot n_i \hspace{0.05cm}.
  • The first summand gives again  L,  the second is proportional to the noise power,  and the last term vanishes since  \underline{x}  and  \underline{n}  are uncorrelated.
  • So for minimizing  d_{\rm E}(\underline{x}, \, \underline{y}),  the sum over  y_i^2  need not be considered since there is no relation to the code sequences  \underline{x}.