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

From LNTwww
Line 3: Line 3:
 
[[File:P_ID2678__KC_Z_3_10.png|right|frame|Overall system model <br>enoder &ndash; channel &ndash; Viterbi]]
 
[[File:P_ID2678__KC_Z_3_10.png|right|frame|Overall system model <br>enoder &ndash; channel &ndash; Viterbi]]
 
The Viterbi algorithm represents the best known realization form for the maximum likelihood decoding of a convolutional code. We assume the following model here:
 
The Viterbi algorithm represents the best known realization form for the maximum likelihood decoding of a convolutional code. We assume the following model here:
* Die Informationssequenz&nbsp; $\underline{u}$&nbsp; wird durch einen Faltungscode in die Codesequenz&nbsp; $\underline{x}$&nbsp; umgesetzt. Es gelte&nbsp; $u_i &#8712; \{0, \, 1\}$. Dagegen werden die Codesymbole bipolar dargestellt &nbsp; &#8658; &nbsp; $x_i &#8712; \{&ndash;1, \, +1\}$.
+
* The information sequence&nbsp; $\underline{u}$&nbsp; is converted into the code sequence&nbsp; $\underline{x}$&nbsp; by a convolutional code. It is valid&nbsp; $u_i &#8712; \{0, \, 1\}$. In contrast, the code symbols are represented bipolar &nbsp; &#8658; &nbsp; $x_i &#8712; \{&ndash;1, \, +1\}$.
* Der Kanal sei durch das&nbsp; [[Channel_Coding/Channel_Models_and_Decision_Structures#Binary_Symmetric_Channel_.E2.80.93_BSC| "BSC&ndash;Modell"]]&nbsp; gegeben &nbsp; &#8658; &nbsp; $y_i &#8712; \{&ndash;1, \, +1\}$&nbsp; oder es wird der&nbsp; [[Channel_Coding/Channel_Models_and_Decision_Structures#AWGN_channel_at_Binary_Input| "AWGN&ndash;Kanal"]]&nbsp; vorausgesetzt &nbsp; &#8658; &nbsp; reellwertige Empfangswerte&nbsp; $y_i$.
+
* Let the channel be given by the&nbsp; [[Channel_Coding/Channel_Models_and_Decision_Structures#Binary_Symmetric_Channel_.E2.80. 93_BSC|"BSC&ndash;Model"]]&nbsp; given &nbsp; &#8658; &nbsp; $y_i &#8712; \{&ndash;1, \, +1\}$&nbsp; or the&nbsp; [[Channel_Coding/Channel_Models_and_Decision_Structures#AWGN_channel_at_Binary_Input| "AWGN&ndash;channel"]]&nbsp; provided &nbsp; &#8658; &nbsp; real-valued received values&nbsp; $y_i$.
* Bei gegebener Empfangssequenz&nbsp; $\underline{y}$&nbsp; entscheidet sich der Viterbi&ndash;Algorithmus für die Codesequenz&nbsp; $\underline{z}$&nbsp; gemäß
+
* Given a receive sequence&nbsp; $\underline{y}$&nbsp; the Viterbi algorithm decides on the code sequence&nbsp; $\underline{z}$&nbsp; according to
 
:$$\underline{z} = {\rm arg} \max_{\underline{x}_{\hspace{0.03cm}i} \hspace{0.03cm} \in \hspace{0.05cm} \mathcal{C}} \hspace{0.1cm} {\rm Pr}( \underline{x}_{\hspace{0.03cm}i} |\hspace{0.05cm} \underline{y} ) \hspace{0.05cm}.$$
 
:$$\underline{z} = {\rm arg} \max_{\underline{x}_{\hspace{0.03cm}i} \hspace{0.03cm} \in \hspace{0.05cm} \mathcal{C}} \hspace{0.1cm} {\rm Pr}( \underline{x}_{\hspace{0.03cm}i} |\hspace{0.05cm} \underline{y} ) \hspace{0.05cm}.$$
  
*Dies entspricht dem&nbsp; [[Channel_Coding/Channel_Models_and_Decision_Structures#Criteria_.C2.BBMaximum-a-posteriori.C2.AB_and_.C2.BBMaximum-Likelihood.C2.AB| "Maximum&ndash;a&ndash;posteriori"]]&nbsp; (MAP)&ndash;Kriterium. Sind alle Informationssequenzen&nbsp; $\underline{u}$&nbsp; gleichwahrscheinlich, so geht dieses in das etwas einfachere&nbsp; [[Channel_Coding/Channel_Models_and_Decision_Structures#Criteria_.C2.BBMaximum-a-posteriori.C2.AB_and_.C2.BBMaximum-Likelihood.C2.AB| "Maximum&ndash;Likelihood&ndash;Kriterium"]]&nbsp; über:
+
*This corresponds to the&nbsp; [[Channel_Coding/Channel_Models_and_Decision_Structures#Criteria_.C2.BBMaximum-a-posteriori.C2.AB_and_.C2.BBMaximum-Likelihood.C2.AB| "Maximum a posteriori"]]&nbsp; (MAP) criterion. If all information sequences&nbsp; $\underline{u}$&nbsp; are equally likely, this transitions to the somewhat simpler&nbsp; [[Channel_Coding/Channel_Models_and_Decision_Structures#Criteria_.C2.BBMaximum-a-posteriori.C2.AB_and_.C2.BBMaximum-Likelihood.C2.AB| "Maximum likelihood criterion"]]&nbsp;:  
 
:$$\underline{z} = {\rm arg} \max_{\underline{x}_{\hspace{0.03cm}i} \hspace{0.05cm} \in \hspace{0.05cm} \mathcal{C}} \hspace{0.1cm} {\rm Pr}( \underline{y}  \hspace{0.05cm}|\hspace{0.05cm} \underline{x}_{\hspace{0.03cm}i} ) \hspace{0.05cm}.$$
 
:$$\underline{z} = {\rm arg} \max_{\underline{x}_{\hspace{0.03cm}i} \hspace{0.05cm} \in \hspace{0.05cm} \mathcal{C}} \hspace{0.1cm} {\rm Pr}( \underline{y}  \hspace{0.05cm}|\hspace{0.05cm} \underline{x}_{\hspace{0.03cm}i} ) \hspace{0.05cm}.$$
  
*Als weiteres Ergebnis gibt der Viterbi&ndash;Algorithmus zusätzlich die Sequenz&nbsp; $\underline{v}$&nbsp; als Schätzung für die Informationssequenz&nbsp; $\underline{u}$&nbsp; aus.
+
*As a further result, the Viterbi&ndash;algorithm additionally outputs the sequence&nbsp; $\underline{v}$&nbsp; as an estimate for the information sequence&nbsp; $\underline{u}$&nbsp;.
  
  
In dieser Aufgabe soll der Zusammenhang zwischen der&nbsp; [[Channel_Coding/Objective_of_Channel_Coding#Important_definitions_for_block_coding| "Hamming&ndash;Distanz"]]&nbsp; $d_{\rm H}(\underline{x}, \, \underline{y})$&nbsp; sowie der&nbsp; [[Channel_Coding/Channel_Models_and_Decision_Structures#Maximum-likelihood_decision_at_the_AWGN_channel| "Euklidischen Distanz"]]
+
In this exercise, the relationship between the&nbsp; [[Channel_Coding/Objective_of_Channel_Coding#Important_definitions_for_block_coding| "Hamming&ndash;Distance"]]&nbsp; $d_{\rm H}(\underline{x}, \, \underline{y})$&nbsp; and the&nbsp; [[Channel_Coding/Channel_Models_and_Decision_Structures#Maximum-likelihood_decision_at_the_AWGN_channel| "Euclidean distance"]]
 
:$$d_{\rm E}(\underline{x}  \hspace{0.05cm}, \hspace{0.1cm}\underline{y}) =
 
:$$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}$$
 
\sqrt{\sum_{i=1}^{L} \hspace{0.2cm}(x_i - y_i)^2}\hspace{0.05cm}$$
  
ermittelt werden. Anschließend ist das obige ML&ndash;Kriterium  zu formulieren mit
+
are determined. Then, the above ML criterion is to be formulated with
* der Hamming&ndash;Distanz&nbsp; $d_{\rm H}(\underline{x}, \, \underline{y})$,
+
* the Hamming distance&nbsp; $d_{\rm H}(\underline{x}, \, \underline{y})$,
* der Euklidischen Distanz&nbsp; $d_{\rm E}(\underline{x}, \, \underline{y})$, und
+
* the Euclidean distance&nbsp; $d_{\rm E}(\underline{x}, \, \underline{y})$, and
* dem&nbsp; [[Channel_Coding/Decoding_of_Convolutional_Codes#Relationship_between_Hamming_distance_and_correlation| "Korrelationswert"]]&nbsp; $&#9001; x \cdot y &#9002;$.
+
* the&nbsp; [[Channel_Coding/Decoding_of_Convolutional_Codes#Relationship_between_Hamming_distance_and_correlation| "correlation value"]]&nbsp; $&#9001; x \cdot y &#9002;$.
  
  
Line 30: Line 30:
  
  
''Hinweise:''
+
Hints:
* Die Aufgabe bezieht sich auf das Kapitel&nbsp; [[Channel_Coding/Decodierung_von_Faltungscodes| Decodierung von Faltungscodes]].  
+
* This exercise refers to the chapter&nbsp; [[Channel_Coding/Decoding_of_Convolutional_Codes| "Decoding of Convolutional Codes"]].  
* Bezug genommen wird insbesondere auf  die  Seite&nbsp; [[Channel_Coding/Decodierung_von_Faltungscodes#Viterbi.E2.80.93Algorithmus.2C_basierend_auf_Korrelation_und_Metriken|Viterbi&ndash;Algorithmus &ndash; basierend auf Korrelation und Metriken]].
+
* Reference is made in particular to the page&nbsp; [[Channel_Coding/Decoding_of_Convolutional_Codes#Viterbi_algorithm_based_on_correlation_and_metrics|"Viterbi algorithm &ndash; based on correlation and metrics"]].
* Zur Vereinfachung wird auf "Tilde" und "Apostroph" verzichtet.
+
* For simplicity, "tilde" and "apostrophe" are omitted.
* Weitere Informationen zu diesem Thema finden Sie auf folgenden Seiten dieses Buches:
+
* For more information on this topic, see the following pages in this book:
** [[Channel_Coding/Kanalmodelle_und_Entscheiderstrukturen#Maximum-a-posteriori.E2.80.93_und_Maximum-Likelihood.E2.80.93Kriterium| MAP&ndash; und ML&ndash;Kriterium]],
+
** [[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/Kanalmodelle_und_Entscheiderstrukturen#Maximum-Likelihood.E2.80.93Entscheidung_beim_BSC.E2.80.93Kanal| ML&ndash;Entscheidung beim BSC&ndash;Kanal]],
+
** [[Channel_Coding/Channel_Models_and_Decision_Structures#Maximum-likelihood_decision_at_the_BSC_channel| "ML decision at the BSC channel"]],
** [[Channel_Coding/Kanalmodelle_und_Entscheiderstrukturen#Maximum-Likelihood.E2.80.93Entscheidung_beim_AWGN.E2.80.93Kanal| ML&ndash;Entscheidung beim AWGN&ndash;Kanal]],
+
** [[Channel_Coding/Channel_Models_and_Decision_Structures#Maximum-likelihood_decision_at_the_AWGN_channel| "ML decision at the AWGN channel"]],
** [[Channel_Coding/Decodierung_linearer_Blockcodes#Blockschaltbild_und_Voraussetzungen| Decodierung linearer Blockcodes]].
+
** [[Channel_Coding/Decoding_of_Linear_Block_Codes#Block_diagram_and_requirements| "Decoding linear block codes"]].
  
  
  
 
+
===Questions===
===Fragebogen===
 
 
<quiz display=simple>
 
<quiz display=simple>
{Wie hängen&nbsp; $d_{\rm H}(\underline{x}, \, \underline{y})$&nbsp; und&nbsp; $d_{\rm E}(\underline{x}, \, \underline{y})$&nbsp; beim BSC&ndash;Modell zusammen?
+
{How are&nbsp; $d_{\rm H}(\underline{x}, \, \underline{y})$&nbsp; and&nbsp; $d_{\rm E}(\underline{x}, \, \underline{y})$&nbsp; related in the BSC&ndash;model?
 
|type="()"}
 
|type="()"}
- Es gilt&nbsp; $d_{\rm H}(\underline{x}, \, \underline{y}) = d_{\rm E}(\underline{x}, \, \underline{y})$.
+
- &nbsp; $d_{\rm H}(\underline{x}, \, \underline{y}) = d_{\rm E}(\underline{x}, \, \underline{y})$ is valid.
- Es gilt&nbsp; $d_{\rm H}(\underline{x}, \, \underline{y}) = d_{\rm E}^2(\underline{x}, \, \underline{y})$.
+
- &nbsp; $d_{\rm H}(\underline{x}, \, \underline{y}) = d_{\rm E}^2(\underline{x}, \, \underline{y})$ is valid.
+ Es gilt&nbsp; $d_{\rm H}(\underline{x}, \, \underline{y}) = d_{\rm E}^2(\underline{x}, \, \underline{y})/4$.
+
+ &nbsp; $d_{\rm H}(\underline{x}, \, \underline{y}) = d_{\rm E}^2(\underline{x}, \, \underline{y})/4$ is valid.
  
{Welche der Gleichungen beschreiben die ML&ndash;Decodierung beim BSC&ndash;Modell? Die Minimierung/Maximierung bezieht sich jeweils auf alle&nbsp; $\underline{x} &#8712;\mathcal{ C}$.
+
{Which of the equations describe the ML decoding in the BSC model? The minimization/maximization refers to all&nbsp; $\underline{x} &#8712;\mathcal{ C}$, respectively.
 
|type="[]"}
 
|type="[]"}
 
+ $\underline{z} = \arg \min {d_{\rm H}(\underline{x}, \, \underline{y})}$,
 
+ $\underline{z} = \arg \min {d_{\rm H}(\underline{x}, \, \underline{y})}$,
Line 57: Line 56:
 
+ $\underline{z} = \arg \min {d_{\rm E}^2(\underline{x}, \, \underline{y})}$,
 
+ $\underline{z} = \arg \min {d_{\rm E}^2(\underline{x}, \, \underline{y})}$,
  
{Welche Gleichung beschreibt die ML&ndash;Entscheidung beim BSC&ndash;Modell?
+
{Which equation describes the ML decision in the BSC model?
 
|type="()"}
 
|type="()"}
 
- $\underline{z} = \arg \min &#9001; \underline{x} \cdot \underline{y} &#9002;$,
 
- $\underline{z} = \arg \min &#9001; \underline{x} \cdot \underline{y} &#9002;$,
 
+ $\underline{z} = \arg \max &#9001; \underline{x} \cdot \underline{y} &#9002;$.
 
+ $\underline{z} = \arg \max &#9001; \underline{x} \cdot \underline{y} &#9002;$.
  
{Welche Gleichungen gelten für die ML&ndash;Entscheidung beim AWGN&ndash;Modell?
+
{What equations apply to the ML decision in the AWGN model?
 
|type="[]"}
 
|type="[]"}
 
- $\underline{z} = \arg \min {d_{\rm H}(\underline{x}, \, \underline{y})}$,
 
- $\underline{z} = \arg \min {d_{\rm H}(\underline{x}, \, \underline{y})}$,
Line 69: Line 68:
 
</quiz>
 
</quiz>
  
===Musterlösung===
+
===Solution===
 
{{ML-Kopf}}
 
{{ML-Kopf}}
'''(1)'''&nbsp; Richtig ist der <u>Lösungsvorschlag 3</u>:
+
'''(1)'''&nbsp; Correct is the <u>proposed solution 3</u>:
*Die zwei Binärfolgen seien $\underline{x}$ und $\underline{y}$ mit $x_i &#8712; \{-1, \, +1\}, \ y_i &#8712; \{-1, \, +1\}$. Die Folgenlänge sei jeweils $L$.
+
*Let the two binary sequences be $\underline{x}$ and $\underline{y}$ with $x_i &#8712; \{-1, \, +1\}, \ y_i &#8712; \{-1, \, +1\}$. Let the sequence length be $L$ in each case.
*Die Hamming&ndash;Distanz $d_{\rm H}(\underline{x}, \, \underline{y})$ gibt die Anzahl der Bit an, in denen sich $\underline{x}$ und $\underline{y}$ unterscheiden, für die also $x_i \, - y_i = &plusmn;2$ &nbsp; &#8658; &nbsp; $ (x_i \, - y_i)^2 = 4$ gilt.  
+
*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 $x_i \, - y_i = &plusmn;2$ &nbsp; &#8658; &nbsp; $ (x_i \, - y_i)^2 = 4$ holds.  
*Gleiche Symbole $(x_i = y_i)$ tragen zur Hamming&ndash;Distanz nicht bei und ergeben $(x_i \, &ndash; y_i)^2 = 0$. Nach dem <u>Lösungsvorschlag 3</u> kann daher geschrieben werden:
+
*Equal symbols $(x_i = y_i)$ 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:
 
:$$ 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>Alle Lösungsvorschläge</u> sind richtig:
+
'''(2)'''&nbsp; <u>All proposed solutions</u> are correct:
*Beim BSC&ndash;Modell ist es allgemein üblich, zum gegebenen Empfangsvektor $\underline{y}$ das Codewort $\underline{x}$ mit der kleinsten Hamming&ndash;Distanz $d_{\rm H}(\underline{x}, \, \underline{y})$ auszuwählen:
+
*In the BSC model, it is common practice to select the codeword $\underline{x}$ 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}
 
:$$\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}.$$
*Entsprechend der Teilaufgabe '''(1)''' gilt aber auch:
+
*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}
 
:$$\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 92: Line 91:
 
\hspace{0.05cm}.$$
 
\hspace{0.05cm}.$$
  
*Der Faktor $1/4$ spielt für die Minimierung keine Rolle. Da $d_{\rm E}(\underline{x}, \, \underline{y}) &#8805; 0$ ist, ist es auch egal, ob die Minimierung hinsichtlich $d_{\rm E}(\underline{x}, \, \underline{y})$ oder $d_{\rm E}^2(\underline{x}, \, \underline{y})$ erfolgt.  
+
*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 $d_{\rm E}(\underline{x}, \, \underline{y})$ or $d_{\rm E}^2(\underline{x}, \, \underline{y})$.  
  
  
  
'''(3)'''&nbsp; Richtig ist der <u>Lösungsvorschlag 2</u>:
+
'''(3)'''&nbsp; Correct is the <u>proposed solution 2</u>:
*Das Quadrat der Euklidischen Distanz kann wie folgt ausgedrückt werden:
+
*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}) =
 
\sum_{i=1}^{L} \hspace{0.2cm}(x_i - y_i)^2 =  
 
\sum_{i=1}^{L} \hspace{0.2cm}(x_i - y_i)^2 =  
Line 104: Line 103:
 
\hspace{0.05cm}.$$
 
\hspace{0.05cm}.$$
  
*Die beiden ersten Summanden sind jeweils gleich $L$ und müssen für die Minimierung nicht berücksichtigt werden.  
+
*The first two summands are each equal to $L$ and need not be considered for minimization.  
*Für den letzten Ausdruck in dieser Gleichung kann $&ndash;2 \cdot &#9001; \underline{x}, \, \underline{y} &#9002;$ geschrieben werden.  
+
*For the last expression in this equation, $&ndash;2 \cdot &#9001; \underline{x}, \, \underline{y} &#9002;$ can be written.  
*Aufgrund des negativen Vorzeichens wird aus der Minimierung eine Maximierung &nbsp; &#8658; &nbsp; <u>Antwort 2</u>.
+
*Due to the negative sign, minimization becomes maximization &nbsp; &#8658; &nbsp; <u>answer 2</u>.
  
  
  
'''(4)'''&nbsp; Richtig sind die <u>Lösungsvorschläge 2 und 3</u>.:
+
'''(4)'''&nbsp; Correct are <u>proposed solutions 2 and 3</u>:
*Für den AWGN&ndash;Kanal kann im Gegensatz zum BSC keine Hamming&ndash;Distanz angegeben werden.  
+
*For the AWGN channel, unlike the BSC, no Hamming distance can be specified.  
*Ausgehend von der Gleichung
+
*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}) =  
 
\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}\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{0.1cm}-2 \cdot \sum_{i=1}^{L} \hspace{0.1cm} x_i \cdot y_i$$
  
:gelten für den ersten und letzten Summanden die gleichen Aussagen wie für das BSC&ndash;Modell &ndash; siehe Teilaufgabe (3).  
+
:the same statements apply for the first and last summands as for the BSC model &ndash; see subtask (3).  
*Für den mittleren Summanden gilt mit $y_i = x_i + n_i$ und $x_i &#8712; \{&ndash;1, \, +1\}$:  
+
*For the middle summand, $y_i = x_i + n_i$ and $x_i &#8712; \{&ndash;1, \, +1\}$ 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}.$$
  
*Der erste Summand ergibt wieder $L$, der zweite ist proportional zur Rauschleistung und der letzte Term verschwindet, da $\underline{x}$ und $\underline{n}$ unkorreliert sind.  
+
*The first summand again gives $L$, the second is proportional to the noise power, and the last term vanishes since $\underline{x}$ and $\underline{n}$ are uncorrelated.  
*Für die Minimerung von $d_{\rm E}(\underline{x}, \, \underline{y})$ muss also die Summe über $y_i^2$ nicht berücksichtigt werden, da kein Bezug zu den Codesequenzen $\underline{x}$ besteht.
+
*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 $\underline{x}$.
 
{{ML-Fuß}}
 
{{ML-Fuß}}
  

Revision as of 22:27, 18 October 2022

Overall system model
enoder – channel – Viterbi

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

  • The information sequence  $\underline{u}$  is converted into the code sequence  $\underline{x}$  by a convolutional code. It is valid  $u_i ∈ \{0, \, 1\}$. In contrast, the code symbols are represented bipolar   ⇒   $x_i ∈ \{–1, \, +1\}$.
  • Let the channel be given by the  "BSC–Model"  given   ⇒   $y_i ∈ \{–1, \, +1\}$  or the  "AWGN–channel"  provided   ⇒   real-valued received values  $y_i$.
  • Given a receive sequence  $\underline{y}$  the Viterbi algorithm decides on the code 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, the relationship between the  "Hamming–Distance"  $d_{\rm H}(\underline{x}, \, \underline{y})$  and the  "Euclidean distance"

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

are determined. Then, the above ML 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
  • the  "correlation value"  $〈 x \cdot y 〉$.





Hints:


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 ML decoding in the BSC model? The minimization/maximization refers to all  $\underline{x} ∈\mathcal{ C}$, respectively.

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

3

Which equation describes the ML 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 ML 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 codeword $\underline{x}$ 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 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 again gives $L$, 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 $\underline{x}$.