<< Chapter < Page Chapter >> Page >

Theorem 51 H p ( z ) forms an order K , unitary filter bank with PCS symmetry iff

H p ( z ) = I 0 0 J i = 1 K A i z - 1 B i - B i z - 1 A i A 0 B 0 - B 0 A 0 I 0 0 J P

where A i B i - B i A i are constant orthogonal matrices. H p ( z ) is characterized by 2 K M / 2 2 parameters.

Characterization of unitary H p ( z ) — linear-phase symmetry

For the linear-phase case,

H p ( z ) = Q W 0 ( z ) W 0 R ( z ) W 1 ( z ) - W 1 R ( z ) I 0 0 J = 1 2 Q I I I - I W 0 ( z ) + W 1 ( z ) W 0 R ( z ) - W 1 R ( z ) W 0 ( z ) - W 1 ( z ) W 0 R ( z ) + W 1 R ( z ) I 0 0 J = def 1 2 Q I I I - I W 0 ' ( z ) W 1 ' ( z ) ( W 1 ' ) R ( z ) ( W 0 ' ) R ( z ) I 0 0 J .

Therefore, we have the following Theorem:

Theorem 52 H p ( z ) of order K , forms a unitary filter bank with linear-phase filters iff

H p ( z ) = 1 2 Q I I I - I i = 1 K A i z - 1 B i B i z - 1 A i A 0 B 0 B 0 A 0 I 0 0 J ,

where A i B i B i A i are constant orthogonal matrices. H p ( z ) is characterized by 2 K M / 2 2 parameters.

Characterization of unitary H p ( z ) — linear phase and pcs symmetry

In this case, H p ( z ) is given by

H p ( z ) = I 0 0 J W 0 ( z ) D W 0 R ( z ) J D W 0 ( z ) V ( - 1 ) M / 2 W 0 R ( z ) J V = def 1 2 I 0 0 J W 0 ' ( z ) D ( W 0 ' ) R ( z ) J - D W 0 ' ( z ) ( W 0 ' ) R ( z ) J P = 1 2 I 0 0 J I D - D I W 0 ' ( z ) 0 0 ( W 0 ' ) R ( z ) I 0 0 J P .

Therefore we have proved the following Theorem:

Theorem 53 H p ( z ) of order K forms a unitary filter bank with linear-phase and PCS filters iff there exists a unitary, order K , M / 2 × M / 2 matrix W 0 ' ( z ) such that

H p ( z ) = 1 2 I D - J D J W 0 ' ( z ) 0 0 ( W 0 ' ) R ( z ) I 0 0 J P .

In this case H p ( z ) is determined by precisely ( M / 2 - 1 ) L + M / 2 2 parameters where L K is the McMillan degree of W 0 ' ( z ) .

Characterization of unitary H p ( z ) — linear phase and ps symmetry

From the previous result we have the following result:

Theorem 54 H p ( z ) of order K forms a unitary filter bank with linear-phase and PS filters iff there exists a unitary, order K , M / 2 × M / 2 matrix W 0 ' ( z ) such that

H p ( z ) = 1 2 I D - J J D W 0 ' ( z ) 0 0 ( W 0 ' ) R ( z ) I 0 0 J P .

H p is determined by precisely ( M / 2 - 1 ) L + M / 2 2 parameters where L K is the McMillan degree of W 0 ' ( z ) .

Notice that Theorems  "Characterization of Unitary H p (z) — PS Symmetry" through Theorem  "Characterization of Unitary H p (z) — Linear Phase and PS Symmetry" give a completeness characterization for unitary filter banks with thesymmetries in question (and the appropriate length restrictions on the filters). However, ifone requires only the matrices W 0 ' ( z ) and W 1 ' ( z ) in the above theorems to be invertible on the unit circle (and notunitary), then the above results gives a method to generate nonunitary PR filter banks with the symmetries considered. Notice however, thatin the nonunitary case this is not a complete parameterization of all such filter banks.

Linear-phase wavelet tight frames

A necessary and sufficient condition for a unitary (FIR) filter bank to give rise to a compactly supported wavelet tight frame (WTF) isthat the lowpass filter h 0 in the filter bank satisfies the linear constraint [link]

n h 0 ( n ) = M .

We now examine and characterize how H p ( z ) for unitary filter banks with symmetries can be constrained to give rise to wavelet tight frames (WTFs).First consider the case of PS symmetry in which case H p ( z ) is parameterized in [link] . We have a WTF iff

first row of H p ( z ) z = 1 = 1 / M ... 1 / M .

In [link] , since P permutes the columns, the first row is unaffected. Hence [link] is equivalent to the first rows of both W 0 ' ( z ) and W 1 ' ( z ) when z = 1 is given by

2 / M ... 2 / M .

This is precisely the condition to be satisfied by a WTF of multiplicity M / 2 . Therefore both W 0 ' ( z ) and W 1 ' ( z ) give rise to multiplicity M / 2 compactly supported WTFs. If the McMillan degree of W 0 ' ( z ) and W 1 ' ( z ) are L 0 and L 1 respectively, then they are parameterized respectively by M / 2 - 1 2 + ( M / 2 - 1 ) L 0 and M / 2 - 1 2 + ( M / 2 - 1 ) L 1 parameters. In summary, a WTF with PS symmetry can be explicitly parameterized by 2 M / 2 - 1 2 + ( M / 2 - 1 ) ( L 0 + L 1 ) parameters. Both L 0 and L 1 are greater than or equal to K .

Questions & Answers

how do you get the 2/50
Abba Reply
number of sport play by 50 student construct discrete data
Aminu Reply
width of the frangebany leaves on how to write a introduction
Theresa Reply
Solve the mean of variance
Veronica Reply
Step 1: Find the mean. To find the mean, add up all the scores, then divide them by the number of scores. ... Step 2: Find each score's deviation from the mean. ... Step 3: Square each deviation from the mean. ... Step 4: Find the sum of squares. ... Step 5: Divide the sum of squares by n – 1 or N.
kenneth
what is error
Yakuba Reply
Is mistake done to something
Vutshila
Hy
anas
hy
What is the life teble
anas
hy
Jibrin
statistics is the analyzing of data
Tajudeen Reply
what is statics?
Zelalem Reply
how do you calculate mean
Gloria Reply
diveving the sum if all values
Shaynaynay
let A1,A2 and A3 events be independent,show that (A1)^c, (A2)^c and (A3)^c are independent?
Fisaye Reply
what is statistics
Akhisani Reply
data collected all over the world
Shaynaynay
construct a less than and more than table
Imad Reply
The sample of 16 students is taken. The average age in the sample was 22 years with astandard deviation of 6 years. Construct a 95% confidence interval for the age of the population.
Aschalew Reply
Bhartdarshan' is an internet-based travel agency wherein customer can see videos of the cities they plant to visit. The number of hits daily is a normally distributed random variable with a mean of 10,000 and a standard deviation of 2,400 a. what is the probability of getting more than 12,000 hits? b. what is the probability of getting fewer than 9,000 hits?
Akshay Reply
Bhartdarshan'is an internet-based travel agency wherein customer can see videos of the cities they plan to visit. The number of hits daily is a normally distributed random variable with a mean of 10,000 and a standard deviation of 2,400. a. What is the probability of getting more than 12,000 hits
Akshay
1
Bright
Sorry i want to learn more about this question
Bright
Someone help
Bright
a= 0.20233 b=0.3384
Sufiyan
a
Shaynaynay
How do I interpret level of significance?
Mohd Reply
It depends on your business problem or in Machine Learning you could use ROC- AUC cruve to decide the threshold value
Shivam
how skewness and kurtosis are used in statistics
Owen Reply
yes what is it
Taneeya
Got questions? Join the online conversation and get instant answers!
Jobilize.com Reply

Get Jobilize Job Search Mobile App in your pocket Now!

Get it on Google Play Download on the App Store Now




Source:  OpenStax, Wavelets and wavelet transforms. OpenStax CNX. Aug 06, 2015 Download for free at https://legacy.cnx.org/content/col11454/1.6
Google Play and the Google Play logo are trademarks of Google Inc.

Notification Switch

Would you like to follow the 'Wavelets and wavelet transforms' conversation and receive update notifications?

Ask