
2013/02/11 14:59:10 -0600
Development of ideas of vector expansion
Most people with technical backgrounds are familiar with the ideas of expansion vectors or basis vectors and of orthogonality; however, the related concepts of biorthogonality or of frames and tight frames are less familiar but also important. In the study of wavelet systems, we find that frames and tight frames are needed and should be understood, at least at a superficial level. One can find details in 12, 2, 1, 8. Another perhaps unfamiliar concept is that of an unconditional basis used by Donoho, Daubechies, and others 3, 10, 2 to explain why wavelets are good for signal compression, detection, and denoising 6, 5. In this chapter, we will very briefly define and discuss these ideas. At this point, you may want to skip these sections and perhaps refer to them later when they are specifically needed.
A set of vectors or functions fk(t)spans a vector space
F (or F is the Span of the set) if any element of
that space can be expressed as a linear combination of members of that
set, meaning: Given the finite or infinite set of functions fk(t), we
define
as the vector space with all elements of the
space of the form

with k∈Z and t,a∈R. An inner product
is usually defined for this space and is denoted 〈f(t),g(t)〉. A norm is defined and is denoted by
.
We say that the set fk(t) is a basis set or a basis
for a
given space F if the set of
in Equation 5.1 are unique
for any particular g(t)∈F. The set is called an orthogonal basis
if 〈fk(t),fℓ(t)〉=0 for all k≠ℓ. If we are in three dimensional Euclidean space, orthogonal
basis vectors are coordinate vectors that are at right (90o) angles to
each other. We say the set is an orthonormal basis if 〈fk(t),fℓ(t)〉=δ(k–ℓ) i.e. if, in addition to being
orthogonal, the basis vectors are normalized to unity norm: ∥fk(t)∥=1 for all k.
From these definitions it is clear that if we have an orthonormal basis, we can express any element in the vector space, g(t)∈F, written as Equation 5.1 by

since by taking the inner product of fk(t) with both sides of Equation 5.1, we get

where this inner product of the signal g(t) with the basis vector fk(t) “picks out" the corresponding coefficient ak. This expansion formulation or representation is extremely valuable. It expresses Equation 5.2 as an identity operator in the sense that the inner product operates on g(t) to produce a set of coefficients that, when used to linearly combine the basis vectors, gives back the original signal g(t). It is the foundation of Parseval's theorem which says the norm or energy can be partitioned in terms of the expansion coefficients ak. It is why the interpretation, storage, transmission, approximation, compression, and manipulation of the coefficients can be very useful. Indeed, Equation 5.2 is the form of all Fourier type methods.
Although the advantages of an orthonormal basis are clear, there are cases
where the basis system dictated by the problem is not and cannot (or
should not) be made orthogonal. For these cases, one can still have the
expression of Equation 5.1 and one similar to Equation 5.2 by using a dual basis set
whose elements are not orthogonal to
each other, but to the corresponding element of the expansion set

Because this type of “orthogonality" requires two sets of vectors, the expansion set and the dual set, the system is called biorthogonal. Using Equation 5.4 with the expansion in Equation 5.1 gives

Although a biorthogonal system is more complicated in that it requires, not only the original expansion set, but the finding, calculating, and storage of a dual set of vectors, it is very general and allows a larger class of expansions. There may, however, be greater numerical problems with a biorthogonal system if some of the basis vectors are strongly correlated.
The calculation of the expansion coefficients using an inner product in Equation 5.3 is called the analysis part of the complete process, and the calculation of the signal from the coefficients and expansion vectors in Equation 5.1 is called the synthesis part.
In finite dimensions, analysis and synthesis operations are simply matrix–vector multiplications. If the expansion vectors in Equation 5.1 are a basis, the synthesis matrix has these basis vectors as columns and the matrix is square and non singular. If the matrix is orthogonal, its rows and columns are orthogonal, its inverse is its transpose, and the identity operator is simply the matrix multiplied by its transpose. If it is not orthogonal, then the identity is the matrix multiplied by its inverse and the dual basis consists of the rows of the inverse. If the matrix is singular, then its columns are not independent and, therefore, do not form a basis.
Using a four dimensional space with matrices to illustrate the ideas of this
chapter, the synthesis formula
becomes

which can be compactly written in matrix form as

The synthesis or expansion Equation 5.1 or Equation 5.7 becomes

with the left-hand column vector g being the signal vector, the matrix F formed with the basis vectors fk as columns, and the right-hand vector a containing the four expansion coefficients, ak.
The equation for calculating the kth expansion coefficient in Equation 5.6 is

which can be written in vector form as

where each ak is an inner product of the kth row of
with g and analysis or coefficient
Equation 5.3 or Equation 5.10 becomes

which together are Equation 5.2 or

Therefore,

is how the dual basis in Equation 5.4 is found.
If the columns of F are orthogonal and normalized, then

This means the basis and dual basis are the same, and Equation 5.12 and Equation 5.13 become

and

which are both simpler and more numerically stable than Equation 5.13.
The discrete Fourier transform (DFT) is an interesting example of a finite dimensional Fourier transform with orthogonal basis vectors where matrix and vector techniques can be informative as to the DFT's characteristics and properties. That can be found developed in several signal processing books.
The Fourier Series is an excellent example of an infinite dimensional composition (synthesis) and decomposition (analysis). The expansion formula for an even function g(t) over 0<x<2π is

where the basis vectors (functions) are
and the expansion coefficients are obtained as

The basis vector set is easily seen to be orthonormal by verifying

These basis functions span an infinite dimensional vector space and the convergence of Equation 5.17 must be examined. Indeed, it is the robustness of that convergence that is discussed in this section under the topic of unconditional bases.
Another example of an infinite dimensional orthogonal basis is Shannon's sampling expansion 9. If f(t) is band limited, then

for a sampling interval
if the spectrum of f(t) is
zero for |ω|>W. In this case the basis functions are the sinc
functions with coefficients which are simply samples of the original
function. This means the inner product of a sinc basis function with a
bandlimited function will give a sample of that function. It is easy to
see that the sinc basis functions are orthogonal by taking the inner
product of two sinc functions which will sample one of them at the points
of value one or zero.
While the conditions for a set of functions being an orthonormal basis are sufficient for the representation in Equation 5.2 and the requirement of the set being a basis is sufficient for Equation 5.5, they are not necessary. To be a basis requires uniqueness of the coefficients. In other words it requires that the set be independent, meaning no element can be written as a linear combination of the others.
If the set of functions or vectors is dependent and yet does allow the expansion described in Equation 5.5, then the set is called a frame. Thus, a frame is a spanning set. The term frame comes from a definition that requires finite limits on an inequality bound 2, 12 of inner products.
If we want the coefficients in an expansion of a signal to represent the signal well, these coefficients should have certain properties. They are stated best in terms of energy and energy bounds. For an orthogonal basis, this takes the form of Parseval's theorem. To be a frame in a signal space, an expansion set ϕk(t) must satisfy

for some 0<A and B<∞ and for all signals g(t) in the space. Dividing Equation 5.22 by ∥g∥2 shows that A and B are bounds on the normalized energy of the inner products. They “frame" the normalized coefficient energy. If

then the expansion set is called a tight frame. This case gives

which is a generalized Parseval's theorem for tight frames. If A=B=1, the tight frame becomes an orthogonal basis. From this, it can be shown that for a tight frame 2

which is the same as the expansion using an orthonormal basis except for the A–1 term which is a measure of the redundancy in the expansion set.
If an expansion set is a non tight frame, there is no strict Parseval's theorem and the energy in the transform domain cannot be exactly partitioned. However, the closer A and B are, the better an approximate partitioning can be done. If A=B, we have a tight frame and the partitioning can be done exactly with Equation 5.24. Daubechies 2 shows that the tighter the frame bounds in Equation 5.22 are, the better the analysis and synthesis system is conditioned. In other words, if