Abstract
In this paper we consider algebras of functions that are constant on the sets of a partition. We describe the crossed product algebras of the mentioned algebras with \(\mathbb {Z}.\) We show that the function algebra is isomorphic to the algebra of all functions on some set. We also describe the commutant of the function algebra and finish by giving an example of piece-wise constant functions on a real line.
Access provided by Autonomous University of Puebla. Download conference paper PDF
Similar content being viewed by others
Keywords
1 Introduction
An important direction of investigation for any class of non-commutative algebras and rings, is the description of commutative subalgebras and commutative subrings. This is because such a description allows one to relate representation theory, non-commutative properties, graded structures, ideals and subalgebras, homological and other properties of non-commutative algebras to spectral theory, duality, algebraic geometry and topology naturally associated with commutative algebras. In representation theory, for example, semi-direct products or crossed products play a central role in the construction and classification of representations using the method of induced representations. When a non-commutative algebra is given, one looks for a subalgebra such that its representations can be studied and classified more easily and such that the whole algebra can be decomposed as a crossed product of this subalgebra by a suitable action.
When one has found a way to present a non-commutative algebra as a crossed product of a commutative subalgebra by some action on it, then it is important to know whether the subalgebra is maximal commutative, or if not, to find a maximal commutative subalgebra containing the given subalgebra. This maximality of a commutative subalgebra and related properties of the action are intimately related to the description and classification of representations of the non-commutative algebra.
Some work has been done in this direction [2, 4, 6] where the interplay between topological dynamics of the action on one had and the algebraic property of the commutative subalgebra in the \(C^*-\)crossed product algebra \(C(X)\rtimes \mathbb {Z}\) being maximal commutative on the other hand are considered. In [4], an explicit description of the (unique) maximal commutative subalgebra containing a subalgebra \(\mathcal {A}\) of \(\mathbb {C}^X\) is given. In [3], properties of commutative subrings and ideals in non-commutative algebraic crossed products by arbitrary groups are investigated and a description of the commutant of the base coefficient subring in the crossed product ring is given. More results on commutants in crossed products and dynamical systems can be found in [1, 5] and the references therein.
In this article, we take a slightly different approach. We consider algebras of functions that are constant on the sets of a partition, describe the crossed product algebras of the mentioned algebras with \(\mathbb {Z}\) and show that the function algebra is isomorphic to the algebra of all functions on some set. We also describe the commutant of the function algebra and finish by giving an example of piece-wise constant functions on a real line.
2 Definitions and a Preliminary Result
Let \(\mathcal {A}\) be any commutative algebra. Using the notation in [4], we let \(\psi :\mathcal {A}\rightarrow \mathcal {A}\) be any algebra automorphism on \(\mathcal {A}\) and define
It can be proved that \(\mathcal {A}\rtimes _{\psi }\mathbb {Z}\) is an associative \(\mathbb {C}-\)algebra with respect to point-wise addition, scalar multiplication and multiplication defined by twisted convolution , \(*\) as follows;
where \(\psi ^k\) denotes the \(k-\)fold composition of \(\psi \) with itself for positive k and we use the obvious definition for \(k\le 0\).
Definition 1
\(\mathcal {A}\rtimes _{\psi }\mathbb {Z}\) as described above is called the crossed product algebra of \(\mathcal {A}\) and \(\mathbb {Z}\) under \(\psi .\)
A useful and convenient way of working with \(\mathcal {A}\rtimes _{\psi }\mathbb {Z}\), is to write elements \(f,g\in \mathcal {A}\rtimes _{\psi }\mathbb {Z}\) in the form \(f=\sum _{n\in \mathbb Z}f_n\delta ^n\) and \(g=\sum _{n\in \mathbb Z}g_m\delta ^m\) where \(f_n=f(n),~g_m=g(m)\) and
Then addition and scalar multiplication are canonically defined and multiplication is determined by the relation
where \(m,n\in \mathbb {Z}\) and \(f_n,g_m\in \mathcal {A}.\)
Definition 2
By the commutant \(\mathcal {A}'\) of \(\mathcal {A}\) in \(\mathcal {A}\rtimes _{\psi }\mathbb {Z}\) we mean
It has been proven [4] that the commutant \(\mathcal {A}'\) is commutative and thus, is the unique maximal commutative subalgebra containing \(\mathcal {A}\). For any \(f,g\in \mathcal {A}\rtimes _{\psi }\mathbb {Z},\) that is, \(f=\sum _{n\in \mathbb Z}f_n\delta ^n\) and \(g=\sum _{m\in \mathbb Z}g_m\delta ^m,\) then \(fg=gf\) if and only if
Now let X be any set and \(\mathcal {A}\) an algebra of complex valued functions on X. Let \(\sigma :X\rightarrow X\) be any bijection such that \(\mathcal {A}\) is invariant under \(\sigma \) and \(\sigma ^{-1},\) that is for every \(h\in \mathcal {A},\ h\circ \sigma \in \mathcal {A}\) and \(h\circ \sigma ^{-1}\in \mathcal {A}.\) Then \((X, \sigma )\) is a discrete dynamical system and \(\sigma \) induces an automorphism \(\tilde{\sigma }:\mathcal {A}\rightarrow \mathcal {A}\) defined by,
Our goal is to describe the commutant of \(\mathcal {A}\) in the crossed product algebra \(\mathcal {A}\rtimes _{\tilde{\sigma }}\mathbb Z\) for the case where \(\mathcal {A}\) is the algebra of functions that are constant on the sets of a partition. First we have the following results.
Definition 3
For any nonzero \(n\in \mathbb Z,\) we set
The following theorem has been proven in [4].
Theorem 1
The unique maximal commutative subalgebra of \(\mathcal {A}\rtimes _{\tilde{\sigma }}\mathbb Z\) that contains \(\mathcal {A}\) is precisely the set of elements
We observe that since \(\tilde{\sigma }(f)=f\circ \sigma ^{-1},\) then
and hence for every \(n\in \mathbb {Z}\), \(\tilde{\sigma }^n(f)=f\circ \sigma ^{-n}.\) Therefore, by taking \(X=\mathbb R\) and \(\mathcal {A}\) as the algebra of constant functions on X we have: for every \(x\in X\) and every \(h\in \mathcal {A}\),
since h is a constant function. It follows that in this case \(Sep_{\mathcal {A}}^n(X)=\emptyset .\) Therefore in this case, \(\mathcal {A}'=\mathcal {A}\rtimes _{\tilde{\sigma }}\mathbb Z.\)
3 Algebra of Piece-Wise Constant Functions
Let X be any set, J a countable set and \(\mathbb {P}=\{X_j~:~j\in J\}\) be a partition of X; that is \(X=\cup _{r\in J}X_r\) where \(X_r \ne \emptyset \) and \(X_r\cap X_{r^{\prime }}=\emptyset \) if \(r\ne r^{\prime }.\)
Let \(\mathcal {A}\) be the algebra of piece-wise constant complex-valued functions on X. That is
Let \(\sigma :X\rightarrow X\) be a bijection on X. The lemma below gives the necessary and sufficient conditions for \((X,\sigma )\) to be a dynamical system.
Lemma 1
The following are equivalent.
-
1.
The algebra \(\mathcal {A}\) is invariant under \(\sigma \) and \(\sigma ^{-1}\).
-
2.
For every \(i\in J\) there exists \(j\in J\) such that \(\sigma (X_i)=X_j.\)
Proof
We recall that the algebra \(\mathcal {A}\) is invariant under \(\sigma \) if and only if for every \(h\in \mathcal {A},~h\circ \sigma \in \mathcal {A}.\)
Obviously, if for every \(i\in J\) there exists a unique \(j\in J\) such that \(\sigma (X_i)=X_j,\) then
Thus \(h\circ \sigma \in \mathcal {A}\).
Conversely, suppose \(\mathcal {A}\) is invariant under \(\sigma \) but 2. does not hold. Let \(x_1,x_2\in X_j\) and \(X_r, X_{r^{\prime }}\in \mathbb {P}\) such that \(\sigma (x_1)\in X_r\) and \(\sigma (x_2)\in X_{r^{\prime }}\). Let \(h:X\rightarrow \mathbb {C}\) be the function defined by
Then \(h\in \mathcal {A} .\) But \(h\circ \sigma (x_1)=1\) and \(h\circ \sigma (x_2)=0.\) Thus \(h\notin \mathcal {A}\), which contradicts the assumption. \(\square \)
The following lemma asserts that any bijection \(\sigma _2:X\rightarrow X\) that preserves the structure of a partition essentially produces the same algebra of functions.
Lemma 2
Let \(\mathbb {P}_1=\{X_j~:~j\in J\}\) and \(\mathbb {P}_2=\{Y_j~:~j\in J\}\) be partitions of the sets X and Y respectively, and let
and
Then \(\mathcal {A}_X\) is isomorphic to \(\mathcal {A}_Y.\)
Proof
Choose points \(x_i\in X\) and \(y_i\in Y\) such that \(x_i\in X_i\) if and only if \(y_i\in Y_i ~ \forall ~i\in J\) and let \(\mu :\mathcal {A}_X\rightarrow \mathcal {A}_Y\) be a function defined by
It is enough to prove that \(\mu \) is an algebra isomorphism.
-
Let \(f,g\in \mathcal {A}_X\) and let \(\alpha ,\beta \in \mathbb {C}.\) Then if \(y\in Y,\) then \(y\in Y_i\) for some \(i\in J,\) therefore,
$$\begin{aligned} \mu (\alpha f+\beta g)(y)=(\alpha f+\beta g)(x_i)\\ =\alpha f(x_i)+\beta g(x_i)\\ = \alpha \mu (f)(y)+\beta \mu (g)(y)\\ =[\alpha \mu (f)+\beta \mu (g)](y). \end{aligned}$$Therefore \(\mu \) is linear since y was arbitrary.
-
For every \(f,g\in \mathcal {A}_X\) and \(y\in Y~(y\in Y_i),\)
$$\begin{aligned} \mu (fg)(y) = (fg)(x_i)\\ = f(x_i)g(x_i)\\ =\mu (f)(y)\mu (g)(y)\\ = [\mu (f)\mu (g)](y). \end{aligned}$$Thus \(\mu \) is a multiplicative homomorphism.
-
Now, suppose \(f,g\in \mathcal {A}_X\) such that \(f\ne g.\) Then there exists \(i\in J\) such that \(f(x_i)\ne g(x_i),~x_i\in X_i.\) Therefore, if \(y\in Y_i,\)
$$\begin{aligned} \mu (f)(y)=f(x_i)\ne g(x_i)=\mu (g)y. \end{aligned}$$Therefore \(\mu \) is injective.
-
Finally, suppose \(h\in \mathcal {A}_Y\) and let \(f\in \mathcal {A}_X\) be defined by \(f(x)=h(y_i)\). If \(y\in Y,\) then \(y\in Y_i\) for some \(i\in J,\) and hence,
$$\begin{aligned} h(y)=h(y_i)=f(x)=f(x_i)=\mu (f)(y). \end{aligned}$$It follows that \(\mu \) is onto and hence an algebra isomorphism.
\(\square \)
Theorem 2
Let \(\mathbb {P}_1=\{X_j :j\in J\}\) and \(\mathbb {P}_2=\{Y_j :j\in J\}\) be partitions of two sets X and Y and \(\mathcal {A}_X\) and \(\mathcal {A}_Y\) be algebras of functions that are constant on the sets of the partitions \(\mathbb {P}_1\) and \(\mathbb {P}_2\) respectively. Let \(\sigma _1 :X \rightarrow X\) and \(\sigma _2 :Y \rightarrow Y\) be bijections such that \(\mathcal {A}_X\) is invariant under \(\sigma _1~(and~\sigma _1^{-1})\) and \(\mathcal {A}_Y\) is invariant under \(\sigma _2~(and~\sigma _2^{-1})\) and that \(\sigma _1(X_i)=X_j\) whenever \(\sigma _2(Y_i)=Y_j\) for all \(i, j \in J\). Suppose \(\tilde{\sigma _1}:\mathcal {A}_X \rightarrow \mathcal {A}_X\) is the automorphism on \(\mathcal {A}_X\) induced by \(\sigma _1\), and \(\tilde{\sigma _2}:\mathcal {A}_Y \rightarrow \mathcal {A}_Y\) is the automorphism on \(\mathcal {A}_Y\) induced by \(\sigma _2\). Then
where \(\mu \) is given by (4). Moreover, for every \(n\in \mathbb {Z},\)
Proof
Let \(y\in X\) such that \(y\in Y_i\) for some \(i\in J.\) Then for every \(f\in \mathcal {A},\)
Since y is arbitrary, we have
for every \(f\in \mathcal {A}.\) And since f is arbitrary,
Now from (5), we have
Therefore the relation (6) holds for \(n=2.\)
Now suppose the relation (6) holds for k. Then:
Therefore, from the induction principle,
\(\square \)
Remark 1
From Theorem 2 above, we get two nice results. The first is that if \(\mathbb {P}_1=\mathbb {P}_2\) are partitions of X and \(\sigma _1,\sigma _2:X\rightarrow X\) are bijections on X which preserve the structure of the partition, they will give rise to the same automorphism. That is, suppose \(\mathbb {P}_1=\{X_j~:~j\in J\}\) is a partition of X and \(\sigma _1,\sigma _2:X\rightarrow X\) are bijections on X such that, if \(\sigma _1 (X_i)=X_j,\) then \(\sigma _2 (X_i)=X_j,\) for all \(i,j\in J.\) Let \(\tilde{\sigma }:\mathcal {A}\rightarrow \mathcal {A}\) be the automorphism on induced by \(\sigma ,\) that is, for every \(h\in \mathcal {A},\)
Then for every \(f\in \mathcal {A},\)
This is given by the fact that if \(\mathbb {P}_1=\mathbb {P}_2\), then in (5), we can take \(\mu =id.\)
The second is the following important theorem.
Theorem 3
Let \(\mathbb {P}_1=\{X_j :j\in J\}\) and \(\mathbb {P}_2=\{Y_j :j\in J\}\) be partitions of two sets X and Y and \(\mathcal {A}_X\) and \(\mathcal {A}_Y\) be algebras of functions that are constant on the sets of the partitions \(\mathbb {P}_1\) and \(\mathbb {P}_2\) respectively. Let \(\sigma _1 :X \rightarrow X\) and \(\sigma _2 :Y \rightarrow Y\) be bijections such that \(\mathcal {A}_X\) is invariant under \(\sigma _1~(and~\sigma _1^{-1})\) and \(\mathcal {A}_Y\) is invariant under \(\sigma _2~(and~\sigma _2^{-1})\) and that \(\sigma _1(X_i)=X_j\) whenever \(\sigma _2(Y_i)=Y_j\) for all \(i, j \in J\). Suppose \(\tilde{\sigma _1}:\mathcal {A}_X \rightarrow \mathcal {A}_X\) is the automorphism on \(\mathcal {A}_X\) induced by \(\sigma _1\), and \(\tilde{\sigma _2}:\mathcal {A}_Y \rightarrow \mathcal {A}_Y\) is the automorphism on \(\mathcal {A}_Y\) induced by \(\sigma _2\). Then the crossed product algebras \(\mathcal {A}\rtimes _{ \tilde{\sigma _1}}\mathbb {Z}\) and \(\mathcal {A}\rtimes _{ \tilde{\sigma _2}}\mathbb {Z}\) are isomorphic.
Proof
We need to construct the an isomorphism between the crossed product algebras \(\mathcal {A}_X\rtimes _{ \tilde{\sigma _1}}\mathbb {Z}\) and \(\mathcal {A}_Y\rtimes _{ \tilde{\sigma _1}}\mathbb {Z}.\) Using the notation in [4], we let \(f:=\sum _{n\in \mathbb {Z}}f_n \delta ^n\) be an element in \(\mathcal {A}_X\rtimes _{\tilde{\sigma _1}}\mathbb {Z}.\) Define a function \(\mu :\mathcal {A}_X\rtimes _{ \tilde{\sigma _1}}\mathbb {Z} \rightarrow \mathcal {A}_Y\rtimes _{ \tilde{\sigma _2}}\mathbb {Z}\) be defined by
where \(\mu \) is defined in (4). Then, since \(\mu \) is an algebra isomorphism, it is enough to prove that \(\tilde{\mu }\) is multiplicative. To this end, we let \(f:=\sum _{n\in \mathbb {Z}}f_n \delta _1^n\) and \(g:=\sum _{m\in \mathbb {Z}}g_m \delta _1^m\) be arbitrary elements in \(\mathcal {A}_X\rtimes _{\tilde{\sigma _1}}\mathbb {Z},\) then we prove that \(\tilde{\mu }\) is multiplicative on the generators \(f_n\delta _1^n\) and \(g_m\delta _1^m\) respectively. Using (1) we have
Therefore \(\tilde{\mu }\) is multiplicative on the generators \(f_n\delta ^n\) and since \(\mu \) is linear, it is multiplicative on the elements \(f=\sum _{n\in \mathbb {Z}}f_n\delta ^n \in \mathcal {A}_X\rtimes _{\tilde{\sigma }}\mathbb {Z}.\) \(\square \)
Remark 2
In Lemma 1 we proved the necessary and sufficient condition on a bijection \(\sigma :X\rightarrow X\) such that the algebra \(\mathcal {A}_X\) is invariant under \(\sigma ,\) that is, for every \(i\in J\) there exists \(j\in J\) such that \(\sigma (X_i)=X_j\) where the \(X_i\) form a partition for X. From this, it can be shown that \(\mathcal {A}\) is isomorphic to \(\mathbb {C}^J,\) where by \(\mathbb {C}^J\) we denote the space of complex sequences indexed by J. This can be done by constructing an isomorphism between \(\mathcal {A}_X\) and \(\mathbb {C}^J\) via \(\sigma \) as follows.
Let \(\tau : J\rightarrow J\) be a map such that \(\tau (i)=j\) is equivalent to \(\sigma (X_i)=X_j\) for all \(i,j\in J.\) Then \(\tau \) is a bijection that plays the same role as \(\sigma _2\) in Lemma 2. Therefore, using the same Lemma, we deduce that the algebra \(\mathcal {A}\) is isomorphic to \(\mathbb {C}^J\). In Theorem 3, we have shown a method of constructing an isomorphism between the crossed product algebras \(\mathcal {A}_X\rtimes _{ \tilde{\sigma _1}}\mathbb {Z}\) and \(\mathcal {A}_Y\rtimes _{ \tilde{\sigma _2}}\mathbb {Z}\), when \(\mathcal {A}_X\) and \(\mathcal {A}_Y\) are isomorphic. It follows that the crossed product algebra \(\mathcal {A}_X\rtimes _{ \tilde{\sigma _1}}\mathbb {Z}\) is isomorphic to \(\mathbb {C}^J\rtimes _{ \tilde{\tau }}\mathbb {Z}\), where \(\tilde{\tau }\) follows the same definition as \(\tilde{\sigma }\).
In the next section we describe the commutant of our algebra \(\mathcal {A}_X\) in the crossed product algebra \(\mathcal {A}_X\rtimes _{ \tilde{\sigma }}\mathbb {Z}.\)
3.1 Maximal Commutative Subalgebra
We take the same partition \(\mathbb {P}=\cup _{j\in J}X_j\) and a bijection \(\sigma :X\rightarrow X\) such that for all \(i\in J,\) there exists \(j\in J\) such that \(\sigma (X_i)=X_j.\) For \(k\in \mathbb {Z}_{> 0},\) let
According to Theorem 1, the unique maximal commutative subalgebra of \(\mathcal {A}\rtimes _{\tilde{\sigma }}\mathbb Z\) that contains \(\mathcal {A}\) is precisely the set of elements
where \(Sep_{\mathcal {A}}^n(X)\) is given by (2). We have the following theorem which gives the description of \(Sep_{\mathcal {A}}^n(X)\) in this case and is crucial in the description if the maximal commutative subalgebra.
Theorem 4
Let \(\sigma :X \rightarrow X\) be a bijection on X as given above, \(\tilde{\sigma }:\mathcal {A}_X\rightarrow \mathcal {A}_X\) be the automorphism on \(\mathcal {A}_X\) induced by \(\sigma \) and \(C_k\) be given by (8). Then for every \(n\in \mathbb Z,\)
where
Proof
-
1.
If \(n\equiv 0\pmod {k}\) and \(x\in X_j\in C_k\), the we can write \(n=mk\) for some \(m\in \mathbb Z.\) Then, since \(\sigma ^k(X_j)=X_j\) it follows that \(\sigma ^{-k}(X_j)=X_j\) and therefore for every \(h\in \mathcal {A},\)
$$\begin{aligned} \tilde{\sigma }^n(h)(x)=\tilde{\sigma }^{mk}(h)(x)=(h\circ \sigma ^{-mk})(x)=h(\sigma ^{-mk}(x))=h(x), \end{aligned}$$since x and \(\sigma ^{-mk}(x)\in X_j\) for all \(m\in \mathbb Z.\)
-
2.
If \(n\not \equiv 0\pmod {k},\) we can write \(n=mk+j\) where \(m,j\in \mathbb Z\) with \(1\le j< k.\) It follows that for every \(x\in X_j\in C_k,\)
$$\begin{aligned} \tilde{\sigma }^n(h)(x)=\tilde{\sigma }^{mk+j}(h)(x)\\ =(h\circ \sigma ^{-mk+j})(x)\\ =h(\sigma ^{-mk+j}(x))\\ =\tilde{\sigma }^j(h)(x). \end{aligned}$$But k is the smallest integer such that \(\sigma ^k(X_j)=X_j.\) Therefore since \(j<k\),
$$\begin{aligned} \tilde{\sigma }^j(h)(x)\ne h(x). \end{aligned}$$Hence
$$\begin{aligned} Sep_{\mathcal {A}}^n(X)= \{x\in X~|~\exists ~h\in \mathcal {A}:h(x)\ne \tilde{\sigma }^n(h)(x)\}\\ ={\left\{ \begin{array}{ll} \left\{ \cup _{j~:~X_j\notin C_k}X_j\right\} ~~~ \text{ if } n\equiv 0\pmod {k},\\ \left\{ \cup _{j~:~X_j\in C_k}X_j\right\} ~~~ \text{ if } n \not \equiv 0\pmod {k}, \end{array}\right. } \end{aligned}$$and if \(x\in C_{\infty },\) then obviously \(x\in Sep_{\mathcal {A}}^n\) for every \(n\ge 1\), or simply
$$\begin{aligned} Sep_{\mathcal {A}}^n(\mathbb R)=\bigcup _{k\not \mid n}C_k\cup C_{\infty }. \end{aligned}$$
From the above theorem, the description of the maximal commutative subalgebra in \(\mathcal {A}\rtimes _{\tilde{\sigma }}\mathbb {Z}\) can be done as follows.
Theorem 5
Let \(\mathcal {A}_X\) be the algebra of piece-wise constant functions \(f:X\rightarrow \mathbb {C}\), \(\sigma :X\rightarrow X\) any bijection on X, \(\tilde{\sigma }:\mathcal {A}_X\rightarrow \mathcal {A}_X\) the automorphism on \(\mathcal {A}_X\) induced by \(\sigma \) and \(C_k\) be as described above. Then the unique maximal commutative subalgebra of \(\mathcal {A}_X\rtimes _{\tilde{\sigma }}\mathbb Z\) that contains \(\mathcal {A}_X\) is given by
Proof
From (9) we have that the unique maximal commutative subalgebra of \(\mathcal {A}_X\rtimes _{\tilde{\sigma }}\mathbb Z\) that contains \(\mathcal {A}_X\) is precisely the set of elements
and from (2),
Combining the two results and using the definition of \(h_n\in \mathcal {A}_X\) as
we get
\(\square \)
It can be observed from the results in Theorem 4 that it is possible to have \(Sep_{\mathcal {A}}^n(X)=X\) for all \(n\in \mathbb {Z}.\) For example, suppose J is infinite and let \(\sigma :X \rightarrow X\) be a bijection such that \(\sigma (X_j)=X_{j+1}\) for every \(j\in J.\) Then it is easily seen that in in this case \(Sep_{\mathcal {A}}^n(X)=X.\) However, this is not possible if J is finite since in this case \(\sigma \) acts like a permutation on a finite group. In the following section, we treat one such a case. We let \(X=\mathbb {R}\) and \(\mathcal {A}_X\) be the algebra of piece-wise constant functions on \(\mathbb {R}\) with N fixed jump points, where \(N\ge 1\) is an integer. In order to work in the setting described before, we treat jump points as intervals of zero length. Then \(\mathbb {R}\) is partitioned into \(2N+1\) sub-intervals.
4 Algebra of Piece-Wise Constant Functions on the Real Line with N Fixed Jump Points
Let \(\mathcal {A}\) be the algebra of piece-wise constant functions \(f:\mathbb R\rightarrow \mathbb R\) with Nfixed jumps at points \(t_1,t_2,\ldots ,t_N.\) Partition \(\mathbb R\) into \(N+1\) intervals \(I_0,I_1,\ldots ,I_N\) where \(I_{\alpha }=]t_{\alpha },t_{\alpha +1}[\) with \(t_0=-\infty \) and \(t_{N+1}=\infty .\) By looking at jump points as intervals of zero length, we can write \(\mathbb {R}=\cup I_{\alpha }\) where \(I_{\alpha }\) is as described above for \(\alpha =0,1,\ldots N\) and \(I_{\alpha }=\{t_{\alpha } \}\) if \(\alpha >N.\) Then for every \(h\in \mathcal {A}\) we have
where \(\chi _{I_{\alpha }}\) is the characteristic function of \(I_{\alpha }.\) As in the preceding section, we let \(\sigma :\mathbb R \rightarrow \mathbb R\) be any bijection on \(\mathbb R\) and let \(\tilde{\sigma }:\mathcal {A}\rightarrow \mathcal {A}\) be the automorphism on \(\mathcal {A}\) induced by \(\sigma \). Then we have the following lemma which gives the necessary and sufficient conditions for \((\mathbb R, \sigma )\) to be a discrete dynamical system.
Lemma 3
The algebra \(\mathcal {A}\) is invariant under both \(\sigma \) and \(\sigma ^{-1}\) if and only if the following conditions hold.
-
1.
\(\sigma \) (and \(\sigma ^{-1}\)) maps the each jump point \(t_k,~k=1,\ldots ,N\) onto another jump point.
-
2.
\(\sigma \) maps every interval \(I_{\alpha },~\alpha =0,1,\ldots N\) bijectively onto any of the other intervals \(I_0,I_1 \ldots I_N.\)
Proof
Obviously, if the two conditions hold, then \(\mathcal {A}\) is invariant under \(\sigma \). So we suppose that \(\mathcal {A}\) is invariant under \(\sigma \) and prove that the two conditions must hold.
-
1.
Suppose \(\sigma (t_k)=t_0\notin \{t_1,t_2,\ldots ,t_N\}\) for some \(k\in \{1,2,\ldots ,N\}.\) Then, since \(\sigma \) is onto, there exists \(x_0\in \mathbb R\) such that \(\sigma (x_0)=t_k,\) that is, there exists a non jump point that is mapped onto a jump point. We show that this is not possible.
Let
$$\begin{aligned} h(x)={\left\{ \begin{array}{ll}1~~ \text{ if } ~~x=t_k,\\ 0~~ \text{ otherwise. } \end{array}\right. } \end{aligned}$$Then \(h\in \mathcal {A}.\) But
$$\begin{aligned} h\circ \sigma (x)={\left\{ \begin{array}{ll}1~~ \text{ if } ~~\sigma (x)=t_k,\\ 0~~ \text{ otherwise, } \end{array}\right. }={\left\{ \begin{array}{ll}1~~ \text{ if } ~~x=x_0,\\ 0~~ \text{ otherwise. } \end{array}\right. } \end{aligned}$$Therefore \(h\circ \sigma \notin \mathcal {A}\) which is a contradiction, implying that \(\sigma \) does not map a non jump point onto a jump point, proving the first condition.
-
2.
Consider the bijection \(\sigma :\mathbb R\rightarrow \mathbb R\) defined by
$$\begin{aligned} \sigma (x)={\left\{ \begin{array}{ll}x~~ \text{ if } ~~x\ne t_k^{'} \text{ or } t_k^{''},\\ t_k^{''}~~ \text{ if } x=t_k^{'},\\ t_k^{'}~~ \text{ if } x=t_k^{''}, \end{array}\right. } \end{aligned}$$(13)where \(t_k^{'}\in I_k\) and \(t_k^{''}\in I_{k+1}\) for some \(k\in \{1,2,\ldots , N\}.\) Then \(\sigma \) is a bijection that permutes the jump points. Let \(h\in \mathcal {A}.\) Then using (12) and for the \(\sigma \) in Eq. (13) above, we have:
$$\begin{aligned} \sigma (x)={\left\{ \begin{array}{ll}h(x)~~ \text{ if } ~~x\ne t_k^{'} \text{ or } t_k^{''},\\ a_{k+1}~~ \text{ if } x=t_k^{'},\\ a_k~~~~~ \text{ if } x=t_k^{''}. \end{array}\right. } \end{aligned}$$Therefore, \(h\circ \sigma \) has jumps at points \(t_1,\ldots , t_N,t_k^{'},t_k^{''}\) implying that \(h\circ \sigma \notin \mathcal {A}\).
\(\square \)
The following theorem gives the description of \(Sep_{\mathcal {A}}^n(\mathbb R)\) for any \(n\in \mathbb Z.\)
Theorem 6
Let \(\mathcal {A}\) be an algebra of piece-wise constant functions with N fixed jumps at points \(t_1,\ldots , t_N\), \(\sigma :\mathbb R \rightarrow \mathbb R\) be any bijection on \(\mathbb R\) such that \(\mathcal {A}\) is invariant under \(\sigma \) and let \(\tilde{\sigma }:\mathcal {A}\rightarrow \mathcal {A}\) be the automorphism on \(\mathcal {A}\) induced by \(\sigma \). Let
Then for every \(n\in \mathbb Z,\)
Proof
See Theorem 4 and observe that \(C_{\infty }=\emptyset \) in this case. \(\square \)
Example 1
Let \(\mathcal {A}\) be the algebra of piece-wise constant functions with \(4-\)fixed jump points at \(t_1,t_2,t_3,t_4.\) Partition \(\mathbb R\) into five subintervals \(I_0,\ldots , I_4\) where \(I_{\alpha }=]t_{\alpha },t_{\alpha +1}[\) with \(t_0=-\infty \) and \(t_5=\infty .\)
Let \(\sigma :\mathbb R\rightarrow \mathbb R\) be a bijection such that \(\sigma (I_0)=I_1,~\sigma (I_1)=I_2,~\sigma (I_2)=I_0, \sigma (I_3)=I_4\) and \(\sigma (I_4)=I_3.\) It follows that \(\sigma ^3 (I_0)=I_0,~\sigma ^3 (I_1)=I_1\) and \(\sigma ^3 (I_2)=I_2.\) But \(\sigma ^j (I_{\alpha })\ne I_{\alpha }\) for \(\alpha =0,1,2\) and \(1\le j<3.\)
Also \(\sigma ^2 (I_3)=I_3,~\sigma ^2 (I_4)=I_4\) but \(\sigma ^j (I_{\alpha })\ne I_{\alpha }\) if \(j\not \equiv 0\pmod {2}\) with \(\alpha =3,4.\) Therefore:
and
From these results we have the following theorem.
Theorem 7
Let \(\mathcal {A}\) be the algebra of piece-wise constant functions \(f:\mathbb R\rightarrow \mathbb R\) with N fixed jumps at points \(t_1,t_2,\ldots ,t_N.\) Partition \(\mathbb R\) into \(N+1\) intervals \(I_0,I_1,\ldots ,I_N\) where \(I_{\alpha }=]t_{\alpha },t_{\alpha +1}[\) with \(t_0=-\infty \) and \(t_{N+1}=\infty \) and \(I_M=\{t_{\alpha } \}\) for \(N+1\leqslant M \leqslant 2N\). Let \(\sigma :\mathbb R \rightarrow \mathbb R\) be any bijection on \(\mathbb R\) such that \(\mathcal {A}\) is invariant under \(\sigma \) and let \(\tilde{\sigma }:\mathcal {A}\rightarrow \mathcal {A}\) be the automorphism on \(\mathcal {A}\) induced by \(\sigma \). Let
Then the unique maximal commutative subalgebra of \(\mathcal {A}\rtimes _{\tilde{\sigma }}\mathbb Z\) that contains \(\mathcal {A}\) is given by
Proof
From (9) we have that the unique maximal commutative subalgebra of \(\mathcal {A}\rtimes _{\tilde{\sigma }}\mathbb Z\) that contains \(\mathcal {A}\) is precisely the set of elements
and from (11),
Combining the two results and using the definition of \(h_n\in \mathcal {A}\) as
we get
\(\square \)
5 Some Examples
In this section we give some examples of how our results hold for well known simple cases. We treat two cases of piece-wise constant functions on the real line; those with one fixed jump point and those with two fixed jump points.
5.1 Piece-Wise Constant Functions with One Jump Point
Let \(\mathcal {A}\) be the collection of all piece-wise constant functions on the real line with one fixed jump point \(t_0.\) Following the methods in the previous section \(\mathbb {R}\) is partitioned into three intervals \(I_0=(-\infty , t_0),\quad I_1=(t_0,\infty )\) and \(I_2=\{t_0\}.\) Then we can write \(h\in \mathcal {A}\) as
Let \(\sigma : \mathbb {R}\rightarrow \mathbb {R}\) be any bijection on \(\mathbb {R}\) and let \(\tilde{\sigma }\) be the automorphism on \(\mathcal {A}\) induced by \(\sigma \). Note that by the first part of Lemma 3, invariance of the algebra \(\mathcal {A}\) implies that \(\sigma (t_0)=t_0.\) It follows therefore that \(\sigma (I_0)=I_0\) or \(\sigma (I_0)=I_1.\) We treat these two cases below.
5.1.1 \(\sigma (I_0)=I_0\)
In this case (and by bijectivity of \(\sigma \)), we have that \(\sigma (I_1)=I_1\) and since \(\sigma (t_0)=t_0,\) then for every \(x\in \mathbb {R},~h\in \mathcal {A}\) and \(n\in \mathbb {Z}\)
since x and \(\sigma ^{-n}(x)\) will lie in the same interval. Therefore, all intervals \(I_{\alpha },~\alpha =0,1,2\) belong to \(C_1\) and hence
Therefore, the maximal commutative subalgebra will be given by
5.1.2 \(\sigma (I_0)=I_1\)
In this case (and by bijectivity of \(\sigma \)), we have that \(\sigma (I_1)=I_0\) and since \(\sigma (t_0)=t_0,\) then for every \(x\in \mathbb {R},~h\in \mathcal {A}\) and \(n\in \mathbb {Z}\text { such that }2\mid n\) we have
since x and \(\sigma ^{-n}(x)\) will lie in the same interval. And for odd n, \(\tilde{\sigma }^n(h)(x)=h(x)\) if and only if \(x=t_0.\) Therefore, we have,
and
Therefore,
Therefore, the maximal commutative subalgebra will be given by
5.2 Piece-Wise Constant Functions with Two Jump Points
Let \(\mathcal {A}\) be the collection of all piece-wise constant functions on the real line with two fixed jump points at \(t_0\) and \(t_1.\) Following the methods in the previous section \(\mathbb {R}\) is partitioned into intervals \(I_0=]-\infty , t_0[,\quad I_1=]t_0,t_1[\quad I_2=]t_1,\infty [,\quad I_3=\{t_0\}\) and \(I_4=\{t_1\}.\) Then we can write \(h\in \mathcal {A}\) as
Let \(\sigma : \mathbb {R}\rightarrow \mathbb {R}\) be any bijection on \(\mathbb {R}\) and let \(\tilde{\sigma }\) be the automorphism on \(\mathcal {A}\) induced by \(\sigma \). Note that by the first part of Lemma 3, invariance of the algebra \(\mathcal {A}\) implies that \(\sigma (t_0)=t_0\) (and \(\sigma (t_1)=t_1\)) or \(\sigma (t_0)=t_1\) (in which case \(\sigma (t_1)=t_0\)). Below we give a description for the maximal commutative subalgebra of \(\mathcal {A}\rtimes _{\tilde{\sigma }}\mathbb {Z}\) for different types of \(\sigma .\)
5.2.1 \(\sigma (I_{\alpha })=I_{\alpha }\) for all \(\alpha =0,\ldots , 4\)
This case is similar to the one in Sect. 5.1.1 in the sense that, for every \(x\in \mathbb {R},~h\in \mathcal {A}\) and \(n\in \mathbb {Z}\)
since x and \(\sigma ^{-n}(x)\) will lie in the same interval. Therefore, all intervals \(I_{\alpha },~\alpha =0,\ldots ,4\) belong to \(C_1\) and hence
Therefore, the maximal commutative subalgebra will be given by
5.2.2 \(\sigma (I_0)=I_1,~\sigma (I_1)=I_0\) and \(\sigma (I_{\alpha })=I_{\alpha },~\alpha =2,3,4\)
In this case (and by bijectivity of \(\sigma \)), we have that \(\sigma (I_1)=I_0\) and therefore for every \(x\in \mathbb {R},~h\in \mathcal {A}\) and \(n\in \mathbb {Z}\text { such that }2\mid n\) we have
since x and \(\sigma ^{-n}(x)\) will lie in the same interval. And for odd n, \(\tilde{\sigma }^n(h)(x)=h(x)\) if and only if \(x\in I_2\cup I_3\cup I_4.\) Therefore, we have,
and
Therefore,
It follows that for \(n\in \mathbb {Z}\) such that \(2\mid n\), the maximal commutative subalgebra will be given by
And for odd n, we have
Therefore, the commutant \(\mathcal {A}'\) is given by:
Similar results can be obtained for the following cases
-
1.
\(\sigma (I_0)=I_1,\quad \sigma (I_1)=I_0,\quad \sigma (I_3)=I_4,\quad \sigma (I_4)=I_3\) and \(\sigma (I_2)=I_2.\)
-
2.
\(\sigma (I_0)=I_2,\quad \sigma (I_2)=I_0\) and \(\sigma (I_{\alpha })=I_{\alpha },\quad \alpha =1,3,4.\)
-
3.
\(\sigma (I_0)=I_2,\quad \sigma (I_2)=I_0,\quad \sigma (I_3)=I_4\quad \sigma (I_4)=I_3\) and \(\sigma (I_1)=I_1.\)
-
4.
\(\sigma (I_1)=I_2,\quad \sigma (I_2)=I_1\) and \(\sigma (I_{\alpha })=I_{\alpha }\quad \alpha =0,3,4.\)
-
5.
\(\sigma (I_1)=I_2,\quad \sigma (I_2)=I_1,\quad \sigma (I_3)=I_4,\quad \sigma (I_4)=I_3\) and \(\sigma (I_0)=I_0.\)
Since in all these cases, \(\sigma ^2(I_{\alpha })=I_{\alpha },~\alpha =0,\ldots ,4.\)
5.2.3 \(\sigma (I_0)=I_1,~\sigma (I_1)=I_2,~\sigma (I_2)=I_0\) and \(\sigma (I_{\alpha })=I_{\alpha },~\alpha =3,4\)
In this case, using similar methods we have,
and
Therefore,
It follows that for \(n\in \mathbb {Z}\) such that \(3\mid n\), the maximal commutative subalgebra will be given by
If \(3\not \mid n,\) then
Therefore:
5.2.4 \(\sigma (I_0)=I_1,~\sigma (I_1)=I_2,~\sigma (I_2)=I_0\) and \(\sigma (I_{3})=I_4,~\sigma (I_4)=I_3\)
In this case, using similar methods we have,
and
Therefore,
It follows that for \(n\in \mathbb {Z}\) such that \(3\mid n\), the maximal commutative subalgebra will be given by
If \(2\mid n,\) then
and for all other values of \(n,~\mathcal {A}'=\mathcal {A}.\) Hence:
References
Carlsen, T.M., Silvestrov, S.D.: On the excel crossed product of topological covering maps. Acta Appl. Math. 108(3), 573–583 (2009)
Li, B.-R.: Introduction to Operator Algebras. World Scientific, Singapore - New Jersey - Hong Kong - London (1992)
Öinert, J., Silvestrov, S.D.: Commutativity and ideals in algebraic crossed products. J. Gen. Lie Theory Appl. 2(4), 287–302 (2008)
Svensson, C., Silvestrov, S.D., de Jeu, M.: Dynamical systems and commutants in crossed products. Int. J. Math. 18, 455–471 (2007)
Svensson, C., Silvestrov, S.D., de Jeu, M.: Dynamical systems associated with crossed products. Acta Appl. Math. 108(3), 547–559 (2009)
Tomiyama, J.: Invitation to \(C^*-\)Algebras and Topological Dynamics. World Scientific, Singapore - New Jersey - Hong Kong - London (1987)
Acknowledgements
This work was partially supported by the Swedish Sida Foundation - International Science Program. Alex Behakanira Tumwesigye thanks the Research environment MAM in Mathematics and Applied Mathematics, Division of Applied Mathematics, School of Education, Culture and Communication, Mälardalens University for providing an excellent environment for research and education.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2016 Springer International Publishing Switzerland
About this paper
Cite this paper
Richter, J., Silvestrov, S., Ssembatya, V., Tumwesigye, A.B. (2016). Crossed Product Algebras for Piece-Wise Constant Functions. In: Silvestrov, S., Rančić, M. (eds) Engineering Mathematics II. Springer Proceedings in Mathematics & Statistics, vol 179. Springer, Cham. https://doi.org/10.1007/978-3-319-42105-6_6
Download citation
DOI: https://doi.org/10.1007/978-3-319-42105-6_6
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-42104-9
Online ISBN: 978-3-319-42105-6
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)