Skip to main content

On the spectral norms of r-circulant matrices with the biperiodic Fibonacci and Lucas numbers

A Correction to this article was published on 22 February 2018

This article has been updated

Abstract

In this paper, we present new upper and lower bounds for the spectral norms of the r-circulant matrices \(Q=C_{r} ( (\frac{b}{a} )^{\frac{\xi (1)}{2}}q_{0}, (\frac{b}{a} )^{\frac{\xi(2)}{2}}q_{1}, (\frac {b}{a} )^{\frac{\xi(3)}{2}}q_{2}, \dots, (\frac{b}{a} )^{\frac{\xi(n)}{2}}q_{n-1} )\) and \(L=C_{r} ( (\frac {b}{a} )^{\frac{\xi(0)}{2}}l_{0}, (\frac{b}{a} )^{\frac{\xi (1)}{2}}l_{1}, (\frac{b}{a} )^{\frac{\xi(2)}{2}}l_{2}, \dots, (\frac{b}{a} )^{\frac{\xi(n-1)}{2}}l_{n-1} ) \) whose entries are the biperiodic Fibonacci and biperiodic Lucas numbers, respectively. Finally, we obtain lower and upper bounds for the spectral norms of Kronecker and Hadamard products of Q and L matrices.

Introduction

For \(n\in\mathbb{N}_{0}\), the Fibonacci and Lucas numbers are defined by \(F_{n+2} = F_{n+1} + F_{n}\) and \(L_{n+2} = L_{n+1} + L_{n}\) with the initial conditions \(F_{0}=0\), \(F_{1}=1\) and \(L_{0}=2\), \(L_{1}=1\), respectively. In recent years, there are several applications and generalizations of Fibonacci and Lucas numbers [112]. For example, Falcon and Plaza introduced the k-Fibonacci sequence by studying the recursive application of two geometrical transformations used in the well-known 4-triangle longest-edge (4TLE) partition [6]. Edson and Yayenie [3] presented a new generalization of the Fibonacci sequence: for \(n\in\mathbb{N}_{0}\),

$$ q_{0} = 0, \qquad q_{1}=1, \qquad q_{n+2} = \textstyle\begin{cases} a q_{n+1} + q_{n} & \text{if $n$ is even}, \\ b q_{n+1} + q_{n} & \text{if $n$ is odd}. \end{cases} $$
(1)

They also obtained an extended Binet formula for this sequence:

$$ q_{n} = \biggl( \frac{a^{1-\xi(n)}}{ab^{\lfloor\frac{n}{2} \rfloor}} \biggr) \frac{\alpha^{n}-\beta^{n}}{\alpha- \beta}, \quad n\in\mathbb{N}_{0}. $$
(2)

Afterward, Bilgici [4] defined generalized the Lucas sequence by the following recurrence relation: for \(n\in\mathbb{N}_{0}\),

$$ l_{0} = 2, \qquad l_{1}=a, \qquad l_{n+2} = \textstyle\begin{cases} b l_{n+1} + l_{n} & \text{if $n$ is even,} \\ a l_{n+1} + l_{n} & \text{if $n$ is odd} \end{cases} $$
(3)

and The Binet formula for this sequence is

$$ l_{n} = \biggl( \frac{a^{\xi(n)}}{ab^{\lfloor\frac{n+1}{2} \rfloor}} \biggr) \bigl( \alpha^{n}+\beta^{n}\bigr), \quad n\in\mathbb{N}_{0}. $$
(4)

In Eqs. (2) and (4), \(\alpha= \frac{ab + \sqrt {a^{2}b^{2}+4ab}}{2}\) and \(\beta= \frac{ab - \sqrt{a^{2}b^{2}+4ab}}{2}\) are the roots of the characteristic equation of \(x^{2}-abx-ab=0\), and \(\xi(n) = n- 2\lfloor\frac{n}{2}\rfloor\).

In recent years, there have been several studies on the norms, determinants, and inverses of circulant and r-circulant matrices whose entries are special integer sequences [1327]. For example, Shen and Cen [18] found upper and lower bounds for the spectral norms of r-circulant matrices in the forms \(A=C_{r}(F_{0}, F_{1}, F_{2}, \dots, F_{n-1})\) and \(B=C_{r}(L_{0}, L_{1}, L_{2}, \dots, L_{n-1})\). They also obtained some bounds for the spectral norms of Kronecker and Hadamard products of A and B. Afterward, Shen and Cen [19] gave the upper and lower bounds for the spectral norms of the matrices \(A=C_{r}(F_{k,0}, F_{k,1}, F_{k,2}, \dots , F_{k,n-1})\) and \(B=C_{r}(L_{k,0}, L_{k,1}, L_{k,2}, \dots, L_{k,n-1})\). They also presented some bounds for the spectral norms of Hadamard and Kronecker products of these matrices. Bahşi [16] studied the norms of r-circulant matrices \(H_{r} = \operatorname{Circr}(H_{0}^{(k)},H_{1}^{(k)},H_{2}^{(k)},\dots,H_{n-1}^{(k)})\) and \(\widehat{H_{r}} = \operatorname{Circr}(H_{k}^{(0)},H_{k}^{(1)},H_{k}^{(2)},\dots ,H_{k}^{(n-1)})\), where \(H_{n}^{(k)}\) denotes the nth hyperharmonic number of order r.

Inspired by these studies, in this paper, we compute spectral norms of r-circulant matrices whose entries are the biperiodic Fibonacci and biperiodic Lucas numbers. This study consists of three sections. The first one is the introduction. In the second section, we give some new theorems, corollaries, and some important results. We give a concise conclusion in the last section.

Definition 1.1

For any given \(c_{0},c_{1},c_{2},\dots,c_{n-1} \in\mathbb{C}\), the r-circulant matrix \(C_{r} = (c_{ij})_{n\times n}\) is defined by

$$ C_{r} = \begin{bmatrix} c_{0} & c_{1} & c_{2} & \ldots& c_{n-2} & c_{n-1} \\ rc_{n-1} & c_{0} & c_{1} & \ldots& c_{n-3} & c_{n-2}\\ rc_{n-2} & rc_{n-1} & c_{0} & \ldots& c_{n-4} & c_{n-3}\\ \vdots & \vdots & \vdots & \ddots & \vdots & \vdots\\ rc_{2} & rc_{3} & rc_{4} & \ldots& c_{0} & c_{1} \\ rc_{1} & rc_{2} & rc_{3} & \ldots& rc_{n-1} & c_{0} \end{bmatrix} . $$
(5)

It is clear that, for \(r=1\), \(C_{r}\) turns into a classical circulant matrix. Let us take any \(A=[a_{ij}] \in M_{n,n}(\mathbb{C})\). The Frobenius norm of the matrix A is defined by

$$\|A\|_{F} = \Biggl[ \sum_{i=1}^{m} \sum_{j=1}^{n} |a_{ij}|^{2} \Biggr]^{\frac{1}{2}}. $$

Also, the spectral norm of the matrix A is given by

$$\|A\|_{2} = \sqrt{\max_{1\leq i\leq n}\lambda_{i} \bigl(A^{H} A\bigr)}, $$

where \(\lambda_{i} (A^{H} A)\) are the eigenvalues of \(A^{H} A\) such that \(A^{H}\) is the conjugate transpose of A. Then, the well-known inequality [28] is given by

$$ \frac{1}{\sqrt{n}}\|A\|_{F} \leq\|A\|_{2} \leq\|A\|_{F}. $$
(6)

Lemma 1.2

[28]

For any matrices \(A,B\in M_{m,n}(\mathbb{C})\), we have

$$ \|A\circ B\|_{2} \leq\|A\|_{2}\|B\|_{2} , $$

where \(A\circ B\) is the Hadamard product of A and B.

Lemma 1.3

[28]

For any matrices \(A\in M_{m,n}(\mathbb{C})\) and \(B\in M_{p,q}(\mathbb {C})\), we have

$$ \|A\otimes B\|_{2} = \|A\|_{2} \|B\|_{2}, $$
(7)

where \(A\otimes B\) is the Kronecker product of A and B.

Lemma 1.4

[29]

For any matrices \(A=[a_{ij}] \in M_{n,n}(\mathbb{C})\) and \(B=[b_{ij}] \in M_{n,n}(\mathbb{C})\), we have

$$ \| A \circ B \|_{2} \leq r_{1}(A)c_{1}(B) , $$

where \(A \circ B\) is the Hadamard product, \(r_{1}(A) = \max_{1\leq i \leq n}\sqrt{\sum_{j=1}^{n}|a_{ij}|^{2}}\), and \(c_{1}(B) = \max_{1\leq j \leq n}\sqrt{\sum_{i=1}^{n}|b_{ij}|^{2}}\).

Theorem 1.5

[5]

For any positive integer n, we have

$$ \sum_{k=1}^{n} \biggl( \frac{b}{a} \biggr)^{\xi(k+1)}q_{k} ^{2} = \biggl( \frac{1}{a} \biggr) q_{n} q_{n+1}. $$
(8)

Main results

In this section, we first give the sum of squares of biperiodic Lucas numbers.

Theorem 2.1

For any positive integer m, we have

$$ \sum_{k=1}^{m} \biggl( \frac{b}{a} \biggr)^{\xi(k)}l_{k} ^{2} = \biggl( \frac {1}{a} \biggr)l_{m+1} l_{m} -2. $$
(9)

Proof

Using the Binet formula of the biperiodic Lucas numbers, we have

$$ \textstyle\begin{cases} l_{k} ^{2} = (\frac{\alpha^{2} }{ab} )^{k} + (\frac{\beta^{2} }{ab} )^{k} + 2 (-1)^{k} & \text{if $k$ is even,} \\ l_{k} ^{2} = ( \frac{a}{b} ) [ (\frac{\alpha^{2} }{ab} )^{k} + (\frac{\beta^{2} }{ab} )^{k} + 2 (-1)^{k} ] & \text{if $k$ is odd}. \end{cases} $$

Therefore, for any \(k\geq1\),

$$ \biggl( \frac{b}{a} \biggr)^{\xi(k)}l_{k} ^{2} = \biggl(\frac{\alpha^{2} }{ab} \biggr)^{k} + \biggl(\frac{\beta^{2} }{ab} \biggr)^{k} + 2 (-1)^{k}. $$

Using the properties \(ab(\alpha+1)=\alpha^{2}\) and \(ab(\beta+1)=\beta ^{2}\), we get

$$ \begin{aligned} \sum_{k=1}^{m} \biggl( \frac{b}{a} \biggr)^{\xi(k)}l_{k} ^{2} &= \sum_{k=1}^{m} \biggl(\frac{\alpha^{2} }{ab} \biggr)^{k} + \sum_{k=1}^{m} \biggl(\frac{\beta^{2} }{ab} \biggr)^{k} + \sum _{k=1}^{m} 2 (-1)^{k} \\ & = \frac{ (\frac{\alpha^{2}}{ab} )^{m+1} - (\frac{\alpha ^{2}}{ab} )}{ (\frac{\alpha^{2}}{ab} ) -1 } + \frac{ (\frac{\beta^{2}}{ab} )^{m+1} - (\frac{\beta^{2}}{ab} )}{ (\frac{\beta^{2}}{ab} ) -1 } + (-1)^{m} - 1 \\ & = \frac{1}{(ab)^{m+1}} \bigl[ \alpha^{2m+1} + \beta^{2m+1} - (-1)^{m} \bigr] -2. \end{aligned} $$

Observe that

$$ \biggl(\frac{1}{a} \biggr)l_{m} l_{m+1} = \frac{1}{(ab)^{m+1}} \bigl[ \alpha ^{2m+1} + \beta^{2m+1} - (-1)^{m} \bigr]. $$

Therefore,

$$ \sum_{k=1}^{m} \biggl( \frac{b}{a} \biggr)^{\xi(k)}l_{k} ^{2} = \biggl(\frac {1}{a} \biggr)l_{m+1} l_{m} -2. $$

 □

Theorem 2.2

Let \(Q=C_{r} ( (\frac{b}{a} )^{\frac{\xi(1)}{2}}q_{0}, (\frac{b}{a} )^{\frac{\xi(2)}{2}}q_{1}, (\frac{b}{a} )^{\frac{\xi(3)}{2}}q_{2}, \dots, (\frac{b}{a} )^{\frac{\xi (n)}{2}}q_{n-1} ) \) be an r-circulant matrix. Then, for \(r \in \mathbb{C}\), we have:

  • if \(|r| \geq1\), then

    $$\sqrt{\frac{q_{n} q_{n-1}}{a}} \leq\|Q\|_{2} \leq|r|\frac{q_{n} q_{n-1}}{a}; $$
  • if \(|r| < 1\), then

    $$|r|\sqrt{\frac{q_{n} q_{n-1}}{a}} \leq\|Q\|_{2} \leq\sqrt{(n-1) \frac{q_{n} q_{n-1}}{a}}. $$

Proof

The matrix Q is of the form

$$ Q = \begin{bmatrix} ( \frac{b}{a} )^{\frac{\xi(1)}{2}}q_{0} & ( \frac {b}{a} )^{\frac{\xi(2)}{2}}q_{1} & ( \frac{b}{a} )^{\frac {\xi{(3)}}{2}} q_{2} & \ldots& ( \frac{b}{a} )^{\frac{\xi {(n)}}{2}} q_{n-1} \\ r ( \frac{b}{a} )^{\frac{\xi{(n)}}{2}} q_{n-1} & ( \frac {b}{a} )^{\frac{\xi{(1)}}{2}} q_{0} & ( \frac{b}{a} )^{\frac{\xi{(2)}}{2}} q_{1} & \ldots& ( \frac{b}{a} )^{\frac {\xi{(n-1)}}{2}} q_{n-2}\\ r ( \frac{b}{a} )^{\frac{\xi{(n-1)}}{2}} q_{n-2} & r ( \frac{b}{a} )^{\frac{\xi{(n)}}{2}} q_{n-1} & ( \frac {b}{a} )^{\frac{\xi{(1)}}{2}} q_{0} & \ldots& ( \frac {b}{a} )^{\frac{\xi{(n-2)}}{2}} q_{n-3}\\ \vdots& \vdots& \vdots& \ddots& \vdots\\ r ( \frac{b}{a} )^{\frac{\xi{(2)}}{2}} q_{1} & r ( \frac {b}{a} )^{\frac{\xi{(3)}}{2}} q_{2} & r ( \frac{b}{a} )^{\frac{\xi{(4)}}{2}} q_{3} & \ldots& ( \frac{b}{a} )^{\frac {\xi{(1)}}{2}} q_{0} \end{bmatrix} . $$
(10)

Then we have

$$ \|Q\|_{F}^{2} = \sum_{k=0}^{n-1}(n-k) \biggl( \frac{b}{a} \biggr)^{\xi {(k+1)}} q_{k}^{2} + \sum_{k=1}^{n-1} k|r|^{2} \biggl( \frac{b}{a} \biggr)^{\xi{(k+1)}} q_{k}^{2}. $$

Hence, for \(|r|\geq1\), using Eq. (8), we obtain

$$ \begin{aligned} \|Q\|_{F}^{2} &\geq\sum _{k=0}^{n-1}(n-k) \biggl( \frac{b}{a} \biggr)^{\xi{(k+1)}} q_{k}^{2} + \sum _{k=1}^{n-1} k \biggl( \frac{b}{a} \biggr)^{\xi{(k+1)}} q_{k}^{2} \\ & = n \sum_{k=0}^{n-1} \biggl( \frac{b}{a} \biggr)^{\xi{(k+1)}} q_{k}^{2} \\ & = n \biggl( \frac{q_{n} q_{n-1}}{a} \biggr), \end{aligned} $$

that is,

$$ \frac{1}{\sqrt{n}}\|Q\|_{F} \geq\sqrt{\frac{q_{n} q_{n-1}}{a}}. $$

From (6) we have

$$ \|Q\|_{2} \geq\sqrt{\frac{q_{n} q_{n-1}}{a}}. $$

Now, for \(|r|\geq1\), we give an bound for the spectral norm of the matrix Q. Let the matrices B and C be

$$ B = \begin{bmatrix} r ( \frac{b}{a} )^{\frac{\xi{(1)}}{2}} q_{0} & 1 & 1 & \ldots& 1 \\ r ( \frac{b}{a} )^{\frac{\xi{(n)}}{2}} q_{n-1} & r ( \frac{b}{a} )^{\frac{\xi{(1)}}{2}} q_{0} & 1 & \ldots& 1\\ r ( \frac{b}{a} )^{\frac{\xi{(n-1)}}{2}} q_{n-2} & r ( \frac{b}{a} )^{\frac{\xi{(n)}}{2}} q_{n-1} & r ( \frac {b}{a} )^{\frac{\xi{(1)}}{2}} q_{0} & \ldots& 1\\ \vdots& \vdots& \vdots& \ddots& \vdots\\ r ( \frac{b}{a} )^{\frac{\xi{(2)}}{2}} q_{1} & r ( \frac {b}{a} )^{\frac{\xi{(3)}}{2}} q_{2} & r ( \frac{b}{a} )^{\frac{\xi{(4)}}{2}} q_{3} & \ldots& r ( \frac{b}{a} )^{\frac {\xi{(1)}}{2}} q_{0} \end{bmatrix} $$
(11)

and

$$ C = \begin{bmatrix} ( \frac{b}{a} )^{\frac{\xi{(1)}}{2}} q_{0} & ( \frac {b}{a} )^{\frac{\xi{(2)}}{2}} q_{1} & ( \frac{b}{a} )^{\frac{\xi{(3)}}{2}} q_{2} & \ldots& ( \frac{b}{a} )^{\frac {\xi{(n)}}{2}} q_{n-1} \\ 1 & ( \frac{b}{a} )^{\frac{\xi{(1)}}{2}} q_{0} & ( \frac {b}{a} )^{\frac{\xi{(2)}}{2}} q_{1} & \ldots& ( \frac {b}{a} )^{\frac{\xi{(n-1)}}{2}} q_{n-2}\\ 1 & 1 & ( \frac{b}{a} )^{\frac{\xi{(1)}}{2}} q_{0} & \ldots& ( \frac{b}{a} )^{\frac{\xi{(n-2)}}{2}} q_{n-3}\\ \vdots& \vdots& \vdots& \ddots& \vdots\\ 1 & 1 & 1 & \ldots& ( \frac{b}{a} )^{\frac{\xi{(1)}}{2}} q_{0} \end{bmatrix} , $$
(12)

so that \(Q = B \circ C\). Then we obtain

$$\begin{gathered} r_{1}(B) = \max_{1\leq i\leq n} \sqrt{\sum _{j=1}^{n}|b_{ij}|^{2} } = \sqrt {|r|^{2}\sum_{k=0}^{n-1} \biggl( \frac{b}{a} \biggr)^{\xi(k+1)} q_{k}^{2} } = |r|\sqrt{\frac{q_{n} q_{n-1}}{a}}, \\c_{1}(C) = \max_{1\leq j \leq n} \sqrt{\sum _{i=1}^{n}|c_{ij}|^{2} } = \sqrt {\sum_{k=0}^{n-1} \biggl( \frac{b}{a} \biggr)^{\xi(k+1)} q_{k}^{2} } = \sqrt {\frac{q_{n} q_{n-1}}{a}}. \end{gathered}$$

By Lemma 1.4 we have

$$ \|Q\|_{2} \leq r_{1}(B)c_{1}(C) = |r| \frac{q_{n} q_{n-1}}{a}. $$

Thus,

$$ \sqrt{\frac{q_{n} q_{n-1}}{a}} \leq\|Q\|_{2} \leq|r|\frac{q_{n} q_{n-1}}{a}. $$

On the other hand, for \(|r|<1\), we have

$$ \begin{aligned} \|Q\|_{F}^{2} &\geq\sum _{k=0}^{n-1}(n-k) |r|^{2} \biggl( \frac {b}{a} \biggr)^{\xi{(k+1)}} q_{k}^{2} + \sum _{k=1}^{n-1} k|r|^{2} \biggl( \frac{b}{a} \biggr)^{\xi{(k+1)}} q_{k}^{2} \\ &= n|r|^{2}\sum_{k=0}^{n-1} \biggl( \frac{b}{a} \biggr)^{\xi{(k+1)}} q_{k}^{2} \\ & = n|r|^{2} \biggl( \frac{q_{n} q_{n-1}}{a} \biggr), \end{aligned} $$

that is,

$$ \frac{1}{\sqrt{n}}\|Q\|_{F}\geq|r|\sqrt{\frac{q_{n} q_{n-1}}{a}}. $$

Thus, we obtain

$$ \|Q\|_{2} \geq|r|\sqrt{\frac{q_{n} q_{n-1}}{a}}. $$

Now, for \(|r|< 1\), we give an upper bound for the spectral norm of the matrix Q. Let the matrices D and E be

$$ D = \begin{bmatrix} ( \frac{b}{a} )^{\frac{\xi{(1)}}{2}} q_{0} & 1 & 1 & \ldots& 1 \\ r & ( \frac{b}{a} )^{\frac{\xi{(1)}}{2}} q_{0} & 1 & \ldots& 1\\ r & r & ( \frac{b}{a} )^{\frac{\xi{(1)}}{2}} q_{0} & \ldots& 1\\ \vdots& \vdots& \vdots& \ddots& \vdots\\ r & r & r & \ldots& ( \frac{b}{a} )^{\frac{\xi{(1)}}{2}} q_{0} \end{bmatrix} $$
(13)

and

$$ E = \begin{bmatrix} ( \frac{b}{a} )^{\frac{\xi{(1)}}{2}} q_{0} & ( \frac {b}{a} )^{\frac{\xi{(2)}}{2}} q_{1} & ( \frac{b}{a} )^{\frac{\xi{(3)}}{2}} q_{2} & \ldots& ( \frac{b}{a} )^{\frac {\xi{(n)}}{2}} q_{n-1} \\ ( \frac{b}{a} )^{\frac{\xi{(n)}}{2}} q_{n-1} & ( \frac {b}{a} )^{\frac{\xi{(1)}}{2}} q_{0} & ( \frac{b}{a} )^{\frac{\xi{(2)}}{2}} q_{1} & \ldots& ( \frac{b}{a} )^{\frac {\xi{(n-1)}}{2}} q_{n-2}\\ ( \frac{b}{a} )^{\frac{\xi{(n-1)}}{2}} q_{n-2} & ( \frac {b}{a} )^{\frac{\xi{(n)}}{2}} q_{n-1} & ( \frac{b}{a} )^{\frac{\xi{(1)}}{2}} q_{0} & \ldots& ( \frac{b}{a} )^{\frac {\xi{(n-2)}}{2}} q_{n-3}\\ \vdots& \vdots& \vdots& \ddots& \vdots\\ ( \frac{b}{a} )^{\frac{\xi{(2)}}{2}} q_{1} & ( \frac {b}{a} )^{\frac{\xi{(3)}}{2}} q_{2} & ( \frac{b}{a} )^{\frac{\xi{(4)}}{2}} q_{3} & \ldots& ( \frac{b}{a} )^{\frac {\xi{(1)}}{2}} q_{0} \end{bmatrix} , $$
(14)

so that \(Q=D \circ E\). Then we obtain

$$\begin{gathered} r_{1}(D) = \max_{1\leq i\leq n} \sqrt{\sum _{j=1}^{n}|d_{ij}|^{2} } = \sqrt { \biggl( \frac{b}{a} \biggr)^{\frac{\xi(1)}{2}} q_{0} ^{2} + (n-1)} = \sqrt{n-1}, \\c_{1}(E) = \max_{1\leq j \leq n} \sqrt{\sum _{i=1}^{n}|e_{ij}|^{2} } = \sqrt {\sum_{k=0}^{n-1} \biggl( \frac{b}{a} \biggr)^{\xi(k+1)} q_{k}^{2} } = \sqrt {\frac{q_{n} q_{n-1}}{a}}. \end{gathered}$$

By Lemma 1.4 we have

$$ \|Q\|_{2} \leq r_{1}(D)c_{1}(E) = \sqrt{(n-1) \frac{q_{n} q_{n-1}}{a}}. $$

Thus,

$$ |r|\sqrt{\frac{q_{n} q_{n-1}}{a}} \leq\|Q\|_{2} \leq\sqrt{(n-1) \frac {q_{n} q_{n-1}}{a}}. $$

 □

Theorem 2.3

Let \(L=C_{r} ( (\frac{b}{a} )^{\frac{\xi(0)}{2}}l_{0}, (\frac{b}{a} )^{\frac{\xi(1)}{2}}l_{1}, (\frac{b}{a} )^{\frac{\xi(2)}{2}}l_{2}, \dots, (\frac{b}{a} )^{\frac{\xi (n-1)}{2}}l_{n-1} ) \) be an r-circulant matrix. Then, for \(r \in\mathbb{C}\), we have:

  • if \(|r| \geq1\), then

    $$\sqrt{\frac{l_{n} l_{n-1}}{a}+2} \leq\|L\|_{2} \leq|r| \biggl( \frac{l_{n} l_{n-1}}{a}+2 \biggr); $$
  • if \(|r| < 1\), then

    $$|r|\sqrt{\frac{l_{n} l_{n-1}}{a}+2} \leq\|L\|_{2} \leq\sqrt{n \biggl( \frac {l_{n} l_{n-1}}{a}+2 \biggr)}. $$

Proof

The matrix L is of the form

$$ L = \begin{bmatrix} ( \frac{b}{a} )^{\frac{\xi(0)}{2}}l_{0} & ( \frac {b}{a} )^{\frac{\xi(1)}{2}}l_{1} & ( \frac{b}{a} )^{\frac {\xi{(2)}}{2}} l_{2} & \ldots& ( \frac{b}{a} )^{\frac{\xi {(n-1)}}{2}} l_{n-1} \\ r ( \frac{b}{a} )^{\frac{\xi{(n-1)}}{2}} l_{n-1} & ( \frac{b}{a} )^{\frac{\xi{(0)}}{2}} l_{0} & ( \frac{b}{a} )^{\frac{\xi{(1)}}{2}} l_{1} & \ldots& ( \frac{b}{a} )^{\frac {\xi{(n-2)}}{2}} l_{n-2}\\ r ( \frac{b}{a} )^{\frac{\xi{(n-2)}}{2}} l_{n-2} & r ( \frac{b}{a} )^{\frac{\xi{(n-1)}}{2}} l_{n-1} & ( \frac {b}{a} )^{\frac{\xi{(0)}}{2}} l_{0} & \ldots& ( \frac {b}{a} )^{\frac{\xi{(n-3)}}{2}} l_{n-3}\\ \vdots& \vdots& \vdots& \ddots& \vdots\\ r ( \frac{b}{a} )^{\frac{\xi{(1)}}{2}} l_{1} & r ( \frac {b}{a} )^{\frac{\xi{(2)}}{2}} l_{2} & r ( \frac{b}{a} )^{\frac{\xi{(3)}}{2}} l_{3} & \ldots& ( \frac{b}{a} )^{\frac {\xi{(0)}}{2}} l_{0} \end{bmatrix} . $$
(15)

Then we have

$$ \|L\|_{F}^{2} = \sum_{k=0}^{n-1}(n-k) \biggl( \frac{b}{a} \biggr)^{\xi {(k)}} l_{k}^{2} + \sum_{k=1}^{n-1} k|r|^{2} \biggl( \frac{b}{a} \biggr)^{\xi{(k)}} l_{k}^{2}. $$

Hence, for \(|r|\geq1\), using Eq. (9), we obtain

$$ \begin{aligned} \|L\|_{F}^{2} &\geq\sum _{k=0}^{n-1}(n-k) \biggl( \frac{b}{a} \biggr)^{\xi{(k)}} l_{k}^{2} + \sum _{k=1}^{n-1} k \biggl( \frac{b}{a} \biggr)^{\xi{(k)}} l_{k}^{2} \\ & = n \sum_{k=0}^{n-1} \biggl( \frac{b}{a} \biggr)^{\xi{(k)}} l_{k}^{2} \\ & = n \biggl( \frac{l_{n} l_{n-1}}{a} +2 \biggr), \end{aligned} $$

that is,

$$ \frac{1}{\sqrt{n}}\|L\|_{F} \geq\sqrt{\frac{l_{n} l_{n-1}}{a}+2}. $$

From (6) we have

$$ \|L\|_{2} \geq\sqrt{\frac{l_{n} l_{n-1}}{a}+2}. $$

Now, for \(|r|\geq1\), we give an upper bound for the spectral norm of the matrix L. Let the matrices F and H be

$$ F = \begin{bmatrix} r ( \frac{b}{a} )^{\frac{\xi{(0)}}{2}} l_{0} & 1 & 1 & \ldots& 1 \\ r ( \frac{b}{a} )^{\frac{\xi{(n-1)}}{2}} l_{n-1} & r ( \frac{b}{a} )^{\frac{\xi{(0)}}{2}} l_{0} & 1 & \ldots& 1\\ r ( \frac{b}{a} )^{\frac{\xi{(n-2)}}{2}} l_{n-2} & r ( \frac{b}{a} )^{\frac{\xi{(n-1)}}{2}} l_{n-1} & r ( \frac {b}{a} )^{\frac{\xi{(0)}}{2}} l_{0} & \ldots& 1\\ \vdots& \vdots& \vdots& \ddots& \vdots\\ r ( \frac{b}{a} )^{\frac{\xi{(1)}}{2}} l_{1} & r ( \frac {b}{a} )^{\frac{\xi{(2)}}{2}} l_{2} & r ( \frac{b}{a} )^{\frac{\xi{(3)}}{2}} l_{3} & \ldots& r ( \frac{b}{a} )^{\frac {\xi{(0)}}{2}} l_{0} \end{bmatrix} $$
(16)

and

$$ H = \begin{bmatrix} ( \frac{b}{a} )^{\frac{\xi{(0)}}{2}} l_{0} & ( \frac {b}{a} )^{\frac{\xi{(1)}}{2}} l_{1} & ( \frac{b}{a} )^{\frac{\xi{(2)}}{2}} l_{2} & \ldots& ( \frac{b}{a} )^{\frac {\xi{(n-1)}}{2}} l_{n-1} \\ 1 & ( \frac{b}{a} )^{\frac{\xi{(0)}}{2}} l_{0} & ( \frac {b}{a} )^{\frac{\xi{(1)}}{2}} l_{1} & \ldots& ( \frac {b}{a} )^{\frac{\xi{(n-2)}}{2}} l_{n-2}\\ 1 & 1 & ( \frac{b}{a} )^{\frac{\xi{(0)}}{2}} l_{0} & \ldots& ( \frac{b}{a} )^{\frac{\xi{(n-3)}}{2}} l_{n-3}\\ \vdots& \vdots& \vdots& \ddots& \vdots\\ 1 & 1 & 1 & \ldots& ( \frac{b}{a} )^{\frac{\xi{(0)}}{2}} l_{0} \end{bmatrix} , $$
(17)

so that \(L = F \circ H\). Then we obtain

$$\begin{gathered} r_{1}(F) = \max_{1\leq i\leq n} \sqrt{\sum _{j=1}^{n}|f_{ij}|^{2} } = \sqrt {|r|^{2}\sum_{k=0}^{n-1} \biggl( \frac{b}{a} \biggr)^{\xi(k)} l_{k}^{2} } = |r|\sqrt{\frac{l_{n} l_{n-1}}{a}+2}, \\c_{1}(H) = \max_{1\leq j \leq n} \sqrt{\sum _{i=1}^{n}|h_{ij}|^{2} } = \sqrt {\sum_{k=0}^{n-1} \biggl( \frac{b}{a} \biggr)^{\xi(k)} l_{k}^{2} } = \sqrt {\frac{l_{n} l_{n-1}}{a}+2}. \end{gathered}$$

By Lemma 1.4 we have

$$ \|L\|_{2} \leq r_{1}(F)c_{1}(H) = |r| \biggl( \frac{l_{n} l_{n-1}}{a} + 2 \biggr). $$

Thus,

$$ \sqrt{\frac{l_{n} l_{n-1}}{a}+2} \leq\|L\|_{2} \leq|r| \biggl( \frac{l_{n} l_{n-1}}{a} + 2 \biggr). $$

On the other hand, for \(|r|<1\), we have

$$ \begin{aligned} \|L\|_{F}^{2} &\geq\sum _{k=0}^{n-1}(n-k) |r|^{2} \biggl( \frac {b}{a} \biggr)^{\xi{(k)}} l_{k}^{2} + \sum _{k=1}^{n-1} k|r|^{2} \biggl( \frac{b}{a} \biggr)^{\xi{(k)}} l_{k}^{2} \\ &= n|r|^{2}\sum_{k=0}^{n-1} \biggl( \frac{b}{a} \biggr)^{\xi{(k)}} l_{k}^{2} \\ & = n|r|^{2} \biggl( \frac{l_{n} l_{n-1}}{a} +2 \biggr), \end{aligned} $$

that is,

$$ \frac{1}{\sqrt{n}}\|L\|_{F}\geq|r|\sqrt{\frac{l_{n} l_{n-1}}{a}+2}. $$

Thus, we obtain

$$ \|L\|_{2} \geq|r|\sqrt{\frac{l_{n} l_{n-1}}{a}+2}. $$

Now, for \(|r|< 1\), we give an upper bound for the spectral norm of the matrix L. Let the matrices G and K be

$$ G = \begin{bmatrix} 1 & 1 & 1 & \ldots& 1 \\ r & 1 & 1 & \ldots& 1\\ r & r & 1 & \ldots& 1\\ \vdots & \vdots & \vdots & \ddots & \vdots\\ r & r & r & \ldots& 1 \end{bmatrix} $$
(18)

and

$$ K = \begin{bmatrix} ( \frac{b}{a} )^{\frac{\xi(0)}{2}}l_{0} & ( \frac {b}{a} )^{\frac{\xi(1)}{2}}l_{1} & ( \frac{b}{a} )^{\frac {\xi{(2)}}{2}} l_{2} & \ldots& ( \frac{b}{a} )^{\frac{\xi {(n-1)}}{2}} l_{n-1} \\ ( \frac{b}{a} )^{\frac{\xi{(n-1)}}{2}} l_{n-1} & ( \frac {b}{a} )^{\frac{\xi{(0)}}{2}} l_{0} & ( \frac{b}{a} )^{\frac{\xi{(1)}}{2}} l_{1} & \ldots& ( \frac{b}{a} )^{\frac {\xi{(n-2)}}{2}} l_{n-2}\\ ( \frac{b}{a} )^{\frac{\xi{(n-2)}}{2}} l_{n-2} & ( \frac {b}{a} )^{\frac{\xi{(n-1)}}{2}} l_{n-1} & ( \frac{b}{a} )^{\frac{\xi{(0)}}{2}} l_{0} & \ldots& ( \frac{b}{a} )^{\frac {\xi{(n-3)}}{2}} l_{n-3}\\ \vdots& \vdots& \vdots& \ddots& \vdots\\ ( \frac{b}{a} )^{\frac{\xi{(1)}}{2}} l_{1} & ( \frac {b}{a} )^{\frac{\xi{(2)}}{2}} l_{2} & ( \frac{b}{a} )^{\frac{\xi{(3)}}{2}} l_{3} & \ldots& ( \frac{b}{a} )^{\frac {\xi{(0)}}{2}} l_{0} \end{bmatrix} , $$
(19)

so that \(L=G \circ K\). Then we obtain

$$\begin{gathered} r_{1}(G) = \max_{1\leq i\leq n} \sqrt{\sum _{j=1}^{n}|g_{ij}|^{2} } = \sqrt{n}, \\c_{1}(K) = \max_{1\leq j \leq n} \sqrt{\sum _{i=1}^{n}|k_{ij}|^{2} } = \sqrt {\sum_{k=0}^{n-1} \biggl( \frac{b}{a} \biggr)^{\xi(k)} l_{k}^{2} } = \sqrt {\frac{l_{n} l_{n-1}}{a}+2}. \end{gathered}$$

By Lemma 1.4 we have

$$ \|L\|_{2} \leq r_{1}(G)c_{1}(K) = \sqrt{n \biggl(\frac{l_{n} l_{n-1}}{a} +2 \biggr)}. $$

Thus,

$$ |r|\sqrt{\frac{l_{n} l_{n-1}}{a}+2} \leq\|L\|_{2} \leq\sqrt{n \biggl( \frac {l_{n} l_{n-1}}{a}+2 \biggr)}. $$

 □

Corollary 2.1

Let \(Q=C_{r} ( (\frac{b}{a} )^{\frac{\xi(1)}{2}}q_{0}, (\frac{b}{a} )^{\frac{\xi(2)}{2}}q_{1}, (\frac{b}{a} )^{\frac{\xi(3)}{2}}q_{2}, \dots, (\frac{b}{a} )^{\frac{\xi (n)}{2}}q_{n-1} )\) and \(L= C_{r} ( (\frac{b}{a} )^{\frac{\xi(0)}{2}}l_{0}, (\frac{b}{a} )^{\frac{\xi (1)}{2}}l_{1}, (\frac{b}{a} )^{\frac{\xi(2)}{2}}l_{2}, \dots, (\frac{b}{a} )^{\frac{\xi(n-1)}{2}}l_{n-1} ) \) be r-circulant matrices, where \(r\in\mathbb{C}\).

  1. (i)

    If \(|r|\geq1\), then

    $$ \|Q\circ L\|_{2} \leq|r|^{2} \frac{q_{n}q_{n-1}}{a} \biggl( \frac{l_{n} l_{n-1}}{a} + 2 \biggr). $$
  2. (ii)

    If \(|r| < 1\), then

    $$ \|Q\circ L\|_{2} \leq\sqrt{n(n-1)\frac{q_{n}q_{n-1}}{a} \biggl( \frac{l_{n} l_{n-1}}{a} +2 \biggr)}. $$

Proof

Since \(\|Q\circ L\|_{2} \leq\|Q\|_{2} \|L\|_{2}\), the proof is trivial by Theorems 2.2 and 2.3. □

Corollary 2.2

Let \(Q=C_{r} ( (\frac{b}{a} )^{\frac{\xi(1)}{2}}q_{0}, (\frac{b}{a} )^{\frac{\xi(2)}{2}}q_{1}, (\frac{b}{a} )^{\frac{\xi(3)}{2}}q_{2}, \dots, (\frac{b}{a} )^{\frac{\xi (n)}{2}}q_{n-1} )\) and \(L= C_{r} ( (\frac{b}{a} )^{\frac{\xi(0)}{2}}l_{0}, (\frac{b}{a} )^{\frac{\xi (1)}{2}}l_{1}, (\frac{b}{a} )^{\frac{\xi(2)}{2}}l_{2}, \dots, (\frac{b}{a} )^{\frac{\xi(n-1)}{2}}l_{n-1} ) \) be r-circulant matrices, where \(r\in\mathbb{C}\).

  1. (i)

    If \(|r|\geq1\), then

    $$ \|Q \otimes L\|_{2} \geq\sqrt{\frac{q_{n}q_{n-1}}{a} \biggl( \frac{l_{n} l_{n-1}}{a} + 2 \biggr)} $$

    and

    $$ \|Q\otimes L\|_{2} \leq|r|^{2} \frac{q_{n}q_{n-1}}{a} \biggl( \frac{l_{n} l_{n-1}}{a} + 2 \biggr)l. $$
  2. (ii)

    If \(|r|< 1\), then

    $$ \|Q \otimes L\|_{2} \geq|r|^{2} \sqrt{\frac{q_{n} q_{n-1}}{a} \biggl( \frac {l_{n} l_{n-1}}{a} + 2 \biggr)} $$

    and

    $$ \|Q\otimes L\|_{2} \leq\sqrt{n(n-1)\frac{q_{n}q_{n-1}}{a} \biggl( \frac {l_{n} l_{n-1}}{a} +2 \biggr)}. $$

Proof

Since \(\|Q\otimes L\|_{2} = \|Q\|_{2}\|L\|_{2}\), the proof is trivial by Theorems 2.2 and 2.3. □

Conclusion

In this paper, we obtain new upper and lower bounds for the spectral norms of the r-circulant matrices Q and L whose entries are the biperiodic Fibonacci and biperiodic Lucas numbers. This study can be reduced to various studies for the specific values of a and b in the literature. For example, if \(a=b=r=1\), \(a=b=1\), and \(a=b=k\) in Q and L, our results reduce to the studies [13, 18], and [19], respectively. Since this study is a generalization of these studies, it contributes to the literature by providing essential information on the spectral norms of r-circulant matrices.

Change history

  • 22 February 2018

    In the publication of this article (Köme and Yazlik in J. Inequal. Appl. 2017(1):192), there are a few errors.

References

  1. Koshy, T: Fibonacci and Lucas Numbers with Applications. Wiley, New York (2011)

    MATH  Google Scholar 

  2. Yazlik, Y, Taskara, N: A note on generalized k-Horadam sequence. Comput. Math. Appl. 63(1), 36-41 (2012)

    MathSciNet  Article  MATH  Google Scholar 

  3. Edson, M, Yayenie, O: A new generalization of Fibonacci sequence & extended Binet’s formula. Integers 9(6), 639-654 (2009)

    MathSciNet  Article  MATH  Google Scholar 

  4. Bilgici, G: Two generalizations of Lucas sequence. Appl. Math. Comput. 245, 526-538 (2014)

    MathSciNet  MATH  Google Scholar 

  5. Yayenie, O: New identities for generalized Fibonacci sequences and new generalization of Lucas sequences. Southeast Asian Bull. Math. 36(5), 739-752 (2012)

    MathSciNet  MATH  Google Scholar 

  6. Falcón, S, Plaza, Á: The k-Fibonacci sequence and the Pascal 2-triangle. Chaos Solitons Fractals 33(1), 38-49 (2007)

    MathSciNet  Article  MATH  Google Scholar 

  7. Falcon, S: On the k-Lucas numbers. Int. J. Contemp. Math. Sci. 6(21), 1039-1050 (2011)

    MathSciNet  MATH  Google Scholar 

  8. Yazlık, Y, Köme, C, Madhusudanan, V: A new generalization of Fibonacci and Lucas p-numbers. J. Comput. Anal. Appl. 25(4), 657-669 (2018)

    MathSciNet  Google Scholar 

  9. Kocer, EG, Tuglu, N, Stakhov, A: On the m-extension of the Fibonacci and Lucas p-numbers. Chaos Solitons Fractals 40(4), 1890-1906 (2009)

    MathSciNet  Article  MATH  Google Scholar 

  10. Kilic, E: The Binet formula, sums and representations of generalized Fibonacci p-numbers. Eur. J. Comb. 29(3), 701-711 (2008)

    MathSciNet  Article  MATH  Google Scholar 

  11. Stakhov, A, Rozin, B: Theory of Binet formulas for Fibonacci and Lucas p-numbers. Chaos Solitons Fractals 27(5), 1162-1177 (2006)

    MathSciNet  Article  MATH  Google Scholar 

  12. Uslu, K, Taskara, N, Kose, H: The generalized k-Fibonacci and k-Lucas numbers. Ars Comb. 99, 25-32 (2011)

    MathSciNet  MATH  Google Scholar 

  13. Solak, S: On the norms of circulant matrices with the Fibonacci and Lucas numbers. Appl. Math. Comput. 160(1), 125-132 (2005)

    MathSciNet  MATH  Google Scholar 

  14. Bahsi, M, Solak, S: On the norms of r-circulant matrices with the hyper-Fibonacci and Lucas numbers. J. Math. Inequal. 8(4), 693-705 (2014)

    MathSciNet  Article  MATH  Google Scholar 

  15. Bahsi, M: On the norms of circulant matrices with the generalized Fibonacci and Lucas numbers. TWMS J. Pure Appl. Math. 6(1), 84-92 (2015)

    MathSciNet  MATH  Google Scholar 

  16. Bahsi, M: On the norms of r-circulant matrices with the hyperharmonic numbers. J. Math. Inequal. 10(2), 445-458 (2016)

    MathSciNet  Article  MATH  Google Scholar 

  17. Kocer, EG, Mansour, T, Tuglu, N: Norms of circulant and semicirculant matrices with Horadam’s numbers. Ars Comb. 85, 353-359 (2007)

    MathSciNet  MATH  Google Scholar 

  18. Shen, S, Cen, J: On the bounds for the norms of r-circulant matrices with the Fibonacci and Lucas numbers. Appl. Math. Comput. 216(10), 2891-2897 (2010)

    MathSciNet  MATH  Google Scholar 

  19. Shen, S, Cen, J: On the spectral norms of r-circulant matrices with the k-Fibonacci and k-Lucas numbers. Int. J. Contemp. Math. Sci. 5(12), 569-578 (2010)

    MathSciNet  MATH  Google Scholar 

  20. Yazlik, Y, Taskara, N: On the norms of an r-circulant matrix with the generalized k-Horadam numbers. J. Inequal. Appl. 2013(1), Article ID 394 (2013)

    MathSciNet  Article  MATH  Google Scholar 

  21. Yazlik, Y, Taskara, N: Spectral norm, eigenvalues and determinant of circulant matrix involving the generalized k-Horadam numbers. Ars Comb. 104, 505-512 (2012)

    MathSciNet  MATH  Google Scholar 

  22. Yazlik, Y, Taskara, N: On the inverse of circulant matrix via generalized k-Horadam numbers. Appl. Math. Comput. 223, 191-196 (2013)

    MathSciNet  MATH  Google Scholar 

  23. Liu, L: On the spectrum and spectral norms of circulant matrices with generalized Horadam numbers entries. Int. J. Comput. Math. 2014, Article ID 795175 (2014)

    Article  Google Scholar 

  24. Kızılateş, C, Tuglu, N: On the bounds for the spectral norms of geometric circulant matrices. J. Inequal. Appl. 2016(1), Article ID 312 (2016)

    MathSciNet  Article  MATH  Google Scholar 

  25. He, C, Ma, J, Zhang, K, Wang, Z: The upper bound estimation on the spectral norm of r-circulant matrices with the Fibonacci and Lucas numbers. J. Inequal. Appl. 2015(1), Article ID 72 (2015)

    MathSciNet  Article  MATH  Google Scholar 

  26. Radičić, B: On k-circulant matrices (with geometric sequence). Quaest. Math. 39(1), 135-144 (2016)

    MathSciNet  Article  MATH  Google Scholar 

  27. Shen, S-Q, Cen, J-M: On the norms of circulant matrices with the \((k, h)\)-Fibonacci and \((k, h)\)-Lucas numbers. Int. J. Contemp. Math. Sci. 6(18), 887-894 (2011)

    MathSciNet  MATH  Google Scholar 

  28. Horn, RA, Johnson, CR: Topics in Matrix Analysis. Cambridge University Press, New York (1991)

    Book  MATH  Google Scholar 

  29. Mathias, R: The spectral norm of a nonnegative matrix. Linear Algebra Appl. 139, 269-284 (1990)

    MathSciNet  Article  MATH  Google Scholar 

Download references

Acknowledgements

The authors are grateful to the anonymous referees who have contributed to improve the quality of the paper. The authors declare that they have not received any financial support for this research.

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Yasin Yazlik.

Additional information

Competing interests

The authors declare that there are no competing interests with any individual or institution.

Authors’ contributions

Both authors contributed equally to the writing of this paper. Both authors read and approved the final manuscript.

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Köme, C., Yazlik, Y. On the spectral norms of r-circulant matrices with the biperiodic Fibonacci and Lucas numbers. J Inequal Appl 2017, 192 (2017). https://doi.org/10.1186/s13660-017-1466-0

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s13660-017-1466-0

Keywords

  • biperiodic Fibonacci number
  • biperiodic Lucas number
  • r-circulant matrix
  • norm