New quantum codes from constacyclic codes over finite chain rings

Yongsheng Tang¹, Ting Yao¹, Heqian Xu¹, Xiaoshan Kai²

1.School of Mathematics and Statistics, Hefei Normal University, Hefei, 230601, Anhui, P.R.China;
2. School of Mathematics, Hefei University of Technology, Hefei, 230009, Anhui, P.R.China

Abstract  Let $R$ be the finite chain ring $\mathbb{F}_{p^{2m}}+{u}\mathbb{F}_{p^{2m}}$, where $\mathbb{F}_{p^{2m}}$ is the finite field with $p^{2m}$ elements, $p$ is a prime, $m$ is a non-negative integer and ${u}^{2}=0.$ In this paper, we firstly define a class of Gray maps, which changes the Hermitian self-orthogonal property of linear codes over $\mathbb{F}_{2^{2m}}+{u}\mathbb{F}_{2^{2m}}$ into the Hermitian self-orthogonal property of linear codes over $\mathbb{F}_{2^{2m}}$. Applying the Hermitian construction, a new class of $2^{m}$-ary quantum codes are obtained from Hermitian constacyclic self-orthogonal codes over $\mathbb{F}_{2^{2m}}+{u}\mathbb{F}_{2^{2m}}.$ We secondly define another class of maps, which changes the Hermitian self-orthogonal property of linear codes over $R$ into the trace self-orthogonal property of linear codes over $\mathbb{F}_{p^{2m}}$. Using the Symplectic construction, a new class of $p^{m}$-ary quantum codes are obtained from Hermitian constacyclic self-orthogonal codes over $R.$

keywords: quantum code; Symplectic construction; Hermitian construction; constacyclic code

1 Introduction

Quantum error-correcting codes provide detection or correction of the errors which occur in a noisy quantum communication channel. In [8], it is shown that it is equivalent finding additive codes over $\mathbb{F}_{4}$ which are self-orthogonal with respect to certain trace inner product to finding binary stabilizer quantum error-correcting codes. The results given in [8] are generalized to stabilizer quantum error-correcting codes over $\mathbb{F}_{q}$ in [21]. Consequently, constructions of quantum error-correcting codes with good parameters have been developed rapidly. Many classes of quantum codes with good parameters were constructed from cyclic codes or constacyclic codes over finite fields by using Calderbank-Shor-Steane (CSS) construction, Hermitian construction and Symplectic construction, respectively (see [1,9,19,20,22]).

Recently, many researchers have applied cyclic codes or constacyclic codes over finite rings to construct quantum codes with good parameters. In these studies, the finite rings divided into two classes in general. One is finite chain rings, the other is finite non-chain rings. We firstly introduce linear codes over finite chain rings have been used to construct quantum codes with good parameters. Quantum codes with good parameters have been constructed from cyclic codes over $\mathbb{F}_{2}+u\mathbb{F}_{2}$ with $u^{2}=0$ in [29]. Later, this construction method is generalized to cyclic codes over $\mathbb{F}_{4}+u\mathbb{F}_{4}$ with $u^{2}=0$ in [18], and $\mathbb{F}_{2^{m}}+u\mathbb{F}_{2^{m}}+\cdots+u^{k}\mathbb{F}_{2^{m}}$ with $u^{k+1}=0$ in [32]. Liu et al. obtained quantum codes with good parameters from linear codes over $\mathbb{F}_{p^{2m}}+u\mathbb{F}_{p^{2m}}$ with $u^{2}=0$ in [25]. Ding et al. obtained quantum codes with good parameters from constacyclic codes over $\mathbb{F}_{p^{m}}+u\mathbb{F}_{p^{m}}$ with $u^{2}=0$ in [12]. Tang et al. obtained quantum codes with good parameters from constacyclic codes over $\mathbb{F}_{2^{m}}+u\mathbb{F}_{2^{m}}$ with $u^{2}=0$ in [33] and [34], respectively.

We secondly introduce linear codes over finite non-chain rings have been used to construct quantum codes with good parameters. Qian obtained quantum codes with good parameters from cyclic codes over $\mathbb{F}_{2}+v\mathbb{F}_{2}$ with $v^{2}=v$ in [30]. Ashraf and Mohammad obtained quantum codes with good parameters from cyclic codes over $\mathbb{F}_{3}+v\mathbb{F}_{3}$ with $v^{2}=1$ in [2], and $\mathbb{F}_{p}+v\mathbb{F}_{p}$ with $v^{2}=v$ in [3], respectively. Sari and Siap obtained quantum codes with good parameters from cyclic codes over $\mathbb{F}_{p^{r}}+v\mathbb{F}_{p^{r}}+\cdots+v^{m-1}\mathbb{F}_{p^{r}}$ with $v^{m}=v$ in [31]. Gao et al. obtained quantum codes with good parameters from cyclic codes over $\mathbb{F}_{q}+v_{1}\mathbb{F}_{q}+\cdots+v_{r}\mathbb{F}_{q}$ with $v_{i}^{2}=v_{i},v_{i}v_{j}=v_{j}v_{i}=0$ in [15]. Dertli et al. obtained quantum codes with good parameters from cyclic codes over $\mathbb{F}_{2}+u\mathbb{F}_{2}+v\mathbb{F}_{2}+uv\mathbb{F}_{2}$, where $u^{2}=u,v^{2}=v,uv=vu$ in [10]. Ashraf and Mohammad obtained quantum codes with good parameters from cyclic codes over $\mathbb{F}_{q}+u\mathbb{F}_{q}+v\mathbb{F}_{q}+uv\mathbb{F}_{q}$ with $u^{2}=u,v^{2}=v,uv=vu$ in [4], and $\mathbb{F}_{p}[u,v]/\langle u^{2}-1,v^{3}-v,uv-vu\rangle$ with $u^{2}=1,v^{3}=v,uv=vu$ in [5], respectively. Bag et al. obtained quantum codes with good parameters from constacyclic codes over $\mathbb{F}_{p^{m}}+v\mathbb{F}_{p^{m}}$ with $v^{2}=1,$ $p$ is odd prime in [6]. Gao et al. obtained quantum codes with good parameters from constacyclic codes over $\mathbb{F}_{p}+u\mathbb{F}_{p}$ with $u^{2}=1$ in [14]. Wang et al. obtained quantum codes with good parameters from constacyclic codes over $\mathbb{F}_{p}+v\mathbb{F}_{p}$ with $v^{2}=v$ in [36]. Ma et al. obtained quantum codes with good parameters from constacyclic codes over $\mathbb{F}_{q}+v\mathbb{F}_{q}+v^{2}\mathbb{F}_{q}$ with $v^{3}=v$ in [26]. They also obtained quantum codes with good parameters from constacyclic codes over $\mathbb{F}_{q}[u,v]/\langle u^{2}-1,v^{2}-v,uv-vu\rangle$ with $u^{2}=1,v^{2}=v,uv=vu$ in [27]. Bag et al. obtained quantum codes with good parameters from a class of constacyclic codes over $\mathbb{F}_{p}[u]/\langle u^{k+1}-u\rangle$ with $p$ is odd prime, $k$ is an integer and $\textrm{gcd}(p,k)=1$ in [11], and $\mathbb{F}_{q^{2}}+u\mathbb{F}_{q^{2}}$ with $u^{2}=\omega^{2}$ in [7], respectively. Wang et al. obtained quantum codes with good parameters from constacyclic codes over $\mathbb{F}_{q^{2}}+v\mathbb{F}_{q^{2}}$ with $v^{2}=v$ in [37], and $\mathbb{F}_{q^{2}}+v_{1}\mathbb{F}_{q^{2}}+\cdots+v_{r}\mathbb{F}_{q^{2}}$ with $v_{i}^{2}=v_{i},v_{i}v_{j}=v_{j}v_{i}=0$ in [38], respectively. Ji and Zhang obtained quantum codes with good parameters from constacyclic codes over $\mathbb{F}_{q}[u_{1},u_{2},\cdots,u_{k}]/\langle u_{i}^{3}-u_{i},u_{i}u_{j}-u_{j}u_{i}\rangle$ in [16].

Motivated by these excellent works, we consider two constructions of quantum codes from Hermitian $\alpha(1+u)$-constacyclic self-orthogonal codes over the finite chain ring $\mathbb{F}_{p^{2m}}+{u}\mathbb{F}_{p^{2m}}$ with $u^{2}=0$. By applying the Hermitian construction and the Symplectic construction, respectively, we obtain a new class of $2^{m}$-ary quantum codes and a new class of $p^{m}$-ary quantum codes from Hermitian $\alpha(1+u)$-constacyclic self-orthogonal codes over $\mathbb{F}_{2^{2m}}+u\mathbb{F}_{2^{2m}}$ and $\mathbb{F}_{p^{2m}}+u\mathbb{F}_{p^{2m}},$ respectively. Compared our obtained quantum codes with the existing codes available in Ref.[13], we find that the parameters of our obtained quantum codes are optimal at present. Moreover, we find that some our obtained quantum codes are satisfying $2d=n-k.$

The rest of this paper is organized as follows. In Section 2, we introduce some basic definitions and notations of linear codes over $\mathbb{F}_{p^{2m}}+{u}\mathbb{F}_{p^{2m}}$. A class of Gray maps $\Phi_{M}$ is defined in Section 3. In Section 4, we construct a new class of $2^{m}$-ary quantum codes based on Hermitian $\alpha(1+u)$-constacyclic self-orthogonal codes over $\mathbb{F}_{2^{2m}}+u\mathbb{F}_{2^{2m}}$ by using the Hermitian construction. Some optimal $p^{m}$-ary quantum codes from Hermitian $\alpha(1+u)$-constacyclic self-orthogonal codes over $\mathbb{F}_{p^{2m}}+u\mathbb{F}_{p^{2m}}$ via the Symplectic construction are obtained in Section 5. Section 6 concludes the paper.

2 Preliminaries

Let $\mathbb{F}_{p^{2m}}$ be the finite field with $p^{2m}$ elements. Let $R$ be the polynomial residue ring $\mathbb{F}_{p^{2m}}+u\mathbb{F}_{p^{2m}}$ with $u^{2}=0$. The ring $R$ is a finite chain ring with a maximal ideal $\langle u\rangle$. The units of $R$ are the elements $\{a+ub\mid a\neq 0\}$ and the residue field is $R/uR\simeq\mathbb{F}_{p^{2m}}$. A code of length $N$ over $R$ is a nonempty subset of $R^{N}$, and a code is linear over $R$ if it is an $R$-submodule of $R^{N}$. Two codes are permutation equivalent if one can be obtained from the other by permuting the coordinates. Any code over $R$ is permutation equivalent to a code $C$ with generator matrix of the form:

$$G=\left(\begin{array}[]{ccc}I_{k_{0}}&A&B\\ 0&uI_{k_{1}}&uD\end{array}\right),$$

where $B$ is a matrix over $R$ and $A,\ D$ are $\mathbb{F}_{p^{2m}}$ matrices. Then $C$ is an abelian group of type $\{k_{0},k_{1}\}$, $C$ contains $p^{4m(2k_{0}+k_{1})}$ codewords, and $C$ is a free $R$-module if and only if $k_{1}=0$.

For any element $c=a+ub\in R,$ where $a,b\in\mathbb{F}_{p^{2m}}$. The conjugation of $c=a+ub$ is defined by $\overline{c}=\overline{a}-u\overline{b},$ where $\overline{a}=a^{p^{m}},\overline{b}=b^{p^{m}}.$ Naturally, the conjugation of $\mathbf{c}=(c_{0},c_{1},\cdots,c_{N-1})\in R^{N}$ is defined as $\overline{\mathbf{c}}=(\bar{c}_{0},\bar{c}_{1},\cdots,\bar{c}_{N-1})$. For any $\mathbf{x}=(x_{0},x_{1},\cdots,x_{N-1})$ and $\mathbf{y}=(y_{0},y_{1},\cdots,y_{N-1})$, the Hermitian inner product of $\mathbf{x}$ and $\mathbf{y}$ is defined as

$\langle\mathbf{x},\mathbf{y}\rangle_{H}=x_{0}\bar{y}_{0}+x_{1}\bar{y}_{1}+\cdots+x_{N-1}\bar{y}_{N-1}$.

The Hermitian dual code $C^{\bot_{H}}$ of $C$ is define as

$C^{\bot_{H}}=\{\mathbf{x}\in R^{N}|\ \langle\mathbf{x},\mathbf{y}\rangle_{H}=0$, for all $\mathbf{y}\in C\}$.

A code $C$ is called Hermitian self-orthogonal if $C\subseteq C^{\bot_{H}}$, and Hermitian self-dual if $C^{\bot_{H}}=C$.

Similar to Proposition 1.2 in Ref.[35], we easily obtain the generator matrix of the Hermitian dual code of $C$.

Proposition 2.1. The Hermitian dual code $C^{\perp H}$ of the linear code $C$ of length $N$ over $R$ with generator matrix above has generator matrix

$$\left(\begin{array}[]{ccc}\overline{B}^{T}+\overline{D}^{T}\ \overline{A}^{T}&\overline{D}^{T}&I_{N-k_{0}-k_{1}}\\ u\overline{A}^{T}&uI_{k_{1}}&0\\ \end{array}\right),$$

where $\overline{A}^{T},\overline{B}^{T}$ and $\overline{D}^{T}$ denote the conjugate transposes of $A,B$ and $D$, respectively. Moreover, $C^{\perp H}$ is an abelian group of type $\{N-k_{0}-k_{1},k_{1}\}$ , and its size is $p^{2m(2N-2{k_{0}}-{k_{1}})}$.

Let $C$ be a code of length $N$ over $R$ and $P(C)$ be its polynomial representation, i.e.,

$$P(C)=\left\{\sum_{i=0}^{N-1}c_{i}x^{i}\mid(c_{0},c_{1},\ldots,c_{N-1})\in C\right\}.$$

Let $\sigma_{\lambda}$ be a map from $R^{N}$ to $R^{N}$ given by

$$\sigma_{\lambda}(c_{0},c_{1},\ldots,c_{N-1})=(\lambda c_{N-1},c_{0},\ldots,c_{N-2}),$$

where $\lambda\in R$. Then $C$ is said to be cyclic if $\sigma_{1}(C)=C$ and $\alpha(1+u)$-constacyclic if $\sigma_{\alpha(1+u)}(C)=C$. A code $C$ of length $N$ over $R$ is cyclic if and only if $P(C)$ is an ideal of $R[x]/{\langle x^{N}-1\rangle}$ and a code $C$ of length $N$ over $R$ is $\alpha(1+u)$-constacyclic if and only if $P(C)$ is an ideal of $R[x]/{\langle x^{N}-\alpha(1+u)\rangle}$. Throughout this paper, we assume that $\alpha(1+u)$-constacyclic codes over $R$ of length $N=p^{e}n,$ where $e$ is a non-negative integer and $n$ and $p$ are coprime.

3 A class of Gray maps over $R$

A class of Gray maps $\Phi_{M}$ from $R=\mathbb{F}_{p^{2m}}+{u}\mathbb{F}_{p^{2m}}$ to $\mathbb{F}_{p^{2m}}^{2}$ is defined as follows.

Definition 3.1 A map $\Phi_{M}$ on $R$ is defined as

$\Phi_{M}:R\rightarrow\mathbb{F}_{p^{2m}}^{2}$,

$r+uq\mapsto(q,r+q)M,$

where $M=$ $\begin{pmatrix}a\ &b\\ s\ &t\\ \end{pmatrix}\in GL_{2}(\mathbb{F}_{p^{2m}}),$ and $GL_{2}(\mathbb{F}_{p^{2m}})$ is the set of all $2\times 2$ invertible matrices over $\mathbb{F}_{p^{2m}}$.
The map $\Phi_{M}$ can be naturally extended to $R^{n}$ as follows:

$\Phi_{M}:R^{N}\rightarrow\mathbb{F}_{p^{2m}}^{2N}$,

$(c_{0},c_{1},\ldots,c_{N-1})\mapsto(aq_{0}+s(r_{0}+q_{0}),\ldots,aq_{N-1}+s(r_{N-1}+q_{N-1}),bq_{0}+t(r_{0}+q_{0}),\ldots,bq_{N-1}+t(r_{N-1}+q_{N-1}))$,

where $c_{i}=r_{i}+uq_{i},$ for $0\leq i\leq N-1$.

The Hamming weight of $\textbf{c}\in C$, denoted by $w_{H}(\textbf{c})$, is the number of its nonzero components. The Hamming distance $d_{H}(\textbf{c},\textbf{c}^{\prime})$ between two codewords c and $\textbf{c}^{\prime}$ is the Hamming weight of the codeword $\textbf{c}-\textbf{c}^{\prime}$. The Hamming distance $d_{H}(C)$ of $C$ is defined as $\min\{w_{H}(\textbf{c}-\textbf{c}^{\prime})|\textbf{c},\textbf{c}^{\prime}\in C,\textbf{c}\neq\textbf{c}^{\prime}\}$. That is, for any linear code, the Hamming distance $d_{H}(C)$ of $C$ is its minimum Hamming weight.

The Gray weight of any element $r+uq\in R$ is defined as the Hamming weight of $\Phi_{M}(r+uq)$, i.e., the number of nonzero components of $\Phi_{M}(r+uq)$. The Gray weight of a codeword $\mathbf{c}=(c_{0},c_{1},\cdots,c_{N-1})\in R^{N}$ is defined to be the rational sum of the Gray weight of its components. For any $\mathbf{x},\mathbf{y}\in R^{N}$, the Gray distance is given by $d_{G}(\mathbf{x},\mathbf{y})=w_{G}(\mathbf{x}-\mathbf{y})$. The Gray distance $d_{G}(C)$ of a code $C$ is the smallest nonzero Gray distance between all pairs of distinct codewords. The minimum Gray weight $w_{G}(C)$ of $C$ is the smallest nonzero Gray weight among all codewords. If $C$ is linear, then the Gray distance $d_{G}(C)$ is the same as the minimum Gray weight $w_{G}(C)$.

Proposition 3.2. The Gray map $\Phi_{M}$ is a distance-preserving map from ($R^{N}$, Gray distance) to ($\mathbb{F}_{p^{2m}}^{2N}$, Hamming distance) and it is also $\mathbb{F}_{p^{2m}}$-linear.

Proof   Let $k_{1},k_{2}\in\mathbb{F}_{p^{2m}}$, $\mathbf{x},\mathbf{y}\in R^{N}$, from the definition of Gray map, we have $\Phi_{M}(k_{1}\mathbf{x}+k_{2}\mathbf{y})=k_{1}\Phi_{M}(\mathbf{x})+k_{2}\Phi_{M}(\mathbf{y})$, which means that $\Phi_{M}$ is an $\mathbb{F}_{p^{2m}}$-linear map. Moreover,

$$d_{G}(\mathbf{x},\mathbf{y}) =w_{G}(\mathbf{x}-\mathbf{y})$$

$$=w_{H}(\Phi_{M}(\mathbf{x}-\mathbf{y}))$$

$$=w_{H}(\Phi_{M}(\mathbf{x})-\Phi_{M}(\mathbf{y}))$$

$$=d_{H}(\Phi_{M}(\mathbf{x}),\Phi_{M}(\mathbf{y})).$$

This completes the proof. ∎

Lemma 3.3. Let $M=$ $\begin{pmatrix}a\ &b\\ s\ &t\\ \end{pmatrix}\in GL_{2}(\mathbb{F}_{p^{2m}})$, the conjugation matrix of $M$ be $\overline{M}=$ $\begin{pmatrix}\bar{a}\ &\bar{b}\\ \bar{s}\ &\bar{t}\\ \end{pmatrix}.$ Then

$M\overline{M}^{T}=\Omega=\begin{pmatrix}\lambda&0\\ 0&\lambda\\ \end{pmatrix}$ $(\lambda\in\mathbb{F}_{p^{2m}}^{*}=\mathbb{F}_{p^{2m}}\backslash\{0\})$

if and only if $a\bar{s}+b\bar{t}=0$ and $a\bar{a}+b\bar{b}=s\bar{s}+t\bar{t}=\lambda\neq 0$.

Proof    Note that $M\overline{M}^{T}=\begin{pmatrix}a\bar{a}+b\bar{b}&a\bar{s}+b\bar{t}\\ \bar{a}s+\bar{b}t&s\bar{s}+t\bar{t}\\ \end{pmatrix}$. The desired result follows. ∎

Theorem 3.4. Let $C$ be a linear code of length $N^{\prime}=2^{e}n$ over $\mathbb{F}_{2^{2m}}+u\mathbb{F}_{2^{2m}},$ where $e$ is a non-negative integer and $n$ is odd. Let $M=$ $\begin{pmatrix}a\ &b\\ s\ &t\\ \end{pmatrix}\in GL_{2}(\mathbb{F}_{2^{2m}})$ and $M\overline{M}^{T}=\Omega$. Then $\Phi_{M}(C^{\bot_{H}})=\Phi_{M}(C)^{\bot_{H}}$. Moreover, if $C^{\bot_{H}}\subseteq C$, then $\Phi_{M}(C^{\bot_{H}})\subseteq\Phi_{M}(C)^{\bot_{H}}$.

Proof    Let $\textbf{c}_{1}=\textbf{r}_{1}+u\textbf{q}_{1}\in C$ and $\textbf{c}_{2}=\textbf{r}_{2}+u\textbf{q}_{2}\in C^{\bot_{H}}$, where $\mathbf{r}_{i}=(r_{i1},r_{i2},\cdots,r_{iN^{\prime}})$ and $\mathbf{q}_{i}=(q_{i1},q_{i2},\cdots,q_{iN^{\prime}})\in\mathbb{F}_{2^{2m}}^{N^{\prime}}$. The Hermitian inner product of $\textbf{c}_{1}$ and $\textbf{c}_{2}$ is

$$\langle\textbf{c}_{1},\textbf{c}_{2}\rangle_{H} =\sum_{i=1}^{N^{\prime}}(r_{1i}+uq_{1i})(\overline{r_{2i}}+u\overline{q_{2i}})$$

$$=\sum_{i=1}^{N^{\prime}}r_{1i}\overline{r_{2i}}+u\sum_{i=1}^{N^{\prime}}(r_{1i}\overline{q_{2i}}+\overline{r_{2i}}q_{1i})$$

$$=\langle\textbf{r}_{1},\textbf{r}_{2}\rangle_{H}+u(\langle\textbf{r}_{1},\textbf{q}_{2}\rangle_{H}+\langle\textbf{q}_{1},\textbf{r}_{2}\rangle_{H}).$$

It follows from $\langle\textbf{c}_{1},\textbf{c}_{2}\rangle_{H}=0$ that $\langle\textbf{r}_{1},\textbf{r}_{2}\rangle_{H}=0$ and $\langle\textbf{r}_{1},\textbf{q}_{2}\rangle_{H}+\langle\textbf{q}_{1},\textbf{r}_{2}\rangle_{H}=0$.

On the other hand,

$$\langle\Phi_{M}(\textbf{c}_{1}),\Phi_{M}(\textbf{c}_{2})\rangle_{H}$$

$$=~{} \sum_{i=1}^{N^{\prime}}[aq_{1i}+s(r_{1i}+q_{1i})]\overline{[aq_{2i}+s(r_{2i}+q_{2i})]}+\sum_{i=1}^{N^{\prime}}[bq_{1i}+t(r_{1i}+q_{1i})]\overline{[bq_{2i}+t(r_{2i}+q_{2i})]}$$

$$=~{} (a\overline{a}+b\overline{b})\langle\textbf{q}_{1},\textbf{q}_{2}\rangle_{H}+(a\overline{s}+b\overline{t})(\langle\textbf{q}_{1},\textbf{r}_{2}\rangle_{H}+\langle\textbf{q}_{1},\textbf{q}_{2}\rangle_{H})+(s\overline{a}+t\overline{b})(\langle\textbf{r}_{1},\textbf{q}_{2}\rangle_{H}+\langle\textbf{q}_{1},\textbf{q}_{2}\rangle_{H})$$

$$+(s\overline{s}+t\overline{t})(\langle\textbf{r}_{1},\textbf{r}_{2}\rangle_{H}+\langle\textbf{r}_{1},\textbf{q}_{2}\rangle_{H}+\langle\textbf{q}_{1},\textbf{r}_{2}\rangle_{H}+\langle\textbf{q}_{1},\textbf{q}_{2}\rangle_{H})$$

$$=~{} \lambda[\langle\textbf{r}_{1},\textbf{r}_{2}\rangle_{H}+\langle\textbf{r}_{1},\textbf{q}_{2}\rangle_{H}+\langle\textbf{q}_{1},\textbf{r}_{2}\rangle_{H}],$$

$$=~{} 0,$$

where the third equation follows from Lemma 3.3. Therefore, $\Phi_{M}(C^{\bot_{H}})\subseteq\Phi_{M}(C)^{\bot_{H}}$.

Suppose $|C|=2^{4m(2k_{0}+k_{1})}$. Since $\Phi_{M}$ is a bijection,

$$|\Phi_{M}(C^{\bot_{H}})| =|C^{\bot_{H}}|$$

$$=2^{2m(2N^{\prime}-2k_{0}-k_{1})}$$

$$=2^{4mN^{\prime}}/|\Phi_{M}(C)|$$

$$=|\Phi_{M}(C)^{\perp}|.$$

Therefore, $\Phi_{M}(C^{\bot_{H}})=\Phi_{M}(C)^{\bot_{H}}$. The second desired result directly follows. ∎

4 Hermitian construction

In this section, we obtain a new family of $2^{m}$-ary quantum codes by using Hermitian $\alpha(1+u)$-constacyclic self-orthogonal codes over $\mathbb{F}_{2^{2m}}+u\mathbb{F}_{2^{2m}}$ of length $N^{\prime}=2^{e}n.$ We firstly introduce some definitions and notations of polynomials in $R[x].$

Let $f(x)=a_{k}x^{k}+a_{k-1}x^{k-1}+\cdots+a_{0}$ be a polynomial in $R[x],$ where $a_{0}$ is an invertible element in $R.$ The conjugation polynomial of $f(x)$ is defined by $\overline{f(x)}=\overline{a_{k}}x^{k}+\overline{a_{k-1}}x^{k-1}+\cdots+\overline{a_{0}}$. The reciprocal polynomial of $f(x)$ is defined by $f^{*}(x)=x^{k}f(x^{-1})$, i.e., $f^{*}(x)=a_{0}x^{k}+a_{1}x^{k-1}+\cdots+a_{k}$. Obviously, $(f^{*}(x))^{*}=f(x)$ and $(f(x)g(x))^{*}=f^{*}(x)g^{*}(x)$. We denote by $f^{\dagger}(x)$ the conjugation of the reciprocal polynomial of $f(x)$, i.e. $f^{\dagger}(x)=\overline{f^{*}(x)}.$

To study the Hermitian dual of $\alpha(1+u)$-constacyclic codes, we need the following result. Its proof is similar to that of Theorem 4.1 in Ref.[32] and is omitted.

Theorem 4.1. Let $C$ be an $\alpha(1+u)$-constacyclic code over $R$ of length $N=p^{e}n$ ($n$ prime to $p$). Then $C^{\bot H}$ is an $\overline{\alpha^{-1}}(1+u)$-constacyclic code over $R$ of length $N$.

Since $\gcd(p^{e},p^{2m}-1)=1$, there exists $\beta\in\mathbb{F}_{p^{2m}}^{*}$ such that $\beta^{p^{e}}=\alpha$. In $R[x]/\langle x^{N}-\alpha(1+u)\rangle$,

$$u ~{}=\alpha^{-1}(x^{N}+\alpha)$$

$$~{}=\alpha^{-1}(x^{n}+\beta)^{p^{e}}.$$

Therefore, $u\in\langle(x^{n}+\beta)^{p^{e}}\rangle$. Let

$$x^{n}-\beta=\prod_{i=1}^{r}f_{i}(x)$$

be the factorization of $x^{n}-\beta$ into irreducible factors over $\mathbb{F}_{p^{2m}}$. Moreover, we know that $\alpha(1+u)$-constacyclic codes over $R$ of length $N$ are precisely ideals of $R[x]/\langle x^{N}-\alpha(1+u)\rangle$ and the equation $(x^{n}-\xi(1+u))^{2L}=(x^{n}-\xi)^{2L}$ always holds in $R[x]$, where $\xi\in\mathbb{F}_{p^{2m}}$ and $L$ is any integer. Similar to Theorem 3.4 of Ref.[17], we get the following result.

Theorem 4.2. Let $x^{n}-\beta=\prod_{i=1}^{r}f_{i}(x)$ be the unique factorization of $x^{n}-\beta$ into a product of monic basic irreducible pairwise coprime polynomials in $\mathbb{F}_{p^{2m}}[x],$ where $\emph{gcd}(n,p)=1$. If $C$ is an $\alpha(1+u)$-constacyclic code of length $N=p^{e}n$ over $R$, then $C=\langle\prod_{i=1}^{r}f_{i}^{k_{i}}(x)\rangle$, where $0\leq k_{i}\leq p^{e+1}$ and $|C|=p^{2m(2N-\sum_{i=1}^{r}k_{i}\emph{deg}(f_{i}(x)))}$.

Now, we give the structure of the Hermitian dual of $\alpha(1+u)$-constacyclic codes of length $N=p^{e}n$ over $R.$

Theorem 4.3. Let $\alpha\in\mathbb{F}_{p^{2m}}$ such that $\alpha\bar{\alpha}=1$. Let $x^{n}-\beta=\prod_{i=1}^{r}f_{i}(x)$ be the unique factorization of $x^{n}-\beta$ into a product of monic basic irreducible pairwise coprime polynomials in $\mathbb{F}_{p^{2m}}[x],$ where $\emph{gcd}(n,p)=1$. If $C$ is an $\alpha(1+u)$-constacyclic code of length $N=p^{e}n$ over $R$, then $C=\langle\prod_{i=1}^{r}f_{i}^{k_{i}}(x)\rangle$, where $0\leq k_{i}\leq p^{e+1}$. Then $C^{\perp H}=\langle\prod_{i=1}^{r}[f_{i}^{\dagger}(x)]^{p^{e+1}-k_{i}}\rangle$ and $|C^{\perp H}|=p^{2mt}$, where $t=\sum\nolimits_{i=1}^{r}{k_{i}\deg(f_{i}(x))}.$

Proof. Let $G=\langle\prod_{i=1}^{r}[f_{i}^{\dagger}(x)]^{p^{e+1}-k_{i}}\rangle$ be an ideal of $R[x]/\langle x^{N}-\alpha(1+u)\rangle$. By direct computation, we get

$$\prod_{i=1}^{r}f_{i}^{k_{i}}(x)\cdot\left(\prod_{i=1}^{r}[f_{i}^{\dagger}(x)]^{p^{e+1}-k_{i}}\right)^{\dagger}=\prod_{i=1}^{r}f_{i}^{p^{e+1}}(x)=[x^{p^{e}n}-\beta]^{2}=0.$$

Hence, $G\subseteq C^{\perp}$. For each $i$, let $a_{i}$ denote the constant of $f_{i}(x)$. Since $\prod_{i=1}^{r}f_{i}(x)=x^{n}-\beta$, we have $\prod_{i=1}^{r}a_{i}=\beta$. It follows that each ${a_{i}}$ is an invertible element of $R$ and ${\bar{a}_{i}}$ is a leading coefficient of ${f_{i}^{\dagger}(x)}$. Hence, for each $i$, there exists a suitable invertible element $b_{i}$ of $R$ such that $b_{i}f_{i}^{\dagger}(x)$ is a monic polynomial. Note that $b_{i}=\bar{a}_{i}^{-1}$ and $\prod_{i=1}^{r}b_{i}=\prod_{i=1}^{r}\bar{a}_{i}^{-1}=\beta$. So

$$\prod\nolimits_{i=1}^{r}b_{i}f_{i}^{\dagger}(x) = \prod\nolimits_{i=1}^{r}b_{i}\cdot\prod\nolimits_{i=1}^{r}\overline{f_{i}^{*}(x)}$$

$$= \beta x^{n}\prod\nolimits_{i=1}^{r}\overline{f_{i}(x^{-1})}$$

$$= \beta x^{n}(\overline{{x^{-n}-\beta}})$$

$$= x^{n}-\beta.$$

Thus, $b_{i}f_{i}^{\dagger}(x)$ ($1\leq i\leq r$) are monic basic irreducible divisors of $x^{n}-\beta$ in $R[x]$. Note that

$$G=\left\langle\prod_{i=1}^{r}[f_{i}^{\dagger}(x)]^{p^{e+1}-k_{i}}\right\rangle=\left\langle\prod_{i=1}^{r}[b_{i}f_{i}^{\dagger}(x)]^{p^{e+1}-k_{i}}\right\rangle,$$

On the other hand, by Theorem 4.1, we have

$$|G| = p^{2m(2N-\sum_{i=1}^{r}(p^{e+1}-k_{i})\textrm{deg}(b_{i}f_{i}^{\dagger}(x))}$$

$$= p^{2m(2N-\sum_{i=1}^{r}(p^{e+1}-k_{i})\textrm{deg}(f_{i}(x))}$$

$$= p^{2m(\sum_{i=1}^{r}k_{i}\textrm{deg}(f_{i}(x))}$$

$$= p^{2mt}.$$

Furthermore, from $|C||G|=p^{4mN}=|C||C^{\bot H}|$, we have that $|G|=|C^{\bot H}|$ . This gives the proof. ∎

Next, we consider the Gray images of $\alpha(1+u)$-constacyclic codes over $R.$

Theorem 4.4. Let $C$ be an $\alpha(1+u)$-constacyclic code of length $N=p^{e}n$ ($n$ prime to $p$) over $R$. Let $M=\begin{pmatrix}a\ &b\\ s\ &t\\ \end{pmatrix}\in GL_{2}(\mathbb{F}_{p^{2m}})$, where $a=t\alpha,s=b\alpha$ and $b\neq t.$ Then $\Phi_{M}(C)$ is an $\alpha^{2}$-constacyclic code of length $2N$ over $\mathbb{F}_{p^{2m}}$.

Proof. Let $\textbf{c}=(c_{0},c_{1},\ldots,c_{N-1})\in R^{N}$, where $c_{i}=r_{i}+uq_{i},$ for $i=0,1,\ldots,N-1.$ From the definition of Gray map, we have

$$\Phi_{M}(\textbf{c}) = (aq_{0}+s(r_{0}+q_{0}),\ldots,aq_{N-1}+s(r_{N-1}+q_{N-1}),bq_{0}+t(r_{0}+q_{0}),\ldots,bq_{N-1}+t(r_{N-1}+q_{N-1})).$$

Therefore

$$\sigma_{\alpha^{2}}(\Phi_{M}(\textbf{c})) = (\alpha^{2}[bq_{N-1}+t(r_{N-1}+q_{N-1})],aq_{0}+s(r_{0}+q_{0}),\ldots,aq_{N-1}+s(r_{N-1}+q_{N-1}),\ldots,$$

$$bq_{0}+t(r_{0}+q_{0}),\ldots,bq_{N-2}+t(r_{N-2}+q_{N-2})).$$

On the other hand

$$\sigma_{\alpha(1+u)}(\textbf{c}) = (\alpha(1+u)c_{N-1},c_{0},\ldots,c_{N-2})$$

$$= (\alpha(r_{N-1}+u(r_{N-1}+q_{N-1})),a_{0}+ub_{0},\ldots,a_{N-2}+ub_{N-2}).$$

We can get

$$\Phi_{M}(\sigma_{\alpha(1+u)}(\textbf{c})) = (\alpha a(r_{N-1}+q_{N-1})+\alpha sq_{N-1},aq_{0}+s(r_{0}+q_{0}),\ldots,aq_{N-2}+s(r_{N-2}+q_{N-2}),$$

$$\alpha b(r_{N-1}+q_{N-1})+\alpha tq_{N-1}bq_{0}+t(r_{0}+q_{0}),\ldots,bq_{N-2}+t(r_{N-2}+q_{N-2})).$$

Hence, $\sigma_{\alpha^{2}}(\Phi_{M}(\textbf{c}))=\Phi_{M}(\sigma_{\alpha(1+u)}(\textbf{c}))$. It follows that $\Phi_{M}(C)$ is an $\alpha^{2}$-constacyclic code of length $2N$. This completes the proof. ∎

Consequently, we can obtain the following corollary.

Corollary 4.5. The Gray image of an $\alpha(1+u)$-constacyclic code of length $N=p^{e}n$(n prime to p) over $R$ under the Gray map $\Phi_{M}$ is a distance-invariant $\alpha^{2}$-constacyclic code of length $2N$ over $\mathbb{F}_{p^{2m}}$.

Define the polynomial Gray map as follows.

$\Phi_{M}:R[x]/\langle x^{n}-\alpha(1+u)\rangle\rightarrow\mathbb{F}_{2^{2m}}[x]/\langle x^{2n}-\alpha^{2}\rangle,$

$c(x)=r(x)+uq(x)\mapsto(a+s)q(x)+sr(x)+x^{n}[(b+t)q(x)+tr(x)],$

where $r(x),q(x)\in\mathbb{F}_{2^{2m}}[x]$. It is obvious that $\Phi_{M}(c(x))$ is the polynomial representation of $\Phi_{M}(c)$.

Theorem 4.6. Let $\alpha\in\mathbb{F}_{2^{2m}}$ such that $\alpha\bar{\alpha}=1.$ Let $C$ be an $\alpha(1+u)$-constacyclic code over $\mathbb{F}_{2^{2m}}+u\mathbb{F}_{2^{2m}}$ of length $N^{\prime}=2^{e}n$($n$ is odd) and $C=\langle\prod_{i=1}^{r}f_{i}^{k_{i}}(x)\rangle.$ Let $M=\begin{pmatrix}t\alpha\ &b\\ b\alpha\ &t\\ \end{pmatrix}\in GL_{2}(\mathbb{F}_{p^{2m}})$ and $b\neq t.$ Then $\Phi_{M}(C^{\perp H})$ is an $\alpha^{2}$-constacyclic code over $\mathbb{F}_{2^{2m}}$ of length $2N^{\prime}$ and $\Phi_{M}(C^{\perp H})=\big{\langle}\prod_{i=1}^{r}[f_{i}^{\dagger}(x)]^{2^{e+1}-k_{i}}\big{\rangle}.$

Proof. By Theorem 4.4, we have $C^{\perp H}=\big{\langle}\prod_{i=1}^{r}[f_{i}^{\dagger}(x)]^{2^{e+1}-k_{i}}\big{\rangle}.$ Let $c(x)$ be any element of $C^{\perp H}.$ Then $c(x)$ reduced modulo $u$ must be in the $\alpha$-constacyclic code over $\mathbb{F}_{2^{2m}}$ of length $N$ with $\big{\langle}g(x)\big{\rangle},$ where $g(x)=\prod_{i=1}^{r}[f_{i}^{\dagger}(x)]^{h_{i}}$ with $1\leq i\leq r$ and $h_{i}=\textrm{min}\{2^{e+1}-k_{i},2^{e}\}.$ Hence $c(x)$ can be written as $c(x)=g(x)f_{1}(x)+uf_{2}(x)$, where $f_{1}(x),f_{2}(x)\in\mathbb{F}_{2^{2m}}[x]$. Then

$\Phi_{M}(c(x))=(b+t)\alpha f_{2}(x)+b\alpha f_{1}(x)g(x)+x^{n}[(b+t)f_{2}(x)+tf_{1}(x)g(x)]$

$=(b+t)f_{2}(x)(x^{n}-\alpha)+f_{1}(x)g(x)(b\alpha+x^{n}t).$

Since $g(x)|x^{n}-\alpha$, there exists $h(x)\in\mathbb{F}_{2^{2m}}[x]$ such that $x^{n}-\alpha=g(x)h(x)$. It follows that

$\Phi_{M}(c(x))=(b+t)f_{2}(x)(x^{n}-\alpha)+f_{1}(x)g(x)(b\alpha+x^{n}t)$

$=(b+t)f_{2}(x)g(x)h(x)+f_{1}(x)g(x)(b\alpha+x^{n}t)$

$=g(x)[(b+t)f_{2}(x)h(x)+f_{1}(x)(b\alpha+x^{n}t)].$

So, $\Phi_{M}(C^{\bot H})\subseteq\langle g(x)\rangle$. Now, if $m(x)\in\Phi_{M}(C^{\bot H})$, then $m(x)=k(x)g(x),$ where $k(x)\in\mathbb{F}_{2^{2m}}[x]$. Let $g(x)s_{1}(x)=x^{n}-\alpha,$ for some $s(x)\in R[x]$ satisfies $s(x)$ reduced modulo $u$ is $s_{1}(x).$ Thus

$$m(x)s_{1}(x)x^{n}=k(x)x^{n}(x^{n}-\alpha)=\Phi(uk(x)).$$

So, $uk(x)\in C^{\bot H}$. Hence $uk(x)=\langle u\prod_{i=1}^{r}[f_{i}^{\dagger}(x)]^{t_{i}}\rangle,$ where $t_{i}=2^{e+1}-k_{i}-\textrm{min}\{2^{e+1}-k_{i},2^{e}\}.$ This shows that $k(x)=\prod_{i=1}^{r}[f_{i}^{\dagger}(x)]^{t_{i}}I(x)$ for some polynomial $I(x)\in\mathbb{F}_{2^{2m}}[x]$. Hence

$$m(x)=\prod_{i=1}^{r}[f_{i}^{\dagger}(x)]^{t_{i}}I(x)\prod_{i=1}^{r}[f_{i}^{\dagger}(x)]^{h_{i}}=I(x)\prod_{i=1}^{r}[f_{i}^{\dagger}(x)]^{2^{e+1}-k_{i}}.$$

This shows that

$$\Phi_{M}(C^{\bot H})\subseteq\left\langle\prod_{i=1}^{r}[f_{i}^{\dagger}(x)]^{2^{e+1}-k_{i}}\right\rangle.$$

Computing their cardinalities, we get that $\Phi_{M}(C^{\bot H})=\left\langle\prod_{i=1}^{r}[f_{i}^{\dagger}(x)]^{2^{e+1}-k_{i}}\right\rangle.$ ∎

Now, we give an example to illustrate the above results as follows. Comparing with the previously known quaternary linear code in Ref.[13], our obtained $4$-ary linear code is optimal.
Example 4.7.   Let $\mathbb{F}_{4}=\{0,1,\omega,\omega^{2}=1+\omega\}$. Consider an $\omega(1+u)$-constacyclic code over $\mathbb{F}_{4}+u\mathbb{F}_{4}$ of length $10$. Note

$$x^{10}-\omega=(f_{0}(x))^{2}(f_{1}(x))^{2}(f_{2}(x))^{2}$$

over $\mathbb{F}_{4}+u\mathbb{F}_{4},$ where $f_{0}(x)=x+\omega,$ $f_{1}(x)=x^{2}+x+\omega^{2},$ $f_{2}(x)=x^{2}+\omega^{2}x+\omega^{2}.$ Let $C=\left\langle f_{0}(x)(f_{1}(x))^{3}(f_{2}(x))^{4}\right\rangle$. Then $C^{\perp H}=\left\langle(f_{0}^{\dagger}(x))^{3}f_{1}^{\dagger}(x)\right\rangle,$ where $f_{0}^{\dagger}(x)=f_{0}(x),\ f_{1}^{\dagger}(x)=f_{1}(x).$ Take $M=\begin{pmatrix}\omega t\ &b\\ \omega b\ &t\\ \end{pmatrix}$, where $b,t\in\mathbb{F}_{4}^{*}$ and $b\neq t.$ Computation of the Hamming distances of $\Phi_{M}(C^{\bot H})$ with the help of the computer algebra system MAGMA, we get the Hamming distance of $\Phi_{M}(C^{\bot H})$ is $4.$ Hence, the Gray image $\Phi_{M}(C^{\bot H})$ is an $\omega^{2}$-constacyclic code with $[20,15,4]$ over $\mathbb{F}_{4}.$ Compared with the previously known quaternary linear code in Ref.[13], which is an optimal code.

Let $\eta$ be a primitive element of $\mathbb{F}_{p^{2m}}$, and let $\alpha\in\mathbb{F}_{p^{2m}}$ with order $r>1$ and $r\mid(p^{m}+1)$, then $\bar{\alpha}\alpha=1$. Since $\gcd(p^{e},r)=1$, there is $f$ with $1\leq f\leq p^{m}$ such that $p^{e}f\equiv 1\pmod{r}$. Suppose $\beta=\alpha^{f}$, then $\beta^{2^{e}}=\alpha^{p^{e}f}=\alpha$. Clearly, $\beta\in\mathbb{F}_{p^{2m}}$ with order $r$. Let $\delta$ be a primitive $rn$-th root of unity in some extension field of $\mathbb{F}_{p^{2m}}$ such that $\delta^{n}=\beta$. Then

$$x^{n}-\beta=\prod\limits_{i=0}^{n-1}(x-\delta^{1+ir}).$$

Let $\Gamma$ be a complete set of $p^{2m}$-cyclotomic coset representatives modulo $rn$ and $I=\{i\in\Gamma|i\equiv 1\pmod{r}\}$. For each $i\in I$, let $C_{i}=\left\{{i,ip^{2m},\ldots,i(p^{2m})^{\ell_{i}-1}}\right\}$ be the $p^{2m}$-cyclotomic coset modulo $rn$ containing $i$, where $\ell_{i}$ is the smallest positive integer such that $i\equiv i(p^{2m})^{\ell_{i}}\pmod{rn}$. A cyclotomic coset $C_{i}$ is called symmetric if $rn-p^{m}i\in C_{i}$ and asymmetric otherwise. Asymmetric cosets $C_{i}$ and $C_{-p^{m}i}$ come in pair, we use $(C_{i},C_{-p^{m}i})$ to denote such an asymmetric pair. Let $M_{i}(x)$ denote the minimal polynomial of $\delta^{i}$ over $\mathbb{F}_{p^{2m}}$, then

$$M_{i}(x)=\prod_{j\in C_{i}}(x-\delta^{j}).$$

It is easily verified that there are $I_{1},I_{2}\subset I$ such that

$$x^{n}-\beta=\prod_{i\in I_{1}}M_{i}(x)\prod_{i\in I_{2}}M_{i}(x)M_{-p^{m}i}(x).$$

By Theorem 4.3, we know that any $\alpha(1+u)$-constacyclic code of length $N=p^{e}n$ over $R$ has the form

$$C=\big{\langle}\prod_{i\in I_{1}}M_{i}^{k_{i}}(x)[\prod_{l\in I_{2}}M_{l}^{r_{l}}(x)M_{-p^{m}l}^{s_{l}}(x)]\big{\rangle},$$

where $0\leq k_{i},r_{l},s_{l}\leq p^{e+1}.$

Theorem 4.8. Let $\alpha\in\mathbb{F}_{p^{2m}}$ such that $\alpha\bar{\alpha}=1$ and let

$$C=\big{\langle}\prod_{i\in I_{1}}M_{i}^{k_{i}}(x)\big{[}\prod_{l\in I_{2}}M_{l}^{r_{l}}(x)M_{-p^{m}l}^{s_{l}}(x)\big{]}\big{\rangle}$$

be an $\alpha(1+u)$-constacyclic code of length $N=p^{e}n$ over $R$, where $\emph{gcd}(n,p)=1$ and $0\leq k_{i},r_{l},s_{l}\leq p^{e+1}$. Then

$$C^{\bot H}=\big{\langle}\prod_{i\in I_{1}}(M_{i}^{{\dagger}}(x))^{p^{e+1}-k_{i}}[\prod_{l\in I_{2}}(M_{l}^{{\dagger}}(x))^{p^{e+1}-r_{l}}(M_{-p^{m}l}^{{\dagger}}(x))^{p^{e+1}-s_{l}}]\big{\rangle}$$

and $|C^{\perp H}|=p^{2m\varepsilon}$, where $\varepsilon=\sum_{i\in I_{1}}k_{i}\emph{deg}[M_{i}(x)]+\sum_{l\in I_{2}}\{r_{l}\emph{deg}[M_{l}(x)]+s_{l}\emph{deg}[M_{-p^{m}l}(x)]\}.$

Theorem 4.9. Let $\alpha\in\mathbb{F}_{p^{2m}}$ such that $\alpha\bar{\alpha}=1$ and let

$$C=\big{\langle}\prod_{i\in I_{1}}M_{i}^{k_{i}}(x)\big{[}\prod_{l\in I_{2}}M_{l}^{r_{l}}(x)M_{-p^{m}l}^{s_{l}}(x)\big{]}\big{\rangle}$$

be an $\alpha(1+u)$-constacyclic code of length $N=p^{e}n$ over $R$, where $\emph{gcd}(n,p)=1$ and $0\leq k_{i},r_{l},s_{l}\leq p^{e+1}$. Then $C\subseteq C^{\bot H}$ if and only if   $p^{e}\leq k_{i}\leq p^{e+1}$ and $p^{e+1}\leq r_{l}+s_{l}\leq p^{e+2}.$

Proof. It is known that

$$C^{\bot H}=\big{\langle}\prod_{i\in I_{1}}(M_{i}^{{\dagger}}(x))^{p^{e+1}-k_{i}}\big{[}\prod_{l\in I_{2}}(M_{l}^{{\dagger}}(x))^{p^{e+1}-r_{l}}(M_{-p^{m}l}^{{\dagger}}(x))^{p^{e+1}-s_{l}}\big{]}\big{\rangle}.$$

Therefore, $C\subseteq C^{\bot H}$ if and only if

$$\prod_{i\in I_{1}}(M_{i}^{{\dagger}}(x))^{p^{e+1}-k_{i}}\big{[}\prod_{l\in I_{2}}(M_{l}^{{\dagger}}(x))^{p^{e+1}-r_{l}}(M_{-p^{m}l}^{{\dagger}}(x))^{p^{e+1}-s_{l}}\big{]}$$

divides

$$\prod_{i\in I_{1}}M_{i}^{k_{i}}(x)\big{[}\prod_{l\in I_{2}}M_{l}^{r_{l}}(x)M_{-p^{m}l}^{s_{l}}(x)\big{]}.$$

Comparing the indexes, we obtain that $p^{e}\leq k_{i}\leq p^{e+1}$ and $p^{e+1}\leq r_{l}+s_{l}\leq p^{e+2}$. ∎

Now, we give an example to illustrate the results of Theorems 4.8 and 4.9 as follows.
Example 4.10. Consider an $\omega(1+u)$-constacyclic code over $\mathbb{F}_{4}+u\mathbb{F}_{4}$ of length $34$. The set $\{0,1,2,3,6\}$ is a complete set of $4$-cyclotomic coset representatives modulo $17$. In $(\mathbb{F}_{4}+u\mathbb{F}_{4})[x]$,

$$x^{34}-\omega=(M_{0}(x)M_{1}(x)M_{2}(x)M_{3}(x)M_{6}(x))^{2},$$

where

$$M_{0}(x)=x+\omega,$$

$$M_{1}(x)=x^{4}+(1+u)x^{3}+\omega x^{2}+(1+u)x+\omega,$$

$$M_{2}(x)=x^{4}+(1+u)x^{3}+\bar{\omega}x^{2}+(1+u)x+\omega,$$

$$M_{3}(x)=x^{4}+(\omega+u\omega)x^{3}+x^{2}+(\omega+u\omega)x+\omega,$$

$$M_{6}(x)=x^{4}+(\bar{\omega}+u\bar{\omega})x^{3}+x^{2}+(\bar{\omega}+u\bar{\omega})x+\omega.$$

Let $C=\langle(M_{0}(x))^{2}(M_{1}(x))^{4}(M_{3}(x))^{4}(M_{6}(x))^{2}\rangle$. By Theorem 4.8, the Hermitian dual code of $C$ is

$$C^{\bot H}=\langle(M_{0}^{{\dagger}}(x))^{2}(M_{2}^{{\dagger}}(x))^{4}(M_{6}^{{\dagger}}(x))^{2}\rangle,$$

where $M_{0}^{{\dagger}}(x)=M_{0}(x),\ M_{2}^{{\dagger}}(x)=M_{1}(x),\ M_{6}^{{\dagger}}(x)=M_{3}(x).$ By Theorem 4.9, we obtain $C\subseteq C^{\bot H}.$

We use the notation $[[n,k,d]]_{q}$ to denote a $q$-ary quantum error-correcting code for $n$ qubits having $q^{k}$ codewords and the minimum distance $d$. For an $[[n,k,d]]_{q}$ quantum code $C,$ the parameters satisfies the Singleton bound, i.e., $k\leq n-2d+2$. If a quantum code attains this bound, i.e., $k=n-2d+2$, it is called a quantum maximum-distance-separable (MDS) code. Now, we recall the Hermitian quantum code construction:

Theorem 4.11. (Hermitian construction [8]) If there exists a $\mathbb{F}_{q^{2}}$-linear code $C=[2N^{\prime},k]_{q^{2}}$ such that $C\subseteq C^{\bot_{H}}$, then there exists a $q$-ary $[[2N^{\prime},2N^{\prime}-2k,\geq d^{\bot H}]]_{q}$ quantum code, where $d^{\bot H}$ is the minimum Hamming weight of $C^{\bot_{H}}$.

Combining Theorem 4.6 with Theorem 4.10, we see that a Hermitian $\alpha^{2}$-constacyclic self-orthogonal code of length $2N^{\prime}$ over $\mathbb{F}_{2^{2m}}$ can be constructed from a Hermitian $\alpha(1+u)$-constacyclic self-orthogonal code over $R$ of length $N^{\prime}$ under the Gray map $\Phi_{M}$. According to Theorem 4.11, we can obtain the following theorem, which can be used for construction of quantum codes.

Theorem 4.12. Let $\alpha\in\mathbb{F}_{2^{2m}}$ such that $\alpha\bar{\alpha}=1$ and let

$$C=\big{\langle}\prod_{i\in I_{1}}M_{i}^{k_{i}}(x)\big{[}\prod_{l\in I_{2}}M_{l}^{r_{l}}(x)M_{-2^{m}l}^{s_{l}}(x)\big{]}\big{\rangle}$$

be an $\alpha(1+u)$-constacyclic code of length $N^{\prime}=2^{e}n$ ($n$ is odd) over $\mathbb{F}_{2^{2m}}+u\mathbb{F}_{2^{2m}}$, and size $2^{2k},$ where $0\leq k_{i},r_{l},s_{l}\leq 2^{e+1}$. Let $M=\begin{pmatrix}t\alpha\ &b\\ b\alpha\ &t\\ \end{pmatrix}\in GL_{2}(\mathbb{F}_{2^{2m}})$, where $b,t\in\mathbb{F}_{2^{2m}}^{*}$ and $b\neq t$ . Suppose that $2^{e}\leq k_{i}\leq 2^{e+1}$ and $2^{e+1}\leq r_{l}+s_{l}\leq 2^{e+2}$. Then there exists a $2^{m}$-ary quantum code with parameters $[[2N^{\prime},2N^{\prime}-2k,\geq d^{\bot H}]]_{2^{m}},$ where $d^{\bot H}$ is the minimum Hamming distances of the $\alpha^{2}$-constacyclic code $\Phi_{M}(C^{\bot H}).$

Now, we give an example to illustrate the above results as follows.
Example 4.13.  Let $\mathbb{F}_{4^{2}}=\{0,1,\omega,\omega^{2},\ldots,\omega^{14}\}$ be a finite field with sixteen elements, where $\omega^{4}=\omega+1.$ Consider a Hermitian $\omega^{6}(1+u)$-constacyclic self-orthogonal code with length $6$ over $\mathbb{F}_{4^{2}}+u\mathbb{F}_{4^{2}}.$ Let $C$ be an $\omega^{6}(1+u)$-constacyclic code of length $6$ over $\mathbb{F}_{4^{2}}+u\mathbb{F}_{4^{2}}.$ Then $\beta=(\omega^{6})^{2^{4-1}}=\omega^{3}.$
Note that

$$x^{6}-\omega^{3}=(f_{0}(x))^{2}(f_{1}(x))^{2}(f_{2}(x))^{2}$$

over $\mathbb{F}_{4^{2}}+u\mathbb{F}_{4^{2}},$ where $f_{0}(x)=x+\omega^{3},$ $f_{1}(x)=x+\omega^{8},$ $f_{2}(x)=x+\omega^{13}.$ Let

$$C=\left\langle(f_{0}(x))^{4}(f_{1}(x))^{4}(f_{2}(x))^{2}\right\rangle.$$

By Theorem 4.9, $C$ is a Hermitian $\omega^{6}(1+u)$-constacyclic self-orthogonal code over $\mathbb{F}_{4^{2}}+u\mathbb{F}_{4^{2}}$ of length $6.$ Take $M=\begin{pmatrix}\omega^{6}t\ &b\\ \omega^{6}b\ &t\\ \end{pmatrix}$, where $b,t\in\mathbb{F}_{4^{2}}^{*}$ and $b\neq t.$ Computation of the Hamming distances of $\Phi_{M}(C^{\bot H})$ with the help of the computer algebra system MAGMA, we get the Hamming distance of $\Phi_{M}(C^{\bot H})$ is $2.$ Using the Hermitian construction, we can obtain a $Q=[[12,10,2]]_{4}$ quantum code, which is a quantum MDS code. Thus, our obtained quaternary quantum code is with good parameters.

5 Symplectic construction

In this section, our goal is to construct $p^{m}$-ary quantum codes with respect to the trace inner product by employing an $\alpha(1+u)$-constacyclic of length $N=p^{e}n$ ($n$ prime to $p$ ) over $R$. In the following, we will apply a map $\phi,$ which is from $R$ to $\mathbb{F}_{p^{2m}}^{2}.$ The map $\phi$ is defined as follows.

Definition 5.1. A map $\phi$ on $R$ is defined as

$$\phi:R \rightarrow\mathbb{F}_{p^{2m}}^{2}$$

$$r+uq \mapsto(q,r),$$

where $r,q\in\mathbb{F}_{p^{2m}}$.

Let $\mathfrak{A}=(\textbf{a}|\textbf{b})$ be any element in $\mathbb{F}_{p^{2m}}^{2N},$ where $\textbf{a}=(a_{0},a_{1},\cdots,a_{N-1}),$ $\textbf{b}=(b_{0},b_{1},\cdots,b_{N-1})$. Then we define the symplectic weight of $\mathfrak{A}$ as

$$\textrm{w}_{S}(\mathfrak{A})=|\{i|(a_{i},b_{i})\neq 0,i=0,1,\ldots,N-1\}|.$$

For $\mathfrak{A}=(\textbf{a}|\textbf{b}),\mathfrak{B}=(\textbf{a}^{\prime}|\textbf{b}^{\prime})\in\mathbb{F}_{p^{2m}}^{2N}$, the symplectic distance $d_{S}((\textbf{a}|\textbf{b}),(\textbf{a}^{\prime}|\textbf{b}^{\prime}))$ between two codewords $(\textbf{a}|\textbf{b})$ and $(\textbf{a}^{\prime}|\textbf{b}^{\prime})$ is the symplectic weight of the codeword $(\textbf{a}-\textbf{a}^{\prime}|\textbf{b}-\textbf{b}^{\prime})$. The minimum symplectic distance $d_{S}(\mathcal{C})$ of $\mathcal{C}$ is defined as $\min\{w_{S}((\textbf{a}|\textbf{b}),(\textbf{a}^{\prime}|\textbf{b}^{\prime}))|(\textbf{a}|\textbf{b}),(\textbf{a}^{\prime}|\textbf{b}^{\prime}))\in\mathcal{C},\textbf{a}-\textbf{a}^{\prime}\neq\textbf{b}-\textbf{b}^{\prime}\}$. For any linear code, the minimum symplectic distance $d_{S}(\mathcal{C})$ of $\mathcal{C}$ is its minimum symplectic weight. The trace inner product of $\mathfrak{A}=(\textbf{a}|\textbf{b}),\mathfrak{B}=(\textbf{a}^{\prime}|\textbf{b}^{\prime})\in\mathbb{F}_{p^{2m}}^{2N}$ is defined to be

$$\langle\mathfrak{A},\mathfrak{B}\rangle_{T}=\textbf{a}\cdot\overline{\textbf{b}^{\prime}}-\textbf{b}\cdot\overline{\textbf{a}^{\prime}}\in\mathbb{F}_{p^{m}}.$$

For a $p^{2m}$-ary linear code $\mathcal{C}$ of length $2N$, the trace dual code $\mathcal{C}^{\perp_{T}}$ is defined as

$$\mathcal{C}^{\perp_{T}}=\{\mathfrak{A}\in\mathbb{F}_{p^{2m}}^{2N}|\langle\mathfrak{A},\mathfrak{B}\rangle_{T}=0\ \textrm{for}\ \textrm{all}\ \mathfrak{B}\in\mathcal{C}\}.$$

A $p^{2m}$-ary linear code $\mathcal{C}$ of length $2N$ is called trace self-orthogonal if $\mathcal{C}\subseteq\mathcal{C}^{\bot_{T}}$, and trace self-dual if $\mathcal{C}^{\bot_{T}}=\mathcal{C}$.

It is known that $a+ub=0$ if and only if $a=b=0.$ So, from the above definitions, we directly obtain the following result.
Theorem 5.2. Let $C$ be a linear code of length $N$ over $R,$ $d_{H}(C)$ be the minimum Hamming weight of the linear code $C,$ $d_{S}(\phi(C))$ be the minimum symplectic weight of the linear code $\phi(C)$ over $\mathbb{F}_{p^{2m}}.$ Then $d_{S}(\phi(C))=d_{H}(C).$

Theorem 5.3.  If $C$ is a self-orthogonal code of length $N$ over $R$ with respect to the Hermitian inner product, then $\phi(C)$ is a $p^{m}$-ary self-orthogonal code of length $2N$ with respect to the symplectic inner product.

Proof   Let $\mathfrak{A}=(\textbf{q}|\textbf{r})$ and $\mathfrak{B}=(\textbf{q}^{\prime}|\textbf{r}^{\prime})$ be two codewords in $\phi(C)$. Then there exist $\textbf{c},\textbf{c}^{\prime}\in R^{n}$ and $\textbf{c}=\textbf{r}+u\textbf{q},\textbf{c}^{\prime}=\textbf{r}^{\prime}+u\textbf{q}^{\prime},$ where $\textbf{r},\textbf{q},\textbf{r}^{\prime},\textbf{q}^{\prime}\in\mathbb{F}_{p^{2m}}^{N}$. Since $C$ is self-orthogonal with respect to the Hermitian inner product, it follows that

$$\langle\textbf{c},\textbf{c}^{\prime}\rangle_{H}=(\textbf{r}+u\textbf{q})\cdot(\overline{\textbf{r}^{\prime}}-u\overline{\textbf{q}^{\prime}})=\textbf{r}\cdot\overline{\textbf{r}^{\prime}}-u(\textbf{r}\cdot\overline{\textbf{q}^{\prime}}-\overline{\textbf{r}^{\prime}}\cdot\textbf{q})=0.$$

This gives that $\textbf{r}\cdot\overline{\textbf{q}^{\prime}}-\overline{\textbf{r}^{\prime}}\cdot\textbf{q}=0$. So we get $\langle\mathfrak{A},\mathfrak{B}\rangle_{T}=0$ and $\mathfrak{B}\in\phi(C)^{\perp_{T}}$. Therefore, $\phi(C^{\bot_{H}})\subseteq\phi(C)^{\bot_{T}}$. If $C\subseteq C^{\bot_{H}}$, then $\phi(C)\subseteq\phi(C^{\bot_{H}})\subseteq\phi(C)^{\bot_{T}}$. The desired result follows. ∎

For a code $C$ of length $N$ over $R$, their torsion and residue codes are codes over $\mathbb{F}_{p^{2m}}$, defined as follows.

$$\textrm{Tor}(C)=\{\textbf{b}\in\mathbb{F}_{p^{2m}}^{N}|u\textbf{b}\in C\},\ \ \ \ \ \textrm{Res}(C)=\{\textbf{a}\in\mathbb{F}_{p^{2m}}^{N}|\exists\textbf{b}\in\mathbb{F}_{p^{2m}}^{n}:\textbf{a}+u\textbf{b}\in C\}.$$

The reduction modulo $u$ from $C$ to $\textrm{Res}(C)$ is given by $\varphi:C\rightarrow\textrm{Res}(C),$ $\varphi(\textbf{a}+u\textbf{b})=\textbf{a}.$ It is known that $\varphi$ is well defined and onto, with $\textrm{Ker}(\varphi)\cong\textrm{Tor}(C),$ and $\varphi(C)=\textrm{Res}(C).$

The code $\textrm{Res}(C)$ is the code over $\mathbb{F}_{p^{2m}}$ generated by

$$\left(\begin{array}[]{ccc}I_{k_{0}}&A&B\end{array}\right).$$

The code $\textrm{Tor}(C)$ is the code over $\mathbb{F}_{p^{2m}}$ generated by

$$\left(\begin{array}[]{ccc}I_{k_{0}}&A&B\\ 0&I_{k_{1}}&D\end{array}\right).$$

Note that $C$ has $p^{4m(2k_{0}+k_{1})}$ codewords and that $|\textrm{Res}(C)||\textrm{Tor}(C)|=p^{4m(2k_{0}+k_{1})}.$ Hence for a code $C$ over $R,$ we have $|C|=|\textrm{Res}(C)||\textrm{Tor}(C)|.$

The following Theorem can be found in Ref.[28].

Theorem 4.1 Let $d_{H}(C)$ be the Hamming distance of the linear code $C$ over $R$, $d_{H}(\emph{Tor }(C))$ be the Hamming distance of the code $\emph{Tor }(C).$ Then $d_{H}(C)=d_{H}(\emph{Tor }(C))$.

Theorem 5.5. Let $G$ be a generator matrix with $(k_{0}+k_{1})\times N$ for the linear code $C$ over $R.$ Then $C$ is a Hermitian self-orthogonal code if and only if the $(k_{0}+k_{1})\times(k_{0}+k_{1})$ matrix $G\overline{G}^{T}$ is the zero matrix.

Proof   First, suppose that a linear code $C$ has a generator matrix $G$ with $(k_{0}+k_{1})\times N$ over $R$ and $C$ is a Hermitian self-orthogonal code over $R$. Then, for any codeword $\textbf{c}\in C$, since $C$ is a Hermitian self-orthogonal code, it follows that $\textbf{c}\overline{G}^{T}=0$. Therefore, $G\overline{G}^{T}=0$.

Conversely, assume that $G$ is a generator matrix with $(k_{0}+k_{1})\times N$ for the linear code $C$ over $R$ and $G\overline{G}^{T}=0.$ For any codewords $\textbf{c}_{1}\in C$ and $\textbf{c}_{2}\in C$, there are $\textbf{s}_{1}$ and $\textbf{s}_{2}\in R^{k_{0}+k_{1}}$ such that $\textbf{c}_{1}=\textbf{s}_{1}^{T}G$ and $\textbf{c}_{2}=\textbf{s}_{2}^{T}G$. Then $\langle\textbf{c}_{1},\textbf{c}_{2}\rangle_{H}=\textbf{s}_{1}^{T}G\overline{G}^{T}\overline{\textbf{s}}_{2}=0$. Therefore, $C$ is a Hermitian self-orthogonal code.∎

The following important construction of quantum codes was proposed by Ashikhmin and Knill in Ref.[1].

Theorem 5.6. (Symplectic construction [1]) Let $\mathcal{C}$ be a $q$-ary self-orthogonal $[2N,k]$ code with respect to the trace inner product. Then there exists an $[[N,N-k,d]]_{q}$ quantum code with $d=d_{S}(\mathcal{C}^{\bot H}).$

Our goal is to construct $p^{m}$-ary self-orthogonal codes with respect to the trace inner product by employing a linear code over $R$.

Theorem 5.7. Let $C$ be a linear code over $R$ of length $N=p^{e}n$ ($n$ prime to $p$ ) with a generator matrix $G$, type $\{k_{0},k_{1}\}$ and the minimum Hamming weight of $C^{\perp H}$ be $d$. If $G\overline{G}^{T}=0,$ then there exists a $2^{m}$-ary quantum code with parameters $[[N,N-2k_{0}-k_{1},d]]_{p^{m}}$.

In the following, we will apply a Hermitian $\alpha(1+u)$-constacyclic self-orthogonal code of length $N=p^{e}n$ ($n$ prime to $p$ ) over $R$ to construct a class of $p^{m}$-ary quantum codes.

Theorem 5.8. Let $\alpha\in\mathbb{F}_{p^{2m}}$ such that $\alpha\bar{\alpha}=1$ and let

$$C=\big{\langle}\prod_{i\in I_{1}}M_{i}^{k_{i}}(x)\big{[}\prod_{l\in I_{2}}M_{l}^{r_{l}}(x)M_{-p^{m}l}^{s_{l}}(x)\big{]}\big{\rangle}$$

be an $\alpha(1+u)$-constacyclic code of length $N=p^{e}n$ over $R$, and size $p^{2k},$ where $\emph{gcd}(n,p)=1$ and $0\leq k_{i},r_{l},s_{l}\leq p^{e+1}$. Then $\emph{Tor}(C^{\bot H})$ and $\emph{Res}(C^{\bot H})$ are both $\alpha$-constacyclic codes of length $N$ over $\mathbb{F}_{p^{2m}}$ with the generator of the following form

$$\emph{Tor}(C^{\bot H})=\big{\langle}\prod_{i\in I_{1}}(M_{i}^{{\dagger}}(x))^{k_{i}-p^{e}}\big{[}\prod_{l\in I_{2}}(M_{l}^{{\dagger}}(x))^{r_{l}-\emph{min}\{p^{e},r_{l}\}}(M_{-p^{m}l}^{{\dagger}}(x))^{s_{l}-\emph{min}\{p^{e},s_{l}\}}\big{]}\big{\rangle}$$

and

$$\emph{Res}(C^{\bot H})=\big{\langle}\prod_{i\in I_{1}}(M_{i}^{{\dagger}}(x))^{p^{e+1}-k_{i}}\big{[}\prod_{l\in I_{2}}(M_{l}^{{\dagger}}(x))^{p^{e+1}-r_{l}}(M_{-p^{m}l}^{{\dagger}}(x))^{p^{e+1}-s_{l}}\big{]}\big{\rangle}.$$

Proof. By Theorem 4.8, we get

$$C^{\bot H}=\big{\langle}\prod_{i\in I_{1}}(M_{i}^{{\dagger}}(x))^{p^{e+1}-k_{i}}\big{[}\prod_{l\in I_{2}}(M_{l}^{{\dagger}}(x))^{p^{e+1}-r_{l}}(M_{-p^{m}l}^{{\dagger}}(x))^{p^{e+1}-s_{l}}\big{]}\big{\rangle}.$$

By the definition of torsion code, we have

$$\textrm{Tor}(C^{\bot H})=\big{\langle}\prod_{i\in I_{1}}(M_{i}^{{\dagger}}(x))^{\textrm{min}\{p^{e+1}-k_{i},p^{e+1}\}-\textrm{min}\{p^{e+1}-k_{i},p^{e}\}}$$

$$\big{[}\prod_{l\in I_{2}}(M_{l}^{{\dagger}}(x))^{\textrm{min}\{p^{e+1}-r_{l},p^{e+1}\}-\textrm{min}\{p^{e+1}-r_{l},p^{e}\}}(M_{-p^{m}l}^{{\dagger}}(x))^{\textrm{min}\{p^{e+1}-s_{l},p^{e+1}\}-\textrm{min}\{p^{e+1}-s_{l},p^{e}\}}\big{]}\big{\rangle}$$

and $\textrm{Tor}(C^{\bot H})$ is an $\alpha$-constacyclic code of length $N$ over $\mathbb{F}_{p^{2m}}.$ Since $0\leq r_{l},s_{l}\leq p^{e+1}$ and $p^{e}\leq k_{i}\leq p^{e+1},$ it follows that

$$\textrm{Tor}(C^{\bot H})=\big{\langle}\prod_{l\in I_{2}}(M_{l}^{{\dagger}}(x))^{p^{e+1}-r_{l}-\textrm{min}\{p^{e},r_{l}\}}(M_{-p^{m}l}^{{\dagger}}(x))^{p^{e+1}-s_{l}-\textrm{min}\{p^{e},s_{l}\}}\big{\rangle}.$$

Similarly, we can get

$$\textrm{Res}(C^{\bot H})=\big{\langle}\prod_{i\in I_{1}}(M_{i}^{{\dagger}}(x))^{p^{e+1}-k_{i}}\big{[}\prod_{l\in I_{2}}(M_{l}^{{\dagger}}(x))^{p^{e+1}-r_{l}}(M_{-p^{m}l}^{{\dagger}}(x))^{p^{e+1}-s_{l}}\big{]}\big{\rangle}$$

and $\textrm{Res}(C^{\bot H})$ is an $\alpha$-constacyclic code of length $N$ over $\mathbb{F}_{p^{2m}}.$ ∎

By Theorems 4.9 and 5.8, we see that a trace self-orthogonal code of length $2N$ over $\mathbb{F}_{p^{2m}}$ can be constructed from a Hermitian $\alpha(1+u)$-constacyclic self-orthogonal code of length $N$ over $R$ under the map $\phi$. According to Theorem 5.6, we can obtain a class of $p^{m}$-ary quantum codes.

Theorem 5.9. Let $\alpha\in\mathbb{F}_{p^{2m}}$ such that $\alpha\bar{\alpha}=1$ and let

$$C=\big{\langle}\prod_{i\in I_{1}}M_{i}^{k_{i}}(x)\big{[}\prod_{l\in I_{2}}M_{l}^{r_{l}}(x)M_{-p^{m}l}^{s_{l}}(x)\big{]}\big{\rangle}$$

be an $\alpha(1+u)$-constacyclic code of length $N=p^{e}n$ over $R$, and size $p^{2k},$ where $\emph{gcd}(n,p)=1$ and $0\leq k_{i},r_{l},s_{l}\leq p^{e+1}$. Suppose that $p^{e}\leq k_{i}\leq p^{e+1}$ and $p^{e+1}\leq r_{l}+s_{l}\leq p^{e+2}$. Then there exists a $p^{m}$-ary quantum code with parameters $[[N,N-k,d]]_{p^{m}},$ where $d$ is the Hamming distance of the code $\emph{Tor}(C^{\bot H}).$

Next, let us use some examples to illustrate our construction method. We are with the help of the computer algebra system MAGMA to find quantum codes with good parameters. We construct some $3$-ary quantum codes by taking advantage of Hermitian $\omega^{4}(1+u)$-constacyclic self-orthogonal codes over $\mathbb{F}_{3^{2}}+u\mathbb{F}_{3^{2}}.$

Example 5.10. Let $\mathbb{F}_{3^{2}}=\{0,1,\omega,\omega^{2},\ldots,\omega^{7}\}$ be a finite field with nine elements, where $\omega^{2}=\omega+1.$ Consider a Hermitian $\omega^{4}(1+u)$-constacyclic self-orthogonal code with length $5$ over $\mathbb{F}_{3^{2}}+u\mathbb{F}_{3^{2}}.$ Let $C$ be an $\omega^{4}(1+u)$-constacyclic code of length $5$ over $\mathbb{F}_{3^{2}}+u\mathbb{F}_{3^{2}}.$ Then $\beta=(\omega^{4})^{3^{2-1}}=\omega^{4}.$
Note that

$$x^{5}-\omega^{4}=f_{0}(x)f_{1}(x)f_{2}(x)$$

over $\mathbb{F}_{3^{2}}+u\mathbb{F}_{3^{2}},$ where $f_{0}(x)=x+1,$ $f_{1}(x)=x^{2}+\omega^{5}x+1,$ $f_{2}(x)=x^{2}+\omega^{7}x+1.$
Let

$$C=\left\langle(f_{0}(x))^{2}(f_{2}(x))^{2}\right\rangle.$$

By Theorem 4.12, we get $C$ is a Hermitian $\omega^{4}(1+u)$-constacyclic self-orthogonal code over $\mathbb{F}_{3^{2}}+u\mathbb{F}_{3^{2}}$ of length $5$ with

$$C^{\bot H}=\left\langle(f_{2}(x))^{2}\right\rangle,$$

where $f_{2}(x)=f_{1}^{{\dagger}}(x).$ By Theorem 5.8, we obtain $\textrm{Tor}(C^{\bot H})=\big{\langle}(f_{2}(x)\big{\rangle}.$ Computation of the Hamming distances of $\textrm{Tor}(C^{\bot H})$ with the help of the computer algebra system MAGMA, we get the Hamming distance of $C^{\bot H}$ is $3.$ Using the Symplectic construction, we can obtain a $Q=[[5,1,3]]_{3}$ quantum code, which is a quantum MDS code. Thus, our obtained $3$-ary quantum code is optimal.

Example 5.11. Consider a Hermitian $\omega^{4}(1+u)$-constacyclic self-orthogonal code with length $10$ over $\mathbb{F}_{3^{2}}+u\mathbb{F}_{3^{2}}.$ Let $C$ be an $\omega^{4}(1+u)$-constacyclic code of length $10$ over $\mathbb{F}_{3^{2}}+u\mathbb{F}_{3^{2}}.$
Note that

$$x^{10}-\omega^{4}=f_{0}(x)f_{1}(x)f_{2}(x)f_{3}(x)f_{4}(x)f_{5}(x)$$

over $\mathbb{F}_{3^{2}}+u\mathbb{F}_{3^{2}},$ where $f_{0}(x)=x+\omega^{2},$ $f_{1}(x)=x+\omega^{6},$ $f_{2}(x)=x^{2}+\omega x+2,$ $f_{3}(x)=x^{2}+\omega^{3}x+2,$ $f_{4}(x)=x^{2}+\omega^{5}x+2,$ $f_{5}(x)=x^{2}+\omega^{7}x+2.$
Let

$$C=\left\langle(f_{0}(x))^{2}(f_{2}(x))^{2}(f_{3}(x))^{2}(f_{4}(x))^{2}\right\rangle.$$

By Theorem 4.12, we get $C$ is a Hermitian $\omega^{4}(1+u)$-constacyclic self-orthogonal code over $\mathbb{F}_{3^{2}}+u\mathbb{F}_{3^{2}}$ of length $10$ with

$$C^{\bot H}=\left\langle(f_{0}(x))^{2}(f_{4}(x))^{2}\right\rangle,$$

where $f_{0}(x)=f_{1}^{{\dagger}}(x),f_{4}(x)=f_{5}^{{\dagger}}(x).$ By Theorem 5.8, we obtain $\textrm{Tor}(C^{\bot H})=\big{\langle}f_{0}(x)f_{4}(x)\big{\rangle}.$ Computation of the Hamming distances of $\textrm{Tor}(C^{\bot H})$ with the help of the computer algebra system MAGMA, we get the Hamming distance of $C^{\bot H}$ is $4.$ Using the Symplectic construction, we can obtain a $Q=[[10,4,4]]_{3}$ quantum code, which is a quantum MDS code. Thus, our obtained $3$-ary quantum code is optimal.

We list some quaternary quantum codes which can be constructed starting from Hermitian $\omega^{4}(1+u)$-constacyclic self-orthogonal codes over $\mathbb{F}_{3^{2}}+u\mathbb{F}_{3^{2}}$ in Table I. Compared the parameters of quantum codes available in Ref.[13], we find that all of our obtained quantum codes are optimal.

TABLE I   Some $3$-ary quantum codes from $\omega^{4}(1+u)$-constacyclic codes over $\mathbb{F}_{3^{2}}+u\mathbb{F}_{3^{2}}$ Length Generator polynomial of $C^{\bot H}$ Hamming distance of $C^{\bot H}$ Quantum codes 14 $(x+\omega^{2})^{2}(x^{3}+\omega^{3}x^{2}+\omega^{3}x+\omega^{2})^{2}$ 4 $[[14,6,4]]_{3}$ 20 $(x+\omega^{3})^{2}(x^{2}+\omega^{2}x+\omega^{2})^{2}$ 3 $[[20,14,3]]_{3}$ 20 $(x+\omega^{3})^{2}(x+\omega^{5})^{2}(x^{2}+\omega^{6}x+\omega^{6})^{2}$ 4 $[[20,12,4]]_{3}$ 40 $(x^{2}+\omega^{5})^{2}(x^{2}+\omega^{3}x+\omega^{7})^{2}(x^{2}+\omega^{5}x+\omega^{3})^{2}(x^{2}+\omega^{7}x+\omega)^{2}$ 6 $[[40,24,6]]_{3}$ 61 $(x^{5}+\omega^{5}x^{3}+\omega^{7}x^{2}+1)^{2}$ 4 $[[61,51,4]]_{3}$ 70 $(x+\omega^{6})^{2}(x^{2}+\omega^{5}x+2)^{2}(x^{3}+\omega x^{2}+\omega x+\omega^{6})^{2}$ 4 $[[70,58,4]]_{3}$ 82 $(x+\omega^{6})^{2}(x^{4}+\omega^{7}x^{3}+\omega x+\omega^{3}x+1)^{2}$ 4 $[[82,72,4]]_{3}$

Code Comparisons 5.12.   We obtain the first three quantum codes $[[14,6,4]]_{3},$ $[[20,14,3]]_{3}$ and $[[20,12,4]]_{3}$ satisfy $2d=n-k.$ Thus, they are all optimal. Compared our obtained the other four quantum codes in Table I with previously known quantum codes in Ref.[13]. We find that our obtained quantum codes $[[40,24,6]]_{3},$ $[[61,51,4]]_{3},$ $[[70,58,4]]_{3},$ and $[[82,72,4]]_{3}$ all meet their lower bounds in Ref.[13]. Thus, they are also optimal. The example shows that optimal quantum codes also can be constructed from constacyclic codes over finite chain rings.

6 Conclusion

In this paper, we define a class of Gray maps from $\mathbb{F}_{2^{2m}}+u\mathbb{F}_{2^{2m}}$ to $\mathbb{F}_{2^{2m}}^{2},$ which preserves the Hermitian self-orthogonal property of linear codes. By employing Hermitian construction, a new class $2^{m}$-ary quantum codes are constructed from the Hermitian $\alpha(1+u)$-constacyclic self-orthogonal codes over $\mathbb{F}_{2^{2m}}+{u}\mathbb{F}_{2^{2m}}.$ We also define another class of maps from $\mathbb{F}_{p^{2m}}+u\mathbb{F}_{p^{2m}}$ to $\mathbb{F}_{p^{2m}}^{2},$ which changes the Hermitian self-orthogonal property of linear codes over $R$ into the trace self-orthogonal property of linear codes over $\mathbb{F}_{p^{2m}}$. By using the Symplectic construction, a new class of $p^{m}$-ary quantum codes are constructed from the Hermitian $\alpha(1+u)$-constacyclic self-orthogonal codes over $\mathbb{F}_{p^{2m}}+{u}\mathbb{F}_{p^{2m}}.$ Our research showed that quantum codes with good parameters also can be constructed from the Hermitian $\alpha(1+u)$-constacyclic self-orthogonal codes over $R$. Our results also enrich the variety of quantum error-correcting codes. It would be interesting to encourage more researchers to find more quantum codes with good parameters from constacyclic code over finite chain rings and to apply them in practice.

Acknowledgement

The authors would like to thank Doctor Sun Zhonghua who gave many helpful suggestions and comments to greatly improve the presentation of the paper.The authors also wish to thank the editor and the anonymous referee whose comments have greatly improved this paper.

Availability of data and materials

Data sharing not applicable to this article as no datasets were generated or analysed during the current study.

Declaration of competing interest

The authors declare that we have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.