Factorization of Additive Polynomials and van der Geer–van der Vlugt curves in characteristic 2 2

Tetsushi Ito, Daichi Takeuchi, Takahiro Tsushima · 2026

Read Factorization of Additive Polynomials and van der Geer–van der Vlugt curves in characteristic 2 2 by Tetsushi Ito, Daichi Takeuchi, Takahiro Tsushima (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-16729.txt.

Introduction

Ito, Takeuchi and Tsushima give a simpler formula for Frobenius eigenvalues of van der Geer-van der Vlugt curves in characteristic 2, via factorization of additive polynomials. An earlier formula from quotients of the curve was explicit but complicated by many choices. The new approach yields a simpler formula useful for explicit computations. As applications, they provide a method for constructing maximal and minimal such curves, show every such curve arises from the construction, and compute examples while studying their periods.

Catalog record

CREATORTetsushi Ito, Daichi Takeuchi, Takahiro Tsushima
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE135509
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2605.16729

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Factorization of Additive Polynomials and van der Geer–van der Vlugt curves in characteristic 2 2 Tetsushi Ito, Daichi Takeuchi, Takahiro Tsushima Abstract . In our previous work, we gave a formula for the Frobenius eigenvalues of van der Geer–van der Vlugt curves in characteristic 2 2 by considering suitable quotients of the curve. Although the formula is explicit, it depends on many choices, which makes the formula complicated. In this article, we take a different approach using a factorization of additive polynomials, and prove a new formula. The resulting formula is simpler and is useful for explicit computations. As applications, we provide a method for constructing maximal and minimal van der Geer–van der Vlugt curves, and show that every such curve arises from this construction. We also compute various examples of van der Geer–van der Vlugt curves and study their periods. 2020 Mathematics Subject Classification Primary: 14F20, 14G15; Secondary: 14H25, 14G10. 1. Introduction Let p_{0} be a prime, p a power of p_{0} , and q a power of p . Let R(x)\in{\mathbb{F}}_{q}[x] be an {\mathbb{F}}_{p} -linearized polynomial, namely a polynomial of the form R(x)=\sum_{i=0}^{e}a_{i}x^{p^{i}}\in{\mathbb{F}}_{q}[x]. Assume that e\geq 1 and a_{e}\neq 0 . We consider the smooth plane curve over {\mathbb{F}}_{q} defined by C_{R}:y^{p}-y=xR(x). We denote by \overline{C}_{R} its smooth compactification. We call C_{R} and \overline{C}_{R} the van der Geer–van der Vlugt curves associated with R(x) . Explicit computation of the Frobenius eigenvalues of van der Geer–van der Vlugt curves is an interesting problem in number theory and coding theory: see [ 1 , 3 , 5 , 6 , 7 , 12 ] . In [ 6 ] and [ 12 ] , when p_{0} is odd, a formula for the Frobenius eigenvalues was given in terms of quadratic Gauss sums. The case p_{0}=2 is excluded there, which reflects the fact that the quadratic Gauss sum is not defined when p_{0}=2 . For this reason, the Frobenius eigenvalues of van der Geer–van der Vlugt curves are more mysterious, especially in characteristic 2 . From now on, we assume that p_{0}=2 . In [ 7 ] , the authors studied the case p_{0}=2 and proved a formula for the Frobenius eigenvalues of van der Geer–van der Vlugt curves in characteristic 2 . However, the formula proved in [ 7 ] depends on many choices, which makes the formula complicated. In this article, we take another approach to the Frobenius eigenvalues and prove a new formula. The resulting formula is simpler and is useful for explicit computations. As applications, we determine the maximal and minimal curves among a family of twists of van der Geer–van der Vlugt curves. We also study the periods and parities of van der Geer–van der Vlugt curves. We describe our results in more detail. In [ 7 ] , the curve C_{R} is studied through an action of a Heisenberg group H_{R} . In loc. cit., we fix a maximal abelian subgroup A\subset H_{R} , and associate an {\mathbb{F}}_{p} -linearized polynomial F_{A} . The formula proved in [ 7 ] is described using such a datum (A,F_{A}) . In this article, we start with an {\mathbb{F}}_{p} -linearized polynomial F , which serves as F_{A} . It turns out that this approach simplifies the computation of the Frobenius eigenvalues, and hence the resulting formula. Note that this computation does not require detailed group-theoretic properties of H_{R} . In this paper, we start with a pair (F,t) where F(x)=\sum_{i=0}^{e}b_{i}x^{p^{i}}\in{\mathbb{F}}_{q}[x] is an {\mathbb{F}}_{p} -linearized polynomial and t\in{\mathbb{F}}_{q} . We further impose the following conditions on F . Let {\mathbb{F}} be an algebraic closure of {\mathbb{F}}_{q} . For each x\in{\mathbb{F}} , define F^{\ast}(x):=\sum_{i=0}^{e}(b_{i}x)^{p^{-i}}\in{\mathbb{F}}. Such an F^{\ast} is called the adjoint of the {\mathbb{F}}_{p} -linearized polynomial F . The adjoint F^{\ast} defines an {\mathbb{F}}_{p} -linear map {\mathbb{F}}\to{\mathbb{F}} , and we denote by W_{F}^{\ast} its kernel. (i) We have e\geq 1 and b_{0}b_{e}\neq 0 . (ii) We have F^{\ast}(1)=0 . (iii) We have W_{F}^{\ast}\subset{\mathbb{F}}_{q} . For such a pair (F,t) , define \displaystyle R_{F}(x):= \displaystyle\sum_{i=1}^{e}a_{i}x^{p^{i}}\in{\mathbb{F}}_{q}[x],\quad\textrm{with }a_{i}:=\sum_{j=0}^{e-i}(b_{j}b_{j+i})^{p^{-j}}, \displaystyle\alpha_{F}(t):= \displaystyle F^{\ast}(t)^{2}+\sum_{0\leq i<j\leq e}b_{i}^{p^{-i}}b_{j}^{p^{-j}}\in{\mathbb{F}}_{q}. We also set R_{F,t}(x):=R_{F}(x)+\alpha_{F}(t)x. We then associate to (F,t) a smooth affine {\mathbb{F}}_{q} -curve D_{F,t}\colon y^{p}-y=xR_{F,t}(x), which is one of the main objects of study in this paper. In our previous notation, the curve D_{F,t} is denoted by C_{R_{F,t}} . If one uses the adjoint operator f\mapsto f^{\ast} for {\mathbb{F}}_{p} -linearized polynomials, whose precise definition is recalled in Definition 2.3 , then one can verify that R_{F} satisfies R_{F}+R_{F}^{\ast}=F^{\ast}\circ F. The new insight presented in this article is that, for a given {\mathbb{F}}_{p} -linearized polynomial R , a factorization of the form R+R^{\ast}=F^{\ast}\circ F allows one to compute the Frobenius eigenvalues of C_{R} explicitly. To some extent, one may say that the curves D_{F,t} form a sufficiently large subclass of the van der Geer–van der Vlugt curves. For instance, every {\mathbb{F}}_{q} -maximal or {\mathbb{F}}_{q} -minimal van der Geer–van der Vlugt curve is isomorphic to D_{F,t} for some pair (F,t) . See Theorem 1.3 . In fact, for a given R(x) , we give a necessary and sufficient condition for R to admit a pair (F,t) satisfying R=R_{F,t} . For the precise statement, see Theorem 4.14 ; see also Corollary 4.16 and Theorem 5.11 . In what follows, we highlight some of the main results of this paper. 1.1. An Explicit Formula for the Frobenius Eigenvalues of D_{F,t} Let (F,t) be a pair as above, and let D_{F,t} be the associated curve. Let \overline{D}_{F,t} denote the smooth compactification of D_{F,t} . The L -polynomial of \overline{D}_{F,t} is described as follows. Let W_{2} denote the Witt vector scheme of length two. Let \ell be an odd prime and let Q_{q}\colon{\mathbb{F}}_{q}\xrightarrow{x\mapsto(x,0)}W_{2}({\mathbb{F}}_{q})\xrightarrow{\Tr_{q/2}}W_{2}({\mathbb{F}}_{2})\xrightarrow{\xi_{2}}\overline{\mathbb{Q}}^{\times}_{\ell}, where \Tr_{q/2}\colon W_{2}({\mathbb{F}}_{q})\to W_{2}({\mathbb{F}}_{2}) denotes the trace map and \xi_{2} is a fixed faithful character of the group W_{2}({\mathbb{F}}_{2})\simeq\mathbb{Z}/4\mathbb{Z} . Set \sqrt{-1}:=\xi_{2}(1,0) , which is a primitive 4 -th root of unity. The function Q_{q} is an even-characteristic analogue of the quadratic form \frac{1}{2}x^{2} , in the sense that Q_{q}(x+y)Q_{q}(x)^{-1}Q_{q}(y)^{-1}=(-1)^{\Tr_{{\mathbb{F}}_{q}/{\mathbb{F}}_{2}}(xy)}. Theorem 1.1 (Theorem 4.12 ) . The L -polynomial L(\overline{D}_{F,t}/{\mathbb{F}}_{q},T):=\det(1-{\rm Fr}_{q}T\mid H^{1}(\overline{D}_{F,t,{\mathbb{F}}},\overline{\mathbb{Q}}_{\ell})) satisfies L(\overline{D}_{F,t}/{\mathbb{F}}_{q},T)=\prod_{v\in W_{F}^{\ast}}(1-\tau_{v}T)^{p-1}, where \tau_{v}:=Q_{q}(t+v)^{-1}(-1-\sqrt{-1})^{[{\mathbb{F}}_{q}:{\mathbb{F}}_{2}]} . The above formula is simpler than the one obtained in [ 7 , Theorem 1.1] . In loc. cit., we impose a certain condition on C_{R} . One can show that this condition is equivalent to the existence of a pair (F,t) as above such that R(x)=R_{F,t}(x) . See Theorem 4.14 for details. 1.2. Maximal and Minimal Twists Fix a polynomial R(x)=\sum_{i=1}^{e}a_{i}x^{p^{i}} with a_{i}\in{\mathbb{F}}_{q} and a_{e}\neq 0 . For each a\in{\mathbb{F}}_{q} , set R_{a}(x):=R(x)+ax , and consider the van der Geer–van der Vlugt curve C_{a}:=C_{R_{a}} . Its compactification is denoted by \overline{C}_{a} . The curves C_{a} , \overline{C}_{a} are called the twists of C_{0} , \overline{C}_{0} , respectively. We regard C_{a} as a family of van der Geer–van der Vlugt curves parametrized by a . Let T_{R,\max} (resp. T_{R,\min} ) be the set of a\in{\mathbb{F}}_{q} for which the twist \overline{C}_{a} is {\mathbb{F}}_{q} -maximal (resp. {\mathbb{F}}_{q} -minimal). Here, a smooth projective curve over {\mathbb{F}}_{q} is said to be {\mathbb{F}}_{q} -maximal (resp. {\mathbb{F}}_{q} -minimal ) if it attains the Hasse–Weil upper bound (resp. lower bound). We also say that \overline{C}_{a} is {\mathbb{F}}_{q} -extremal if it is either {\mathbb{F}}_{q} -maximal or {\mathbb{F}}_{q} -minimal. For the precise definition, see

Read the complete text