An upper bound on the number of perfect quadratic forms

Abstract: In a recent preprint on arXiv Roland Bacher showed that the number

p_{d}

of non-similar perfect

d

-dimensional quadratic forms satisfies

e^{O m e g a (d)} < p_{d} < e^{O (d^{3} l o g (d))}

. We improve the upper bound to

e^{O (d^{2} l o g (d))}

by a volumetric argument based on Voronoi's first reduction theory.

Review text

A positive definite quadratic form $\mathcal Q$ of $d$ variables is called \textit{perfect} if $\{xx^{t} \ |\ x\in {\mathbb Z}^{d}\), $\mathcal Q(x) = \min(\mathcal{Q}) \} $ spans the vector space $ \mathbb{R}^{d(d+1)/2} $ (here $ \min(\mathcal{Q}) = \min \{\mathcal{Q}(x) \ | \ 0\ne x \in {\mathbb Z}^{d} \} $). Two positive quadratic forms $ \mathcal{Q},\mathcal{Q'} $ of $ d $ variables are called \textit{similar}, if there exist $ M\in\mathrm{GL}_d({\mathbb Z}) $ and $ \lambda\in {\mathbb R}_{>0} $ such that $ \mathcal{Q}'(x) = \lambda^2 \mathcal{Q}(Mx) $ for all $ x\in {\mathbb Z}^{d} $ (if $ \lambda =1 $, then $ \mathcal{Q} $ and $ \mathcal{Q}' $ are called \textit{arithmetically equivalent}). \textit{R. Bacher} showed that the number $ p_d $ of non-similar perfect positive definite quadratic forms of $ d $ variables satisfies $$ e^{d^{1-\varepsilon}} < p_d < e^{d^{3+\varepsilon}} $$ for arbitrary small positive $ \varepsilon $ (see [J. Théor. Nombres Bordx. 30, No. 3, 917--945 (2018; Zbl 1443.11132)]). Denote by $ Q\in {\mathbb R}^{d\times d} $ the matrix associated to the positive definite quadratic form $ \mathcal{Q} $. Furthermore, let $$ \mathcal{S}_{\geqslant 0}^{d} = \{Q \in {\mathbb R}^{d\times d} \colon\mathcal{Q}\text{ is positive semidefinite} \}. $$ In this paper, the author proves the following upper bound $ p_d \leqslant e^{d^2 \log(d)} $ (Theorem 3.1). The proof bases on the Voronoi's first reduction theory (see [\textit{G. Vorono\"{i}}, J. Reine Angew. Math. 133, 97--178 (1908; JFM 38.0261.01)]) and on the volumetric argument applied to the set $ \{ Q \in \mathcal{S}_{\geqslant 0}^{d} \colon\mathrm{Tr}(Q)\leqslant 1 \} $. Moreover, the author shows (Lemma 3.3) that if $ \mathcal{Q} $ is a perfect quadratic form, then there exists a quadratic form $ \mathcal{Q}' $, such that $ \mathcal{Q}' $ arithmetically equivalent to $ \mathcal{Q} $ and $ \{x^{t}x \ |\ x\in {\mathbb Z}^{d},\ \ \mathcal{Q}'(x) = \min(\mathcal{Q}') \} \subseteq (0,d^3(d+7)/8] $. Moreover, if $ \mathcal{Q} $ is an integral primitive perfect quadratic form of $ d $ variables, then $ \min(\mathcal{Q})\leqslant 2^{-(n+d/2)}\cdot (d^3(d+7))^{n/2} $ where $ n = d(d+1)/2 $ (Theorem 4.1).

Recommendations

Cites work

Cited in

(8)

This page was built for publication: An upper bound on the number of perfect quadratic forms

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2308288)