arXiv subjects

B. Sury

Publications and source records attributed to B. Sury.

18 recordsLinked to original sources

arXiv2024

Gerth's heuristics for a family of quadratic extensions of certain Galois number fields

Gerth generalised Cohen-Lenstra heuristics to the prime $p=2$. He conjectured that for any positive integer $m$, the limit $$ \lim_{x \to \infty} \frac{\sum_{0 < D \le X, \atop{ \text{squarefree} }} |{\rm Cl}^2_{\Q(\sqrt{D})}/{\rm Cl}^4_{\Q(\sqrt{D})}|^m}{\sum_{0 < D \le X, \atop{ \text{squarefree} }} 1} $$ exists and proposed a value for the limit. Gerth's conjecture was proved by Fouvry and Kluners in 2007. In this paper, we generalize their result by obtaining lower bounds for the average value of $|{\rm Cl}^2_{\L}/{\rm Cl}^4_{\L}|^m$, where $\L$ varies over an infinite family of quadratic extensions of certain Galois number fields. As a special case of our theorem, we obtain lower bounds for the average value when the base field is any Galois number field with class number $1$ in which $2\Z$ splits.

math.NT↗

arXiv2024

Davenport constant and its variants for some non-abelian groups

We define two variants $e(G)$, $f(G)$ of the Davenport constant $d(G)$ of a finite group $G$, that is not necessarily abelian. These naturally arising constants aid in computing $d(G)$ and are of potential independent interest. We compute the constants $d(G)$, $e(G)$, $f(G)$ for some nonabelian groups G, and demonstrate that, unlike abelian groups where these constants are identical, they can each be distinct. As a byproduct of our results, we also obtain some cases of a conjecture of J. Bass. We compute the $k$-th Davenport constant for several classes of groups as well. We also make a conjecture on $f(G)$ for metacyclic groups and provide evidence towards it.

math.CO↗

arXiv2024

Linear Congruences in several variables with congruence restrictions

In this article, we consider systems of linear congruences in several variables and obtain necessary and sufficient conditions as well as explicit expressions for the number of solutions subject to certain restriction conditions. These results are in terms of Ramanujan sums and generalize the results of Lehmer \cite{DNL13} and Bibak et al. \cite{BBVRL17}. These results have analogues over $\mathbb{F}_q[t]$ where the proofs are similar, once notions such as Ramanujan sums are defined in this set-up. We use the recent description of Ramanujan sums over function fields as developed by Zhiyong Zheng \cite{ZZ18}. This is discussed in the last section. We illustrate the formulae obtained for the number of solutions through some examples. Over the integers, such problems have a rich history, some of which seem to have been forgotten - a number of papers written on the topic re-prove known results. The present authors also became aware of some of these old articles only while writing the present article and hence, we recall very briefly some of the old work by H. J. S. Smith, Rademacher, Brauer, Butson and Stewart, Ramanathan, McCarthy, and Spilker \cite{AB26, BS55, PJM76, HR25, KGR44, HJSS61, JS96}.

math.NT↗

arXiv2024

Linear Congruences and a Conjecture of Bibak

We address three questions posed by Bibak \cite{KB20}, and generalize some results of Bibak, Lehmer and K G Ramanathan on solutions of linear congruences $\sum_{i=1}^k a_i x_i \equiv b \Mod{n}$. In particular, we obtain explicit expressions for the number of solutions where $x_i$'s are squares modulo $n$. In addition, we obtain expressions for the number of solutions with order restrictions $x_1 \geq \cdots \geq x_k$ or, with strict order restrictions $x_1> \cdots > x_k$ in some special cases. In these results, the expressions for the number of solutions involve Ramanujan sums and are obtained using their properties.

math.NT↗

arXiv2023

Polynomial identities and Fermat quotients

We prove some polynomial identities from which we deduce congruences modulo $p^2$ for the Fermat quotient $\frac{2^p-2}{p}$ for any odd prime $p$ (Proposition 1 and Theorem 1). These congruences are simpler than the one obtained by Jothilingam in 1985 which involves listing quadratic residues in some order. On the way, we also observe some more congruences for the Fermat quotient that generalize Eisenstein's classical congruence. Using such polynomial identities, we obtain some sums involving harmonic numbers. We also prove formulae for binomial sums of harmonic numbers of higher order.

math.NT↗

arXiv2023

Binary Cubic Forms and Rational Cube Sum Problem

In this note, we use integral binary cubic forms to study the rational cube sum problem. We prove (unconditionally) that for any positive integer $d$, infinitely many primes in each of the residue classes $ 1 \pmod {9d}$ as well as $ -1 \pmod {9d}$, are sums of two rational cubes. Among other results, we prove that every non-zero residue class $a \pmod {q}$, for any prime $q$, contains infinitely many primes which are sums of two rational cubes. Further, for an arbitrary integer $N$, we show there are infinitely many primes $p$ in each of the residue classes $ 8 \pmod 9$ and $1 \pmod 9$, such that $Np$ is a sum of two rational cubes.

math.NT↗

arXiv2022

Cyclic Cubic Extensions of Q

In this article we explicitly describe irreducible trinomials X^3-aX+b which gives all the cyclic cubic extensions of Q. In doing so, we construct all integral points (x,y,z) with GCD(y,z)=1, of the curves X^2+3Y^2 = 4DZ^3 and X^2+27Y^2=4DZ^3 as D varies over cube-free positive integers. We parametrise these points using well known parametrisation of integral points (x,y,z) of the curve X^2+3Y^2=4Z^3 with GCD(y,z)=1. As an accidental byproduct of our result we show that there are infinitely many primes congruent to 1 or 8 modulo 9, can be expressed as sum of two rational cubes.

math.NT↗

arXiv2021

Fruit Diophantine Equation

We show that the Diophantine equation given by X^3+ XYZ = Y^2+Z^2+5 has no integral solution. As a consequence, we show that the family of elliptic curve given by the Weierstrass equations Y^2-kXY = X^3 - (k^2+5) has no integral point.

math.HO↗

arXiv2021

Degrees of Faithful Irreducible Representations of Certain Metabelian Groups and a Question of Sim

In this paper, we answer affirmatively a question of H S Sim on representations in characteristic $0$, for a class of metabelian groups. Moreover, we provide examples to point out that the analogous answer is no longer valid if the solvable group has derived length larger than 2. Let $F$ be a field of characteristic $0$ and $\overline{F}$ be its algebraic closure. We prove that if $G$ is a finite metabelian group containing a maximal abelian normal subgroup which is a p-group with abelian quotient, all possible faithful irreducible representations over $F$ have the same degree and that the Schur index of any faithful irreducible $\overline{F}$-representation with respect to $F$ is always $1$ or $2$. H S Sim had proven such a result for metacyclic groups when the characteristic of $F$ is positive and posed the question in characteristic $0$. Our result answers this question for the above class of metabelian groups affirmatively. We also determine explicitly the Wedderburn component corresponding to any faithful irreducible $\overline{F}$-representation in the group algebra $F[G]$.

math.RT↗

arXiv2021

Rational and Quasi-Permutation Representations of Holomorph of Cyclic p-Groups

For a finite group $G$, let $p(G)$ denote the minimal degree of a faithful permutation representation of $G$. The minimal degree of a faithful representation of $G$ by quasi-permutation matrices over the fields $\mathbb{C}$ and $\mathbb{Q}$ are denoted by $c(G)$ and $q(G)$ respectively. In general $c(G)\leq q(G)\leq p(G)$ and either inequality may be strict. In this paper, we study the representation theory of the group $G =$ Hol$(C_{p^{n}})$, which is the holomorph of a cyclic group of order $p^n$, $p$ a prime. This group is metacyclic when $p$ is odd and metabelian but not metacyclic when $p=2$ and $n \geq 3$. We explicitly describe the set of all isomorphism types of irreducible representations of $G$ over the field of complex numbers $\mathbb{C}$ as well as the isomorphism types over the field of rational numbers $\mathbb{Q}$. We compute the Wedderburn decomposition of the rational group algebra of $G$. Using the descriptions of the irreducible representations of $G$ over $\mathbb{C}$ and over $\mathbb{Q}$, we show that $c(G) = q(G) = p(G) = p^n$ for any prime $p$. The proofs are often different for the case of $p$ odd and $p=2$.

math.RT↗

arXiv2019

Some Observations on Khovanskii's Matrix Methods for extracting Roots of Polynomials

In this article we apply a formula for the $n$-th power of a $3\times 3$ matrix (found previously by the authors) to investigate a procedure of Khovanskii's for finding the cube root of a positive integer. We show, for each positive integer $α$, how to construct certain families of integer sequences such that a certain rational expression, involving the ratio of successive terms in each family, tends to $α^{1/3}$. We also show how to choose the optimal value of a free parameter to get maximum speed of convergence. We apply a similar method, also due to Khovanskii, to a more general class of cubic equations, and, for each such cubic, obtain a sequence of rationals that converge to the real root of the cubic. We prove that Khovanskii's method for finding the $m$-th ($m \geq 4$) root of a positive integer works, provided a free parameter is chosen to satisfy a very simple condition. Finally, we briefly consider another procedure of Khovanskii's, which also involves $m \times m$ matrices, for approximating the root of an arbitrary polynomial of degree $m$.

math.NT↗

arXiv2018

Powers of a matrix and combinatorial identities

In this article we obtain a general polynomial identity in $k$ variables, where $k\geq 2$ is an arbitrary positive integer. We use this identity to give a closed-form expression for the entries of the powers of a $k \times k$ matrix. Finally, we use these results to derive various combinatorial identities.

math.CO↗

arXiv2016

A Dedekind Domain with Nontrivial Class Group

Analytic properties of function spaces over the real and the complex fields are different in some ways. This reflects in algebraic properties which are different at times and similar in some other respects. For instance, the ring of real-valued continuous functions on a closed interval like $[0,1]$ behaves similarly to the corresponding ring of complex-valued functions; they depend only on the topology of $[0,1]$. The ring $\mathbf{R}[X,Y]/(X^2+Y^2-1)$ of real-valued polynomial functions on the unit circle is not a unique factorization domain - witness the equation $$\cos^2(t) = (1+ \sin(t))(1- \sin(t)).$$ On the other hand, the ring $\mathbf{C}[X,Y]/(X^2+Y^2-1) \cong \mathbf{C}[X+iY, 1/(X+iY)]$ is a principal ideal domain. Again, the rings of convergent power series (over either of these fields) with radius of convergence larger than some number $\rho$ is a Euclidean domain (and hence, a principal ideal domain) - this can be seen by using for a Euclidean "norm" function, the function which counts zeroes (with multiplicity) in the disc $|z| \leq \rho$. In this note, we consider the rings $C_{an}(S^1;\mathbf{R})$ of real-analytic functions on the unit circle $\mathit{S}^1$ which are real-valued and the corresponding ring $C_{an}(S^1; \mathbf{C})$ of analytic functions that are complex-valued. We will see that the latter is a principal ideal domain while the former is a Dedekind domain which is not a principal ideal domain - the class group having order $2$.

math.RA↗

arXiv2011

Gauss decomposition for Chevalley groups, revisited

In the 1960's Noboru Iwahori and Hideya Matsumoto, Eiichi Abe and Kazuo Suzuki, and Michael Stein discovered that Chevalley groups $G=G(Φ,R)$ over a semilocal ring admit remarkable Gauss decomposition $G=TUU^-U$, where $T=T(Φ,R)$ is a split maximal torus, whereas $U=U(Φ,R)$ and $U^-=U^-(Φ,R)$ are unipotent radicals of two opposite Borel subgroups $B=B(Φ,R)$ and $B^-=B^-(Φ,R)$ containing $T$. It follows from the classical work of Hyman Bass and Michael Stein that for classical groups Gauss decomposition holds under weaker assumptions such as $\sr(R)=1$ or $\asr(R)=1$. Later the second author noticed that condition $\sr(R)=1$ is necessary for Gauss decomposition. Here, we show that a slight variation of Tavgen's rank reduction theorem implies that for the elementary group $E(Φ,R)$ condition $\sr(R)=1$ is also sufficient for Gauss decomposition. In other words, $E=HUU^-U$, where $H=H(Φ,R)=T\cap E$. This surprising result shows that stronger conditions on the ground ring, such as being semi-local, $\asr(R)=1$, $\sr(R,Λ)=1$, etc., were only needed to guarantee that for simply connected groups $G=E$, rather than to verify the Gauss decomposition itself.

math.GR↗

arXiv2011

Unitriangular factorisations of Chevalley groups

Lately, the following problem has attracted a lot of attention in various contexts: find the shortest factorisation $G=UU^-UU^-...U^{\pm}$ of a Chevalley group $G=G(Φ,R)$ in terms of the unipotent radical $U=U(Φ,R)$ of the standard Borel subgroup $B=B(Φ,R)$ and the unipotent radical $U^-=U^-(Φ,R)$ of the opposite Borel subgroup $B^-=B^-(Φ,R)$. So far, the record over a finite field was established in a 2010 paper by Babai, Nikolov, and Pyber, where they prove that a group of Lie type admits unitriangular factorisation $G=UU^-UU^-U$ of length 5. Their proof invokes deep analytic and combinatorial tools. In the present paper we notice that from the work of Bass and Tavgen one immediately gets a much more general result, asserting that over any ring of stable rank 1 one has unitriangular factorisation $G=UU^-UU^-$ of length 4. Moreover, we give a detailed survey of triangular factorisations, prove some related results, discuss prospects of generalisation to other classes of rings, and state several unsolved problems. Another main result of the present paper asserts that, in the assumption of the Generalised Riemann's Hypothesis, Chevalley groups over the ring $\Int\Big[\displaystyle{1\over p}\Big]$ admit unitriangular factorisation $G=UU^-UU^-UU^-$ of length 6. Otherwise, the best length estimate for Hasse domains with infinite multiplicative groups that follows from the work of Cooke and Weinberger, gives 9 factors.

math.GR↗

arXiv2007

The congruence kernel of an arithmetic lattice in a rank one algebraic group over a local field

Let k be a global field and let k_v be the completion of k with respect to v, a non-archimedean place of k. Let \mathbf{G} be a connected, simply-connected algebraic group over k, which is absolutely almost simple of k_v-rank 1. Let G=\mathbf{G}(k_v). Let Γbe an arithmetic lattice in G and let C=C(Γ) be its congruence kernel. Lubotzky has shown that C is infinite, confirming an earlier conjecture of Serre. Here we provide complete solution of the congruence subgroup problem for \Gamm$ by determining the structure of C. It is shown that C is a free profinite product, one of whose factors is \hat{F}_ω, the free profinite group on countably many generators. The most surprising conclusion from our results is that the structure of C depends only on the characteristic of k. The structure of C is already known for a number of special cases. Perhaps the most important of these is the (non-uniform) example Γ=SL_2(\mathcal{O}(S)), where \mathcal{O}(S) is the ring of S-integers in k, with S=\{v\}, which plays a central role in the theory of Drinfeld modules. The proof makes use of a decomposition theorem of Lubotzky, arising from the action of Γon the Bruhat-Tits tree associated with G.

math.GR↗

arXiv2007

Primes in a prescribed arithmetic progression dividing the sequence a^k+b^k

Given positive integers a,b,c and d such that c and d are coprime we show that the primes p=c(mod d)dividing a^k+b^k for some k>=1 have a natural density and explicitly compute this density. We demonstrate our results by considering some claims of Fermat that he made in a 1641 letter to Mersenne.

math.NT↗

arXiv2006

Quadratic Factors of $f(X)-g(Y)$

We classify the pairs of polynomials $f,g$ over a field $K$, such that $f(X)-g(Y)$ has a factor of total degree at most 2. This was done by Y. Bilu for characteristic 0 fields $K$. As his method does not work in positive characteristic, we use a quite different approach.

math.NT↗