Explicit bounds for small prime nonresidues
Shilin Ma, Kevin J. McGown, Devon Rhodes, Mathias Wanner

TL;DR
This paper provides explicit upper bounds for the smallest prime nonresidues of Dirichlet characters modulo a prime, improving understanding of their distribution with precise bounds depending on the prime and the number of nonresidues.
Contribution
It establishes explicit bounds for the smallest prime nonresidues of Dirichlet characters, with constants depending on given parameters, advancing previous theoretical results.
Findings
Bounds depend on prime p and number of nonresidues n
Explicit constants are provided for practical bounds
Results hold for sufficiently large p and fixed n
Abstract
Let be a Dirichlet character modulo a prime~. We give explicit upper bounds on , the smallest prime nonresidues of . More precisely, given and there exists an absolute constant such that whenever and .
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsAnalytic Number Theory Research · Finite Group Theory Research · Limits and Structures in Graph Theory
Explicit bounds for small prime nonresidues
Shilin Ma
Kevin McGown
Devon Rhodes
Mathias Wanner
Abstract
Let be a Dirichlet character modulo a prime . We give explicit upper bounds on , the smallest prime nonresidues of . More precisely, given and there exists an absolute constant such that whenever and .
1 Introduction
Let be a nonprincipal Dirichlet character modulo a prime . If , then we refer to as a nonresidue of . Let denote the smallest prime nonresidues of . Giving an upper bound on is an important classical problem that has received much attention. Indeed, in the case of the Legendre symbol, is the least quadratic nonresidue mod . In 1963, Burgess showed that for each , one has (see [3, 4]), and this result has stood as the state of the art since this time, save a recent improvement to the “” in the quadratic case (see [1]). In 2015, Pollack proved the following result (see [9]): For each , there are numbers and such that for all and each nonprincipal character modulo , there are more than prime nonresidues of not exceeding . In particular, for all and all , one has , although this hides the dependence on and . As we alluded to a moment ago, Banks and Guo have recently shown that in the case where is the Legendre symbol (see [1]), provided .
Often in applications (see, for example, [7, 5, 10, 2]) one requires estimates that are completely explicit, and one is willing to accept a weaker asymptotic in order to obtain constants of a reasonable magnitude. Our goal here is to give an explicit upper bound on , the th smallest prime nonresidue. Naturally, our upper bounds are asymptotically weaker than those given in [9] and [1]. The following is our main result from which one can easily derive bounds of the desired form.
Theorem 1**.**
Let be a nonprincipal Dirichlet character modulo . Let be squarefree and assume all its prime factors are less than . Set where is the number of distinct prime factors of . Suppose that whenever and . Then
[TABLE]
where
[TABLE]
using
[TABLE]
provided and .
Corollary 2**.**
Fix two integer constants and such that and . Then there exists an explicit constant such that
[TABLE]
for all and .
To our knowledge, the previous corollary constitutes the first explicit upper bound on when . When , there is the work of Norton (see [8]) that was later superceded by Treviño (see [13]) and when there is a paper by McGown (see [6]). The proof of our result involves a modification of McGown’s work (see [6]), which is based on the method of Burgess (see [3, 4]), and the adoption of Treviño’s results (see [12]).
2 Preparations
Lemma 3**.**
Let be a nonprincipal Dirichlet character to a prime modulus . Let be positive integers satisfying and . Then
[TABLE]
Proof.
From Theorem 1.1 of [13] we have
[TABLE]
Using the explicit version of Stirling’s formula given in [11], we have
[TABLE]
∎
Definition 1**.**
Let with . For constants , , define the following intervals:
[TABLE]
Lemma 4**.**
Let be a real number such that . Then the intervals , where and are disjoint subintervals of , except for .
Proof.
This is Lemma 2 of [6]. ∎
Lemma 5**.**
Let where with each prime and with each prime . Suppose is Dirichlet character modulo such that for all with . If with and , then
[TABLE]
Proof.
Suppose with and . Then for all , . Therefore . Since and and , this will equal if . This can only fail once for each divisor of .
To see this, suppose and for two different values . Since we also know . We have and thus . Now we have , which is a contradiction.
Application of the triangle inequality now gives the result; indeed, we have
[TABLE]
The proof when is similar. In this case we have which implies if . The result follows as before. ∎
Lemma 6**.**
Let be a real number. Then
[TABLE]
where
[TABLE]
Proof.
This is Lemma 3.2 of [13]. ∎
Proposition 7**.**
Let with with each prime and with each . Suppose is a Dirichlet character modulo a prime such that for all satisfying . Assume and set . Then
[TABLE]
provided .
Proof.
First, observe that by Lemma 4 we have
[TABLE]
Applying Lemma 5, and noting that has at least elements, we obtain
[TABLE]
Replacing with the above is equal to
[TABLE]
Now we may apply Lemma 6 to conclude
[TABLE]
∎
Lemma 8**.**
Let and be positive integers with . Then
[TABLE]
Proof.
By the convexity of the logarithm, we know for all . It follows that, for ,
[TABLE]
which implies
[TABLE]
and therefore
[TABLE]
∎
3 Proof of the main result
Proof of Theorem 1.
The conditions and guarantee that all the denominators in the expression for are positive. Notice also that the condition implies that, in particular, .
Let and with and . One verifies that is also satisfied. Indeed, and therefore the condition suffices.
Write . Let be composed of the distinct prime factors of less than and let be composed of the prime factors greater than or equal to as in Proposition 7, so . Note that our choices of , will allow us to apply Lemma 8, and moreover, we have from which it follows that
[TABLE]
both of these facts will be employed forthwith.
We may assume that or there would be nothing to prove. Using this and , we get as a lower bound for ; indeed,
[TABLE]
Using the upper and lower bounds for given in Lemma 3 and Proposition 7 respectively, we find
[TABLE]
which implies
[TABLE]
Substituting our values for and and using the fact , together with (1), gives us
[TABLE]
Since is composed of prime factors strictly less than , we have
[TABLE]
Using this, and substituting our values of and , we find
[TABLE]
Since , we can show
[TABLE]
This is less than 1 whenever . Since we can use this to drop some terms from the product, which yields
[TABLE]
Isolating , and noting that , gives
[TABLE]
Taking the square root of both sides and rearranging gives the desired result. ∎
Proof of Corollary 2.
This follows immediately from Theorem 1, letting be the product of the first prime nonresidues. The fact that this holds for all and can be verified by showing that is decreasing with and increasing with under the conditions given. This is not hard to verify. Indeed, calculus can be used to show that for , the expression
[TABLE]
is increasing with , and that increases with and decreases with . Similarly, the term
[TABLE]
is increasing with and decreasing with . ∎
Acknowledgements
This research was completed as part of the Research Experience for Undergraduates and Teachers program at California State University, Chico funded by the National Science Foundation (DMS-1559788). We would also like to thank the anonymous referee for their helpful suggestions which improved the quality of this paper.
The reference list from the paper itself. Each links out to its DOI / PubMed record.
- 1[1] Banks, William D.; Guo, Victor Z. Quadratic nonresidues below the Burgess bound. Int. J. Number Theory 13 (2017), no. 3, 751–759.
- 2[2] Booker, Andrew R. On Mullin’s second sequence of primes. Integers 12 (2012), no. 6, 1167–1177.
- 3[3] Burgess, D. A. On character sums and primitive roots. Proc. London Math. Soc. (3) 12 (1962), 179–192.
- 4[4] Burgess, D. A. A note on the distribution of residues and non-residues. J. London Math. Soc. 38 (1963), 253–256.
- 5[5] Lezowski, Pierre; Mc Gown, Kevin J. The Euclidean algorithm in quintic and septic cyclic fields. Math. Comp. 86 (2017), no. 307, 2535–2549.
- 6[6] Mc Gown, Kevin J. On the second smallest prime non-residue. J. Number Theory 133 (2013), no. 4, 1289–1299.
- 7[7] Mc Gown, Kevin J. Norm-Euclidean cyclic fields of prime degree. Int. J. Number Theory 8 (2012), no. 1, 227–254.
- 8[8] Norton, Karl K. Numbers with small prime factors, and the least kth power non-residue. Memoirs of the American Mathematical Society, No. 106 American Mathematical Society, Providence, R.I. 1971 ii+106 pp.
