A Density Condition on Point Sets with Slowly-Scaling Distinct Dot Products
Gandhi
The distinct dot products problem, a variation on the ErdÅs distinct distance problem, asks "Given a set $P_n$ of $n$ points in $\mathbb{R}^2$, what is the minimum number $|D(P_n)|$ of distinct dot products formed between them, asymptotically?" The best proven lower-bound is $|D(P_n)| \gtrsim n^{2/3+7/1425}$, due to work by Hanson$\unicode{x2013}$Roche-Newton$\unicode{x2013}$Senger, and a recent improvement by Kokkinos. However, the slowest-scaling known constructions have $|D(P_n)|\sim n$, leaving quite a large gap in the bound. Finding a sublinearly-scaling construction, or disproving its existence, would narrow this gap. We provide a condition that a sequence of point configurations $(P_n)_{n \in \mathbb{N}}$ must satisfy in order for $|D(P_n)|$ to scale 'slowly' i.e. $|D(P_n)| \ll n^{3/4}$. Namely, we prove that any such configuration must contain a point-rich line that gets arbitrarily 'dense' as the sequence progresses.
academic
A Density Condition on Point Sets with Slowly-Scaling Distinct Dot Products
This paper investigates the distinct dot products problem, a variant of the classical Erdős distinct distances problem. The problem asks: given a set Pn of n points in R2, what is the asymptotic behavior of the minimum number of distinct dot products ∣D(Pn)∣ formed between them? The current best lower bound is ∣D(Pn)∣≳n2/3+7/1425, while the slowest-growing known construction achieves ∣D(Pn)∣∼n, leaving a significant gap. This paper provides conditions that point configuration sequences (Pn)n∈N must satisfy for ∣D(Pn)∣ to grow "slowly," i.e., ∣D(Pn)∣≪n3/4. Specifically, it is proven that any such configuration must contain a point-rich line that becomes arbitrarily "dense" as the sequence progresses.
This paper studies the distinct dot products problem, a variant of the famous Erdős distinct distances problem. Given n points in the plane, the problem determines the minimum number of distinct dot products that can be formed between them. This is a fundamental problem in combinatorial geometry with significant theoretical importance.
Theoretical Importance: The problem is classical in combinatorial geometry, connecting to multiple mathematical branches including additive combinatorics and harmonic analysis
Technical Challenge: There exists a significant gap between upper and lower bounds, with the current best lower bound approximately n2/3, while known constructions only achieve linear growth n
Methodological Value: Techniques developed for this problem may apply to other related combinatorial problems
Lower Bound Techniques: Work by Hanson-Roche-Newton-Senger and Kokkinos provides bounds of n2/3+c, but still falls short of the linear upper bound
Construction Methods: Known slowest-growing constructions (such as geometrically spaced collinear points or equally-spaced points on circles) all achieve linear growth ∼n
Theoretical Gap: Lack of deep understanding regarding the possibility of sublinear growth
This paper aims to fill the theoretical gap by identifying structural conditions that slowly-growing point configurations must satisfy, providing new insights toward ultimately resolving the upper-lower bound gap.
Given a point configuration sequence (Pn)n∈N, where each Pn is a set of n distinct points in R2, define the dot product set D(Pn):={pi⋅pj∣pi,pj∈Pn}. The goal is to characterize necessary conditions on configurations satisfying ∣D(Pn)∣≪n3/4.
Lemma 3.6 (Existence of Popular Lines): For configuration sequences with ≪nα dot products, there must exist a "popular line" containing ≫n2−2α points.
Lemma 4.6 (Existence of Popular Circles): For configuration sequences with ≪nα dot products, there must exist a "popular circle" containing ≫n1−α points.
Partition the real axis into "buckets" Bi, each corresponding to intervals between consecutive terms in a geometric series. Analyze the projections of complex dot products across buckets to count distinct dot products.
Definition 6.2 (b-dense): A set L of ℓ collinear points is called b-dense if there exist ∼ℓ pairs of adjacent points p,q∈L such that ∣p∣/∣q∣ falls in the interval (b,1).
By proving that if all point-rich lines satisfy good spacing conditions, then ∣D(Pn)∣≳n3/4, derive the density conditions for slowly-growing configurations.
Theorem 6.3 (Density Condition for Slow Growth):
Let (Pn)n∈N be a sequence of point configurations, where each Pn is a set of n distinct points in R2 with ∣D(Pn)∣≪n3/4. Then for all b∈(0,1), there exists a subsequence such that each configuration in the subsequence contains a b-dense point set L arranged along a line through the origin with ∣L∣≳n1/2.
Lemma 3.1: n collinear points in geometric progression produce ∼n distinct dot products.
Lemma 3.2: Any n collinear points produce ≳n distinct dot products.
Proposition 5.1: A configuration containing N(n) equally-spaced points on a circle and M(n) geometrically-spaced collinear points produces ≳N(n)M(n) dot products.
This paper proves that any slowly-growing point configuration must contain a dense line structure approximating arithmetic progressions. This provides important insight into understanding the essence of the dot products problem.
This paper provides a new theoretical framework for the distinct dot products problem, potentially inspiring subsequent research and advancing the field. While not completely resolving the upper-lower bound gap, it makes important contributions to understanding the problem's essence.
The paper cites major works in the field, including foundational results by Hanson-Roche-Newton-Senger and recent related progress, demonstrating comprehensive command of the literature.