Research notes on every problem, and a library of the papers behind them. Built from the open erdos repository.
Updated
Second order construction
Statement and scope
There is an unbounded sequence of integers for which
Equivalently,
The construction includes diagonal pair sums and has only one possible exceptional integer sum. This does not improve the leading constant or settle the proposed upper bound.
The idea is to compress the reflected construction by an amount of order , while deleting only order points to eliminate the new collisions. Averaging over the compression parameter reduces the required relation count to an ordinary Cartesian-product count; no sparse four-variable counting lemma is assumed.
A cyclic Sidon set and its translate variance
For a prime power , put . There is a set , of size , in which every nonzero ordered difference occurs exactly once.
Here is a self-contained verification of the standard Singer construction used below. Let , and let be a two-dimensional -subspace of . The projective points of give elements of the cyclic group , which has order . For , the subspaces and intersect in exactly one dimension. They cannot coincide: otherwise for , the three elements would belong to , contradicting the degree-three minimal polynomial of . Their unique common projective point gives exactly one representation of the quotient class of as a quotient of two points of . Under a cyclic identification, this is the required difference property. It implies ordinary modular Sidonicity, including diagonal pairs.
For comparison, see §3.4 and entry [4] of source record. The proof above supplies all properties used here.
For any set of residues, and a uniformly random translate , let
The perfect-difference property gives
Since , it follows that
Simultaneous distribution in two short endpoint intervals
Fix , and put . For every sufficiently large , one can choose a translate of , represented by , so that both measures
converge weakly to Lebesgue measure on , as tends to infinity through prime powers.
For completeness, apply (2) to the endpoint prefixes of relative length , where . Their normalized counts have expectation , uniformly in , and variance . Chebyshev's inequality, with normalized tolerance , bounds each failure probability by . There are prefixes in both windows, so the probability of any failure is . A good translate exists. Monotonicity fills the gaps between grid points and gives simultaneous uniform convergence of the two distribution functions. In particular, both endpoint intervals contain elements of .
No independence between the two endpoint measures is asserted or needed. For this deterministic good translate, their Cartesian products converge to the corresponding products of Lebesgue measure.
Averaging the new forbidden relations
Fix also . For each integer , let count pairs , with an unordered multiset of three inputs and , satisfying
Every input belongs to , and every output to . On writing
equation (4) is equivalent to . Thus the sum over these of the ordered relation counts equals the number of endpoint quadruples in the simplex shell
with the lower boundary adjusted by for rounding. The limiting product measure assigns the boundary measure zero. By (3), this count is
Indeed the simplex of nonnegative quadruples with sum at most has volume .
Quadruples with a repeated input contribute only
choose which two input positions coincide, their common value, the remaining input, and the output. For all other tuples there are exactly six input orderings. Consequently
There are parameters. With , some therefore satisfies
The repeated-input error becomes after averaging. Here and below all errors tend to zero with for fixed .
Deletion, reflection, and the unique exception
Delete from every output participating in (4), and call the remaining set . At most elements are deleted. Put
For large , . The set is ordinary Sidon because is, and (4) is precisely the condition that an element of be a sum of three elements of . Hence
The positive-half characterization now shows that has unique unordered pair sums away from zero. Explicitly, positive sums are either sums of two elements of or positive differences of two elements of . Each type is unique by Sidonicity; a collision between the types would give the excluded three-to-one relation. Negative sums follow by reflection. This checks diagonal pairs as well.
A translation gives the integer set
This avoids any parity restriction on . The two copies are disjoint for large , and
Their exceptional sum is ; it is the only integer sum that can repeat.
Optimizing the gain
Uniformly for ,
Using and (6) in (7) yields
For fixed , this expression is maximized by
and its coefficient becomes
Choose a sequence , take the corresponding , and then choose increasing prime powers large enough that the normalized error in (8) is at most . Since , this proves (1).
The diagonal choice proves an unbounded sequence statement, not the same second-order inequality for every sufficiently large . The established leading bounds remain those on the bounds page.