Around Pelikán's conjecture on very odd sequences

From MaRDI portal
(Redirected from Publication:2487840)



Abstract: Very odd sequences were introduced in 1973 by J. Pelikan who conjectured that there were none of length >=5. This conjecture was disproved by MacWilliams and Odlyzko in 1977 who proved there are in fact many very odd sequences. We give connections of these sequences with duadic codes, cyclic difference sets, levels (Stufen) of cyclotomic fields, and derive some new asymptotic results on their lengths and on S(n), which denotes the number of very odd sequences of length n.


A sequence \(a_1,a_2,\dots,a_n\) of 0's and 1's is called very odd, if all the numbers \(\sum_{1\leq i\leq n-k}a_ia_{i+k}\), \(0\leq k<n\) are odd. An example of a very odd sequence of length 4 is 1011. The authors discuss the connection between \(S(n)\), the number of very odd sequences of length \(n\), and coding theory, resp. the level (Stufe) of cyclotomic fields. They prove a formula for \(S(n)\) and its value distribution. Some special results are: \(S(n)>0\) iff the order of 2 modulo \(2n-1\) is odd; this occurs for infinitely many \(n\); \(S(n)>0\) implies \(S(n)=2^e\) for some integer \(e\); assume GRH, \(e\) odd or \(e\equiv 0\pmod 4\), then \(S(n)=2^e\) for infinitely many \(n\).



Cites work









This page was built for publication: Around Pelikán's conjecture on very odd sequences

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2487840)