The L_4 norm of Littlewood polynomials derived from the Jacobi symbol

Littlewood raised the question of how slowly the L_4 norm ||f||_4 of a Littlewood polynomial f (having all coefficients in {-1,+1}) of degree n-1 can grow with n. We consider such polynomials for odd square-free n, where \phi(n) coefficients are determined by the Jacobi symbol, but the remaining coefficients can be freely chosen. When n is prime, these polynomials have the smallest known asymptotic value of the normalised L_4 norm ||f||_4/||f||_2 among all Littlewood polynomials, namely (7/6)^{1/4}. When n is not prime, our results show that the normalised L_4 norm varies considerably according to the free choices of the coefficients and can even grow without bound. However, by suitably choosing these coefficients, the limit of the normalised L_4 norm can be made as small as the best known value (7/6)^{1/4}.

[1]  K. Williams,et al.  Gauss and Jacobi sums , 2021, Mathematical Surveys and Monographs.

[2]  D. J. Newman.  Norms of Polynomials , 1960 .

[3]  D. Newman,et al.  The L 4 norm of a polynomial with coefficients , 1990 .

[4]  Marcel J. E. Golay.  The merit factor of Legendre sequences , 1983, IEEE Trans. Inf. Theory.

[5]  Marcel J. E. Golay,et al.  The merit factor of long low autocorrelation binary sequences , 1982, IEEE Trans. Inf. Theory.

[6]  Jonathan Jedwab,et al.  A Survey of the Merit Factor Problem for Binary Sequences , 2004, SETA.

[7]  Andrew Granville,et al.  Zeros of Fekete polynomials , 1999 .

[8]  R. Lockhart,et al.  The expected _{} norm of random polynomials , 2000 .

[9]  W. Marsden.  I and J , 2012 .

[10]  Henry H. Hill.  Lexington , 1936 .

[11]  J. E. Littlewood,et al.  On Polynomials ∑ ±nzm,∑ eαminzm,z=e0i , 1966 .

[12]  R. Lockhart,et al.  THE EXPECTED Lp NORM OF RANDOM POLYNOMIALS , 2001 .

[13]  Paul Erdös,et al.  An inequality for the maximum of trigonometric polynomials , 1962 .

[14]  J. Littlewood.  Some problems in real and complex analysis , 1968 .

[15]  Tom Høholdt,et al.  The merit factor of binary sequences related to difference sets , 1991, IEEE Trans. Inf. Theory.

[16]  Peter B. Borwein,et al.  Binary sequences with merit factor greater than 6.34 , 2004, IEEE Transactions on Information Theory.

[17]  H. Montgomery.  An exponential polynomial formed with the Legendre symbol , 1980 .

[18]  Tom Høholdt,et al.  Determination of the merit factor of Legendre sequences , 1988, IEEE Trans. Inf. Theory.

[19]  Peter Borwein,et al.  Computational Excursions in Analysis and Number Theory , 2002 .

[20]  Kenneth G. Paterson,et al.  On the existence and construction of good codes with low peak-to-average power ratios , 2000, IEEE Trans. Inf. Theory.

[21]  P. Borwein,et al.  Merit Factors of Polynomials Formed by Jacobi Symbols , 2001, Canadian Journal of Mathematics.

[22]  Noga Alon,et al.  The Probabilistic Method , 2015, Fundamentals of Ramsey Theory.

[23]  Matthew G. Parker,et al.  Two binary sequence families with large merit factor , 2009, Adv. Math. Commun..

[24]  Jonathan I. Hall,et al.  Modifications of Modified Jacobi Sequences , 2011, IEEE Transactions on Information Theory.

[25]  E. T..  An Introduction to the Theory of Numbers , 1946, Nature.

[26]  J. Beck.  Flat Polynomials on the unit Circle—Note on a Problem of Littlewood , 1991 .

[27]  J. Bernasconi.  Low autocorrelation binary sequences : statistical mechanics and configuration space analysis , 1987 .

[28]  Kwok-Kwong Stephen Choi,et al.  Explicit merit factor formulae for Fekete and Turyn polynomials , 2001 .

[29]  J. Spencer.  Six standard deviations suffice , 1985 .

[30]  T. Apostol.  Introduction to analytic number theory , 1976 .

[31]  P. Borwein,et al.  An extremal property of Fekete polynomials , 2000 .

[32]  N. Alon,et al.  The Probabilistic Method: Alon/Probabilistic , 2008 .

[33]  Jonathan I. Hall,et al.  Construction of Even Length Binary Sequences With Asymptotic Merit Factor $6$ , 2008, IEEE Transactions on Information Theory.