A lower bound for the canonical height on elliptic curves over abelian extensions. (Q1427990)

The Lehmer conjecture was originally stated for number fields \(K\). Its analogue for elliptic curves \(E\) over a number field \(K\) asserts that the canonical height \(\widehat{h}\) on \(E/K\) satisfies a lower estimate \[ \widehat{h}(P) \geq [K:{\mathbb Q}]C \] for all non-torsion points \(P \in E(K)\) with a constant \(C = (E/K)\). The author proves this analogue (without the field degree factor \([K:{\mathbb Q}]\)) for elliptic curves \(E/K\) and non-torsion points \(P \in E(K^{ab})\), where \(K^{ab}\) is the maximal abelian extension of \(K\). In this way he extends work of M. Baker who had proved this estimate for elliptic curves \(E/K\) with fractional \(j\)-invariant or with complex multiplication. The reason why this extension of the estimate of M. Baker turns out to be successful is that the author works with formal groups which has the additional advantage that computations in formal groups are rather easy. Formal groups come in if for a point \(P \in E(L)\) and for a finite extension \(L\) of \(K\) the field of definition \(K(P)\) of \(P\) ramifies over \(K\). We mention that a lower estimate for the canonical height \(\widehat{h}\) of an elliptic curve \(E\) over a number field \(K\) (especially if \(K = {\mathbb Q}\) is the field of rational numbers) plays a prevailing role in Manin's conditional algorithm for determining the rank and generators of the curve \(E\) over the field \(K\) [see \textit{Yu. I. Manin}, Cyclotomic fields and modular curves, Russ. Math. Surv. 26, 7--78 (1971); translation from Usp. Mat. Nauk 26, No. 6(162), 7--71 (1971; Zbl 0241.14014), and \textit{H. G. Zimmer}, On Manin's conditional algorithm, Bull. Soc. Math. Fr., Suppl., Mém. 49--50, 211--224 (1977; Zbl 0371.14014), and \textit{H. G. Zimmer}, Generalization of Manin's conditional algorithm, Proc. 1976 ACM Symp. Symb. Alg. Comput., 285--299 (1976; Zbl 0463.14006)]. In this connection, it should also be mentioned that occasionally it is easier to work with elliptic curves \(E\) over certain number fields \(K\) having integral \(j\)-invariant, for instance if one wishes to determine torsion points [see e.g. \textit{S. Schmitt} and \textit{H. G. Zimmer}, Elliptic curves - a computational approach. Berln: W. de Gruyter (2003; Zbl 1195.11078)].