A modular interpretation of BBGS towers (Q2220448)

The Ihara constant of a finite field \(\mathbb{F} = \mathbb{F}_{q}\) is \[ A(q) = \limsup N_{q}(g)/g, \] where \(N_{q}(g)\) is the maximal number of \(\mathbb{F}\)-rational points of a (connected, smooth, projective) algebraic curve of genus \(g\) over \(\mathbb{F}\), and \(g\) tends to infinity. Hundreds of papers have appeared in the last 40 years or so which aim to give upper and lower bounds on \(N_{q}(g)\) and \(A(q)\). An important method in the study of \(A(q)\) is the construction of towers \((X_{i})_{i \in \mathbb{N}}\) of curves \(X_{i}/\mathbb{F}\) with genus tending to infinity, and whose numbers \(N(X_{i}, \mathbb{F})\) of rational points may be estimated from below. Such a tower is \textbf{good} if \(\limsup N(X_{i}, \mathbb{F})/g(X_{i})\) is positive (in which case it provides lower estimates for \(A(q)\)), and is \textbf{recursive} if the associated sequence of function fields \(\mathbb{F}(X_{i})\) is recursively defined through a polynomial equation independent of \(i\) (see the article for details). While recursiveness has little theoretical significance, it is important for practical and computational purposes. The authors describe three different types of towers over the field \(\mathbb{F} = \mathbb{F}_{q^{m}}\), where the constituents are Drinfeld modular curves with different sorts of level structures. In their main theorem they show that these towers are all recursive, and the recursion equations are explicitly given. Thus the paper bridges the gap between the arithmetic-geometric theory of (classical elliptic, Shimura, or Drinfeld) modular curves, and the algorithmic-oriented approach of function field towers.