On monomial graphs of girth eight (Q2467318)

In this paper the authors introduce monomial graphs over a finite field \(\mathbb{F}\) with odd characteristic as bipartite graphs with vertex partitions \(P=\mathbb{F}^3\) and \(L=\mathbb{F}^3\) and a vertex \((a,b,c)\in P\) is adjacent with a vertex \((x,y,z)\in L\) if \(b+y=f_2(x,a)\) and \(c+z=f_3(x,a)\), with \(f_2\) and \(f_3\) monomials in two variables over \(\mathbb{F}\). Motivated by the unique known class of examples \(f_2(x,a)=ax\) and \(f_3(x,a)=ax^2\) that have girth at least 8, and which give rise to the symplectic quadrangles (they are isomorphic to the geometries far away from flags in such quadrangles), they ask themselves whether other monomial graphs possibly exist. They prove that it is very unlikely by establishing some rather restrictive necessary conditions on the monomial \(f_2\) and \(f_3\). By the way, they apparently fail to note that, whenever \(\mathbb{F}\) contains at least 3 elements, then the graph automatically has girth at most 8, and so they are really looking at graphs with girth exactly 8. The motivation of the paper is largely to say more about generalized quadrangles of odd order. But in this case, this approach is slightly less general than the coordinatization introduced by the reviewer and \textit{G. Hanssens} in [Ann. Discrete Math. 37, 195--207 (1988; Zbl 0643.51011)] since the former would automatically produce generalized quadrangles with a rather large automorphism group (at least size \((| \mathbb{F}| -1)^2| \mathbb{F}| ^2\)) containing elations and generalized homologies. That path has been walked before.