A gap result for Cameron-Liebler \(k\)-classes (Q2400562): Difference between revisions

A set of \(k\)-subspaces, \(k \geq 1\), of \(\mathrm{PG}(2k+1,q)\) with the property that every regular \(k\)-spread contains the same number \(\chi\) of \(k\)-subspaces is a Cameron-Liebler \(k\) class of \(\mathrm{PG}(2k+1,q)\). Using a technique originally used to extend nets to planes, the author improves results by \textit{M. Rodgers}, \textit{L. Storme} and \textit{A. Vansweevelt} [``Cameron-Liebler \(k\)-classes in \(\mathrm{PG}(2k+1, q)\)'', Combinatorica (to appear), \url{doi:10.1007/s00493-016-3482-y}] to show the following results. (1) There does not exist a Cameron-Liebler 2-class in \(\mathrm{PG}(5,q)\) with parameter \(\chi\) satisfying \(2< \chi< {{q}\over{3}}\). (2) There exists a positive integer \(q_0\) with the following properties: If \(k \geq 3\) is an integer and \(q\) is a prime power with \(q \geq q_0\) and \(k<q log(q) -q -1\), then \(\mathrm{PG}(2k+1,q)\) has no Cameron-Liebler \(k\)-class with parameter \(\chi\) for \(2 <\chi< {{q}\over{5}}\).