Theories of bundles with additional structures (Q843623)

The author introduces a concept of (vector) \(n\)-bundles with extra homotopy conditions and studies their classifying spaces. Moreover, he studies similar structures on Matrix bundles and shows that their classifying spaces are in fact matrix Grassmannians. In the last section he then derives obstructions for embedding matrix bundles into trivial bundles. The author defines a 0-bundle over \(X\) to be a \(k\)-dimensional vector bundle \(\xi_{k}\) over \(X\) and a 1-bundle over \(X\) to be a collection of two vector bundles \(\xi_{k}\) and \(\xi_{l}\) over \(X\), together with a bundle \(Z\) over \(X\times[0,1]\) such that \(Z|_{X\times\{0\}}=\xi_{k}\otimes [l]\) and \(Z|_{X\times\{1\}}=[k]\otimes \xi_{l}\) (where \([k]\) and \([l]\) denote trivial bundles of the corresponding dimensions). \(Z\) is here thought of as a morphism from \(\xi_{k}\) to \(\xi_{l}\), and two 1-bundles are defined to be equivalent if their constituents are so. Similarly, a 2-bundle is a collection of 3 vector bundles, together with morphisms (in the above sense) and morphisms between morphisms (again in the above sense). The author then shows that a classifying space for 1-bundles (with fixed \(k\) and \(l\)) is given by a double coset space of \(\text{SU}(kl)\). It is this double coset construction, which forces the author to assume \(k\) and \(l\) to be coprime. He then proceeds to a similar description of the classifying space of 2-bundles if the dimensions of the 3 aforementioned bundles are pair-wise coprime. Although the explicit derivation of classifying spaces for the used concept of higher bundles invites to a comparison with other, more commonly used concepts [cf. \textit{J. C. Baez} and \textit{D. Stevenson}, The classifying space of a topological 2-group. Algebraic topology. The Abel symposium 2007. Proceedings of the fourth Abel symposium, Oslo, Norway, August 5--10, 2007. Berlin: Springer. Abel Symposia 4, 1-31 (2009; Zbl 1182.55015)]), the author refrains from doing so.