scientific article; zbMATH DE number 2046136

From MaRDI portal
Publication:4451098





Classification of quantum (noncommutative) differential structures of quantum groups and quantum Riemannian manifolds based on quantum group principal bundles, mainly due to the author, are reviewed. Let \(M\) be the `coordinate' of a noncommutative space. Then a \(M\)-bimodule \(\Omega^1\) with a derivation \(d: M\to\Omega^1\) such that \(M\times M\to\Omega^1\), \(a\otimes b\mapsto \text{ad\,}b\) is surjective, is called a cotangent space or a space of differential 1-forms of \(M\). When \(M\) has a Hopf algebra structure with coproduct \(\Delta\) and counit \(\varepsilon\), then \(\Omega^1\) is assumed to be bicovariant, that is \(\Omega^1\) is a bicomodule with coactions \(\Delta_L\) and \(\Delta_R\), and \(d\) is a bicomodule map. A differential calculus is said to be noncommutative, or quantum, if the left and right multiplication of forms by functions do not coincide. A calculus is said to be coirreducible if it has no proper quotient (sec. 2). Then classifications of differential calculi for \(M = k[x]\), \(\mathbb{C}[G]\) and \(\mathbb{C} G\), the group algebra of \(G\), are stated [\textit{S. Majid}, J. Geom. Phys. 25, 119--140 (1998; Zbl 0952.17012), hereafter referred to as [1]]. General results, such as existence of 1-1 correspondence between the bicovariant differentials of a quantum group \(M\) and its twist \(M^\chi\) (th. 1), Tannaka-Krein reconstruction arguments deducted from this 1-1 correspondence [\textit{S. Majid} and \textit{R. Oeckl}, Commun. Math. Phys. 205, 617--655 (1999; Zbl 0939.58007)] are also stated. After exposing the braided-Lie algebra structure of the quantum tangent space of the factorisable quantum group \(M = \mathbb{C}_q[G]\) (prop. 4, cf.[1]), sec. 2 is concluded to remark that quantum cohomology of the discrete manifold, obtained from a covering of a manifold, recovers the Čech cohomology [\textit{T. Brzezinski} and \textit{S. Majid}, Acta Appl. Math. 54, 185--232 (1998; Zbl 0946.58005)]. The rest of the paper deals with quantum bundles (sec. 3) [cf. \textit{T. Brzezinski} and \textit{S. Majid}, Commun. Math. Phys. 157, 591--638 (1993; Zbl 0817.58003) and ibid. 167, 235 (1995; Zbl 0823.58006), hereafter referred to as [2]] and quantum Riemannian manifold (sec. 4) [cf. \textit{S. Majid}, J. Geom. Phys. 30, 113--146 (1999; Zbl 0940.58004), hereafter refered to as [3]]. Since a local argument is difficult to apply in noncommutative spaces, a global picture is used to develop the theory of quantum bundles and connections, and a Hopf algebra \(H\) is taken in the role of functions on the structure group of bundels. Precise definitions together with definitions of quantum connections are given in sec. 2. In other context of Hopf algebra theory, quantum bundles can be formulated as a Hopf-Galois extension, but the differential geometric picture is more powerful. After descibing the \(q\)-monopole due to the author (prop. 5) [cf. \textit{P. M. Hajac} and \textit{S. Majid}, Commun. Math. Phys. 206, 247--264 (1999; Zbl 0936.46052)], he says if we can construct a theory of quantum flat bundles, then it might provide a homotopy quantum group \(\pi_1(M)\) for any algebra \(M\) (sec. 3). A frame resolution of \(M\) is a quantum principal bundle \((P, H, \Delta_R)\) over \(M\) and a comodule \(V\) and an equivalent soldering form \(\theta: V\to P\Omega^1M\subset \Omega^1 M\) such that the induced map \(E^*\to \Omega^1 M\), \(p\otimes v\mapsto p\theta(v)\) is an isomorphism ([3]). Every quantum group \(M\) has a framing by \(H= M\), \(P= M\otimes M\), \(V= \ker\varepsilon\) (th. 6, cf. [3]). The author says to apply noncommutative geometry to physics, a slight generalization of conventional Riemannian geometry is needed: (i): We should allow any group \(G\) in the `frame bundle'. (ii): The generalized metric \(g_{\mu\nu}= \theta^{*a}_\mu\theta_{\nu a}\) is nondegenerate but need not to be symmetric. (iii): The generalized Levi-Cività connection defined as having vanishing torsion and vanishing cotorsion respects the metric only in a skew sense: \(\nabla_\mu g_{\nu\rho}- \nabla_\nu g_{\mu\rho}= 0\), and need not to be uniquely determined ([2]). This generalization of Riemannian geometry includes special cases of symplectic geometry and metrics where the generalized metric is totally antisymmetric. Metrics with antisymmetric part are what is needed in string theory to establish \(T\)-duality (sec.4). For the entire collection see [Zbl 0980.00028].











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4451098)