We discuss a spacetime having the topology of $S^{3}\times\mathbb{R}$ but with a different smoothness structure. The initial state of the cosmos in our model is identified with a wildly embedded 3-sphere (or a fractal space). In previous work we showed that a wild embedding is obtained by a quantization of a usual (or tame) embedding. Then a wild embedding can be identified with a (geometrical) quantum state. During a decoherence process this wild 3-sphere is changed to a homology 3-sphere. We are able to calculate the decoherence time for this process. After the formation of the homology 3-sphere, we obtain a spacetime with an accelerated expansion enforced by a cosmological constant. The calculation of this cosmological constant gives a qualitative agreement with the current measured value.

In this paper we discuss a spacetime having the topology of S^3 x R but with a different smoothness structure leading to a geometric model for inflation, called geometric inflation. In particular this spacetime is not globally hyperbolic and we obtain a time line with a spatial topology change from the 3-sphere to a homology 3-sphere and back. The topology of the spacetime remains invariant. Among the infinite possible smoothness structures of this spacetime, we choose a homology 3-sphere constructed from the knot 8_{10} with hyperbolic geometry, i.e. admitting a homogenous metric of negative scalar curvature. We discuss the accelerated expansion for FLRW cosmology caused by the topology change. In contrast to other inflation models, this process stops after a finite time. Alternatively, the topology change can be also described by a SU(2)-valued scalar field. Then we calculate the expansion rate (having more than 60 e-folds) and the energy time scale. The coupling to matter is also interpreted geometrically and the reheating process (as well the supercooled expansion during inflation) is naturally obtained. The model depends only on a single parameter, a topological invariant of the homology 3-sphere, and assumes a Planck size universe of S^3-topology. The dependence of the model on the initial state and the a geometric interpretation of quantum fluctuations are also discussed.

In this paper we discuss the restrictions of the spacetime for the standard model of cosmology by using results of the differential topology of 3- and 4-manifolds. The smoothness of the cosmic evolution is the strongest restriction. The Poincare model (dodecaeder model), the Picard horn and the 3-torus are ruled out by the restrictions but a sum of two Poincare spheres is allowed.

In this paper we discuss a space-time having the topology of S^{3}xR but with different smoothness structure. This space-time is not a global hyperbolic space-time. Especially we obtain a time line with a topology change of the space from the 3-sphere to a homology 3-sphere and back but without a topology-change of the space-time. Among the infinite possible smoothness structures of this space-time, we choose a homology 3-sphere with hyperbolic geometry admitting a homogenous metric. Then the topology change can be described by a time-dependent curvature parameter k changing from k=+1 to k=-1 and back. The solution of the Friedman equation for dust matter (p=0) after inserting this function shows an exponential growing which is typical for inflation. In contrast to other inflation models, this process stops after a finite time.

Usually, the topology of a 4-manifolds $M$ is restricted to admit a global hyperbolic structure $\Sigma\times\mathbb{R}$. The result was obtained by using two conditions: existence of a Lorentz structure and causality (no time-like closed curves). In this paper we study the influence of the smoothness structure to show its independence of the two conditions. Then we obtain the possibility for a topology-change of the 3-manifold $\Sigma$ keeping fix its homology. We will study the example $S^{3}\times\mathbb{R}$ with an exotic differential structure more carefully to show some implications for cosmology. Especially we obtain an interpretation of the transition in topology as dark energy.

This paper has been withdrawn by the authors due to the too limited range of the current version of the model (the dark energy fraction is constant, but it has to be a function of the scale factor). We will resubmit the paper after extension.