In an accretion disk, the boundary layer is the region where the disk attaches to the star. Unlike in the disk proper, the boundary layer has a rotation profile which rises with radius, since the rotation profile must transition from that of the more slowly rotating star to that of the more quickly rotating disk. The rising rotation profile makes the boundary layer (linearly) stable to the MRI instability. Therefore, a different mechanism must be responsible for transporting angular momentum in the boundary layer and decelerating the disk material. The details of this mechanism are observationally significant since up to half of the total disk luminosity comes from the boundary layer region. Working with Roman Rafikov and James Stone, I have discovered a new instability which is capable of transporting angular momentum and driving accretion in the boundary layer via waves rather than turbulence. This instability is related to the Papaloizou-Pringle class of instabilities, in that it is sourced by a corotation resonance in the boundary layer. However, it is distinct from the traditional Papaloizou-Pringle instability; it is also robust, feeding off the large degree of shear in the boundary layer. I will present both analytical and numerical results, which explain the mechanism of the new instability (which we call the sonic or acoustic instability) and its role in angular momentum transport in accretion disk boundary layers. I will also describe why no reasonable alpha prescription for the viscosity can be reconciled with this instability. This has significant ramifications for semi-analytical models of boundary layers, which typically assume some form of alpha viscosity.