In this paper, we propose FoVCast, a 360° virtual reality (VR) video transmission framework, to implement efficient and robust performance for users with heterogeneous channels. In FoVCast, the overlapped field of views (FoVs) requested by multiple users are transmitted in the broadcast mode to improve the utilization efficiency of wireless bandwidth. To combat heterogeneous wireless broadcast environments, robust video transmission is exploited to provide users with received quality proportional to their channel conditions. The basic idea is to replace nonlinear quantization and coding with linear transform and power scaling to achieve graceful performance over heterogeneous channels. Furthermore, a joint power allocation and subcarrier matching problem is formulated to minimize the distortion of overlapped FoVs. To solve the problem efficiently, an auction-based collaborative optimization algorithm is proposed. Simulation results demonstrate that the proposed scheme can achieve significant performance gains in terms of peak signal-to-noise ratio (PSNR) compared with baseline schemes.