Time-sensitive communications (TSC) in wireless networks is an emerging paradigm that gains research momentum as an enabler of the industrial Internet of Things. As compared to ultra-reliable low-latency communications (URLLC) in fifth-generation cellular networks, TSC has stricter requirements in terms of latency and reliability, and also demands absolute time-synchronization and on-time delivery of packets for deterministic and isochronous real-time applications. In this regard, a key question is how to schedule TSC traffic flows effectively. This paper presents a radio resource allocation strategy for deterministic downlink TSC flows leveraging traffic pattern knowledge. Taking physical layer control channel effects into account, a comparison of semi-persistent and dynamic packet scheduling methods is presented as well as necessary enhancements to link adaptation and interference coordination procedures. With the proposed methods, the network capacity in terms of number of supported TSC flows can be more than doubled compared to traditional dynamic scheduling methods as commonly assumed for URLLC applications.