Integrated Sensing and Communications Systems (ISAC) sense radio frequency spectrum and transfer wireless data jointly. They operate in a shared and congested spectrum. We are considering ISAC systems that cooperate or are co-designed for mutual benefits. Co-designed systems may share HW and antenna resources as well as awareness about the state of the radio spectrum. The ISAC systems have a number of operational parameters that can be adjusted either by using structured optimization or machine learning.  We focus on multicarrier waveforms used by most current and emerging wireless communication systems. Similarly, multicarrier waveforms have been employed for radar purposes. We will present waveform optimization, reinforcement learning, interference management and signal processing methods for co-designed ISAC systems that share channel and interference awareness. Model-based reinforcement learning approach is taken to exploit the rich structural knowledge of man-made communication and sensing systems and propagation effects to choose optimal actions and resource use. .Optimizing operational parameters can be modeled as a radar-centric or communications-centric constrained optimization problem where the minimum desired performance levels for other sub-systems impose the constraints. The developed OFDM radar signal processing algorithms in ISAC can take advantage of nonidealities such as carrier offsets that are commonly considered an impairment in wireless communications. We demonstrate the achieved performance gains in different sensing and communication tasks and interference management through extensive simulation and analytical results.