Nanonetworking is an emerging field of research, where nanotechnology and communication engineering are applied on a common ground. Molecular Communication (MC) is a bio-inspired paradigm, where Nanonetworks, i.e., the interconnection of devices at the nanoscale, are based on the exchange of molecules. Amongst others, diffusion-based MC is expected to be suitable for covering short distances (nm-m). In this work, we explore the main characteristics of diffusion-based MC through the use of N3Sim, a physical simulation framework for MC. N3Sim allows for the simulation of the physics underlying the diffusion of molecules for different scenarios. Through the N3Sim results, the Linear Time Invariant (LTI) property is proven to be a valid assumption for the free diffusion-based MC scenario. Moreover, diffusion-based noise is observed and evaluated with reference to already proposed stochastic models. The optimal pulse shape for diffusion-based MC is provided as a result of simulations. Two different pulse-based coding techniques are also compared through N3Sim in terms of available bandwidth and energy consumption for communication.