Communication networks have vital roles in our daily lives, both in small-scale and large-scale perspective. First, in small-scale networks, nanonetworks consisting of nano-scale devices have been proposed to solve problems in environmental monitoring, healthcare (in the form of nanomedicine), industrial, and military fields. An emerging means of supporting communication between nano-scale devices is molecular communication (MC), where information is encoded in the type, quantity or timing of molecules released by a transmitter. These molecules are then either carried, drift, or diffuse from the transmitter to a receiver, where decoding can take place. In the large-scale networks, with the exponential growth of the data traffic and the massive increase in the number of connected devices in the age of Internet-ofThings (IoT), future wireless communication networks have to satisfy some main requirements, such as a capability of serving many devices simultaneously, providing high per-user data rate, and serving realtime applications. Cell-free massive multiple-input multiple-output (MIMO), where many access points (APs) coherently serve all users in the networks, is one potential candidate as it can offer very high spectral efficiency (SE), energy efficiency (EE), and coverage probability. The major challenge to both MC and cell-free massive MIMO, which significantly affects on system performance, is noise and interference management. Therefore, the main motivation of this thesis is to propose methods to manage noise and interference in both MC and cellfree masive MIMO systems. In the first part, we focus on noise management in the MC systems. Two main challenges in MC, namely noise and coexistence, are investigated. In particular, we develop a detection framework that can cope with noise in scenarios where the molecules propagate according to anomalous diffusion instead of the conventional Brownian motion. Then, we develop a framework to establish coexistence conditions based on the theory of chemical reaction networks. We specialize our framework in two settings: an enzyme-aided MC system; and a low-rate MC system near a biochemical system. In the second part, we focus on design and optimisation for cell-free massive MIMO by applying interference management techniques. We propose a new method which aims at improving the channel estimation by using pilot power control technique to reduce the mean square channel estimation error. Then, we design and optimise cell-free massive MIMO systems with multi-antennas at users. A simple and insightful closed-form expression for the SE is derived. This closed-form expression enables us to formulate a data power control problem to improve the system SE. Moreover, we propose the framework to achieve sub-optimal of SE regardless the number of users in the system.