There are many types of networks in computing. You can define networks from the way that computers connect to the internet and each other (WAN/LAN) to the way that internet addresses identify themselves (TCP/IP). When it comes to bitcoin, there are two network models that matter most: P2P and client-server.
P2P networks communicate between “peer” computers which are connected to each other via the internet. There is no central server in a P2P network, so files and data move peer-to-peer through the network rather than through a central repository.
P2P networks are highly secure because there is no central repository which holds information about the other peers in the system. However, P2P networks require more care to ensure that the individual connections between peers are active.
Client-server networks use a central server to connect clients in a network. These networks inherently are less secure and require more security attention. If the central server were to be compromised, then all the clients are vulnerable as well. That said, many if not all of the “cloud” and “web 2.0” applications of today are run with a client-server model because they are highly efficient at moving data and files quickly without much maintenance.
Note: In both cases, “peers” and “clients” refer to individual computers. In bitcoin, peers and clients are more commonly known as nodes.
Bitcoin owes its resilient and fraud resistant reputation to the marriage between P2P network architecture and the distributed ledger blockchain. In this way, Bitcoin is the world’s preeminent decentralized money.