Bridges provide an inexpensive and easy way to connect network segments, much as hubs and switches do. Like switches, they connect two segments on a network. Like a switch, a bridge operates at Layer 2 on the OSI Reference Model. Bridges and switches both isolate and contain collision domains within a segment. They both transmit broadcasts from one segment to another (which can lead to broadcast storms). Both also “learn” where nodes are located based on MAC addresses.
What sets a bridge apart from a switch is that switches allow simultaneous communications between any two nodes. Switches also can create LANs, much as a hub does, where bridges are used primarily to segment networks. So think of a switch as designed to communicate with individual nodes while a bridge communicates with and between network segments. Switches also can create Virtual LANs or VLANs, in which collisions are completely eliminated and broadcast domains programmed by software.
When designing a network with more than one segment, the debate often comes down to whether to use a bridge or a switch or to opt for a router. Setting up a router can be complicated. A bridge's best use is to join together networks of different media types, such as UTP to coaxial. This is especially helpful in creating larger networks, and to keep network segments free of data that doesn't belong in a particular segment.
Click
Here to View a Good Diagram of a Bridged Network
HomeNetHelp
Bridge Tutorial
A router, which operates at Layer 3 of the OSI Model, can create and connect several logical networks. However—and here’s the key difference between a router and a bridge or a switch—a router also will allow two different network topologies, such as Ethernet and Token Ring, to connect into a single network. A router provides multiple paths (compared to only one on a bridge) between segments, and will map nodes on a segment and the connecting paths with a routing protocol and internal routing tables.
Routing over a segmented network is no different than routing over an internetwork. The router uses the destination IP address (this is what makes it a Layer 3 device. Remember, bridges and switches use the Layer 2 MAC address.) to determine where a frame should go. If the destination IP address is on a segment directly connected to the router, then the router will forward the frame out the appropriate port to that segment. If not, the router will search its routing table.
When you’re thinking about hubs, bridges, switches and routers, remember that routers are the only devices of the four that will allow you to share a single IP address among multiple network clients.
This is a good place to discuss the "brouter." A brouter is a router that can also function as a bridge. A brouter can process some information at Layer 2 (MAC addresses) and other information at Layer 3 (IP or IPX addresses). How it will do this is determined by how it is configured. However, they tend to negate their own value. The most useful feature of a LAN router is to isolate certain types of traffic (such as broadcasts and multicasts) from other networks. The brouter defeats this purpose because its bridge portion will pass on those broadcasts.
RouterGod:
The online gossip rag and rumor mill for Cisco professionals
How Routers
Work
“Gateway" is a blanket term for any hardware or software system that joins together two dissimilar networks. In other words, it’s a network point where one network can enter another (like a “gate,” get it?). By this definition, many routers are also gateways. These systems are the most complex of all the network devices CompTIA expects you to know about because they translate at multiple layers of the OSI Reference Model. So hold onto your hats, we’ll be moving between the layers.
For instance, let’s say you have a gateway that connects an LAN with a mainframe. You’ll find few environments that are so different from each other. In a LAN, you’ve got distributed processing, baseband communications, and the ASCII character set. Mainframe networks use centralized processing, broadband and communications and he EBCDIC character set. A gateway, when properly configured, will translate each LAN protocol into its mainframe counterpart and vice versa.
Gateways can be entirely software, entirely hardware or a combination of the two. Depending on their implementation, gateways can operate at any level of the OSI model, though they generally operate from the Transport Layer (Layer 4) to the Application Layer (Layer 7). Gateways exist on the borders of a network, which means they are functionally related to firewalls.
Short for Channel Service Unit/Data Service Unit, the CSU/DSU is a common device found in equipment rooms where the network is connect via T-series data connectors (in other words, like a T1 or T3). The CSU/DSU connects a digital carrier (such as the T1) to the network equipment (usually a router). The CSU terminates the line at the customer site while the DSU does the actual transmission through the CSU. The CSU also can provide diagnostics and remote testing while the DSU provides buffering and data flow control. Typically, the two devices are packaged together as a single unit. Think of it as a very high-powered, very expensive modem. Such a device is required for both ends of a T1 or T3 connection and both ends must be set to the same communications standard.
CSU/DSU
A’La Alliance Datacom
GDC White
Paper: CSU/DSU Non-integrated vs Router-integrated