I’ve written about the importance of networking in the organizational introduction sense, so now I’m turning to the importance of understanding the networking technology that has become vital to successful clinical device operations. The network, big or small, provides the necessary infrastructure to exchange data and while big networks usually correspond with more IT staff, it doesn’t necessarily mean they’ll give any more attention or expertise to clinical engineering operations. Conversely, a small network doesn’t necessarily mean that it is easy to configure or operate.

Either way, it’s extremely important for CEs to take the time to understand and cooperate in the integration of network medical devices on the wired and wireless infrastructure. I’m only going to scratch the surface and present a selective overview, so know that there is a tremendous amount more detail on numerous aspects of what I’ve written below.

The first thing a CE can do is to understand the network topology in his or her facility. The topology is the layout of how the network systems interconnect. Systematically, this architecture is represented in the connections (physical or otherwise) and their supporting configurations. In design, however, this architecture is frequently represented by a set of diagrams, which represent the nodes, switches, routers, and other hardware that make up the local-, wide-, metropolitan-, (and other) networks.

At the end are all the network devices (nodes) that connect to a switch or access point. In the case of wired networks, the nodes interconnect using copper wire or fiber optics. For copper wire, usually the first connection (patch) is made to a wall jack which traces back to a wiring closet with labeled patch panels corresponding to the labeled wall jack. Those patch panels provide a structured capability to connect the devices patched to the wall jack more easily into the switching infrastructure. For wireless devices, the nodes negotiate connections directly to wireless access points (WAPs) by radio. Those wireless access points either connect directly to switches or connect through an intermediary wireless access controller device, which coordinates wireless nodes through multiple WAPs.

The first layer of switches to which nodes connect are typically referred as “edge” switches. This is because the edge switches are usually connected to yet more switches referred to as “core” switches. The purpose of a network switch device is to control the flow of network traffic to the most appropriate switch port. In most cases a network device sends a packet to the switch with a specific target address, but in some cases the device is either not aware of the target address (yet) and/or has a general broadcast to make. In the case of broadcast traffic the switch’s most appropriate port could potentially be all ports. Any target or broadcast address that does not exist as a device on the current switch could then be passed to additional connected switches in a best effort attempt to arrive at the target. In larger network closets, many edge switches are arranged in a stack which allows for them to operate more efficiently together as if they were one large switch with many ports.

Networks are arranged in sets of subdivided addressing space. These are called subnets. Nodes that send packets to the switch with subnet space outside their own should designate a default network gateway, which can help route the packet to the correct network. The switch(es) will pass the data along (“uplink”) until it arrives at the designated router node. A router will consult a table matching subnets to network interfaces (ports) much as a switch has a table matching node addresses and ports. If the router is directly connected to the designated subnet it will pass the packet on that interface, otherwise much like the initial node, the router may have a default network gateway path of its own. Typically Local Area Networks (LANs) and Wide Area Networks (WANs) have routers that are fully programmed and aware of one another to make the most efficient routing decisions and provide failover capabilities. Border routers exist to route traffic outside of internal LANs and WANs to Internet Service Provider routers that permit passing traffic on to the Internet.

This very brief overview of basic networking systems provides context for how devices connect and pass network packets with one another. The next level is to understand more about what those packets contain and the nested data protocols that facilitate information transfer. In a future article I will continue to cover common protocols such as ARP, DNS, STP – many other alphabet soups.