Making connections

12 Jan 2010

by: By John Adams

Common connectivity devices electronic security installers will encounter on networks include repeaters, gateways, hubs, brouters, switches, bridges, routers and gateways. Integrate security to a network and you’ll run into some or all of these.

WHEN electronic security installers get involved in networked security solutions they’re going to find themselves facing a host of connectivity devices, the most common of which are repeaters, especially if the network is employing Cat-5 to get around a relatively large multi-story facility.

Repeaters are essential if a larger network is going to operate effectively. These devices connect different sections of the cable plant, receiving the signal and pumping it back up to full strength to combat attenuation caused by cable impedance. This done, the signal is then sent on its way. Think of repeaters as existing in the physical layer of a network where they support network media comprising the cable plant.

Depending on how much you spend on repeaters, there are 2 different ways to buffer a signal – the first being that you can re-amplify it. This is an effective technique, especially if there are not too many repeaters in the system. The trouble with amplification is that it will also boost signal noise so that over time a re-amplified signal will degrade. Baseband systems reamplify signals by kicking in power as the signal passes through the repeater.

The best repeater amplification method is signal regeneration. This is easier for digital systems because they’re built around a digital code and signals can be completely recreated and then re-transmitted fresh - as if they’d just left the originating NIC. Broadband systems always regenerate signals.

Next come bridges, routers and brouters. A bridge is a device that links network segments and represents an integral component of both segments. When you think of network geography consider that a NIC and the network device it connected together comprised a node, with a number of local nodes connected to a hub. Once signals travel through the hub, the next piece of hardware they’ll come to is the bridge. Imagine 2 groups of nodes, each connecting to their own hub and network cable running from each hub and into a bridge. The bridge operates on the network’s data link layer.

“When you think of network geography consider that a NIC and the network device it connected together comprised a node, with a number of local nodes connected to a hub. Once signals travel through the hub, the next piece of hardware they’ll come to is a bridge”

In terms of network geography, this bridge is a component of both network segments – it has an address on both parts of the network and all the packets of data sent from all the nodes attached to both hubs will pass through the bridge. They’ll then be passed on to other bridges responsible for different network segments.

Inside the bridge is a list of addresses and the bridge uses this list to check whether an incoming packet is destined for one of its nodes. If so, the packet is passed to the appropriate node NIC for processing. Bear in mind that bridges aren’t the perfect answer on large networks, though they work fine in smaller or carefully designed architectures.

The weakness of a bridge is that if the address of a packet received is not on the network segments that bridge holds addresses for, it will send out a network broadcast looking for the address of the homeless packet. In a quiet network such an event is generally not an issue but if there are many such packets loose in a network then mass signaling can cause performance drops.

Routers are similar to bridges in some ways and function by breaking a large network into segments. The beauty of a router is that it can not only direct traffic to nodes on its own segments – it has a full list of network addresses and is able to direct traffic to the right location wherever that might be on the network.

In a large system, maintaining router tables is virtually a full time job, as every router on the network needs to have a fully updated version of all addresses on a LAN. These addresses will include those related to DVRs and security management servers, though in many cases these systems may be running on a local VPN between a select number of machines.

A cool thing about routers is that the larger a network and the more subnets linked together by routers, the more paths there are allowing packets to get through to their destination should there be traffic jams anywhere in the network. Routers are also designed to establish the quickest way for a packet to get through a network.

Routers operate in the network layer of a system and they have a logical network address allowing them to pass data packets to NICs located outside their own network segment. An excellent feature of routers is that they can not only connect to a primary Cat-5 network segment. These devices are also able to link with other network media such as coaxial cables if routable protocols, a bridge or a brouter are installed. 

Routers pass information between themselves using either router information protocol – that’s the RIP – or the open shortest path first protocol (the OSPF). When RIP is employed the connected routers pass their entire routing table between themselves, while OSPF updates variations to the table only.

Meanwhile, brouters function at both the data layer and physical layers of a network. For these devices, operation depends on the protocol that reaches them across the network. Routable protocol packets are routed, while non-routable packets are sent on their way using the physical address of a node is a connected subnet – it works the same way as a bridge.

Last of all the network connecting devices is the gateway, which converts signals from completely unrelated networks and allows communication between them. Such devices exist in the form of software on a server and might translate communications between a Cat-5 PC-based network and UNIX-based operating system for instance.