Network speed is often thought of as being the bandwidth of the link, however there are many sources of latency (delay) that can affect overall performance and throughput. Low latency networks are typically described as being networks that operate close to their predetermined bandwidth, whilst high latency networks often suffer from large delays, that might not effect throughput but will show considerable delay. An example of a low latency network is short distance optical fiber network, which is very low latency and will provide very high throughput with minimal delay. On the other hand a satellite link will provide reasonable throughput but large delay – due to the distances traveled.
The reasons for latency though, or delay, can be down not just to distance – you can have severe delay on a LAN – it can come about for a variety of reasons. Some of the most common reasons for latency is network congestion. This can come about through the network being too busy, all the bandwidth is being consumed, packets are being dropped, which requires resending, and that can compound the problem. Other reasons are poor cabling, kinks or twists in cabling or poorly constructed or wired cable connectors. On optical networks, transceivers may need cleaning or just periodically reseated as that is another common reason for unexplained latency.
With regards voice networks latency is a concern, but not to the extent most people imagine. Voice typically requires one way latency between parties of 150ms, but that is a large figure and easily obtained even over the internet – remember a ping is two ways so that will equate to a round trip delay of 300ms.
If on a WAN you are experiencing unacceptable latency you should take this up with your service provider. Failing that you could check it out yourself by examining the traffic crossing the link, and either filtering at the router unwanted traffic or looking into deploying a WAN optimizer.