What is latency, how serious is it and how do I get rid of it?

Latency is the amount of time it takes for a packet of data to go from the starting location to the final location. Together the amount of latency and bandwidth dictate the quality, speed and capacity of a network.

Once latency or a delay is present, it cannot be optimized. This means that the cause of the latency will have to be removed before the network can work at the optimal performance.

The delay or latency we are talking about is measured in nanoseconds. These nanoseconds are measured so to know the difference in the time of the request and the completion.

Therefore, you would need to use an internal device rather than an external device to remove the problematic latency caused by the serial port. To maximize your throughput, you will want to use the highest amount of bandwidth you can get.

Latency is often referred to as wasted time. With many of the objectives we do that require bandwidth, speed, clarity etc in a network latency is a proven hassle. Therefore, it is important to remove as much latency as possible.

The delay in the packets sent in VoIP is going to affect the overall sound and call quality. This is important because the way you sound and the call stability is a factor in communication that needs to be the highest quality.

Have you ever had a slow computer that when you play a game or watch a video it is choppy, therefore it is hard to pay attention, get the point of the program or down right so frustrating you give up? Well with the higher levels of latency in a network, it can be like this. Severe latency can also sound much like a skipping CD or record.

Maybe on a very light end of it, because there are services and technologies now that remove the latency in the packets travel time, but originally, when the technology was new this was a problem.

The users usually really begin noticing latency as a service level issue when the completed roundtrip latency is greater than 250 milliseconds (ms). The ITU or International Telecommunications Union usually recommends that the delay or latency never exceed 150 ms one way (from speaker to listener).

In considering that an analog call is the benchmark by which VoIP is graded, 75 Ms of latency is acceptable. Most VoIP researchers are those who consider VoIP to be a toll by-pass or a alternative application that allows for a certain amount of reduced quality to achieve a higher amount of costs savings or more quality features set. Given the trade off, an acceptable conversation for business grade communication can be achieved in the 60 to 80 Ms range.
In order to effectively control latency, administrators or the technician who hook up your network or internet should focus on reducing delay at the network endpoint and therefore prioritizing traffic over the network. Optimization of packet size are good first steps to improving service quality at the endpoint. Endpoint delays are usually reduced by adhering to a standard packet size, using the G.711 codec, and avoiding asynchronous trans-coding.
Prioritizing VoIP traffic over the network at the working Layers 2 and 3 also help to yield latency improvements. Policy based network management, that much like the appropriate use of bandwidth, bandwidth reservation that like allotted badwidth per users, the correct Type of Service, Class of Service, and Multi-Protocol Label Switching (MPLS) are all widely used techniques for prioritizing VoIP.
In summation it is important to clear up as much latency as possible. The best ways to fix this is to consider the traffic going through the network. It is also important to have enough bandwidth to cover the uses you want for your internet and telephone services.