Network consolidation has been with us since the 1990s, driven by the simple requirement to reduce the costs of business communication. For IT, it is a matter of controlling CapEx and OpEx. For service providers, it is a matter of offering multiservice solutions at a competitive cost. (Remember when “triple play” was the buzzword of the day?) Consolidation has been so successful that you seldom encounter an organization these days that runs separate data and telephony networks. Voice and video over IP is proven, reliable, and cheap. And modern service providers—whether they got their start as a telephony, cable, long distance, or Internet provider—now run all of their services over an IP core.

Treating all communications as data, and sending it all over a shared IP infrastructure—or series of IP infrastructures—has also revolutionized our modern lives from smart phones to shopping to entertainment to travel. For myself, one of the most interesting impacts of technology has been how different my teenagers’ social lives are from my own when I was a teenager. Their activities are more spontaneous, their social groups are larger, and always-available communications make their activities safer.

And consolidation is still evolving. These days the excitement is around virtualization, improving the utilization of our existing communications resources.

From the beginning, one of the biggest challenges of consolidating all communications onto an IP infrastructure stems from the fact that not all data is equal. As users we expect a certain Quality of Experience (QOE) related to the service we’re using. So QOE for voice is different than QOE for videoconferencing, both of which are different from high-definition entertainment. Each kind of data stream requires different treatment within the network to meet users’ QOE expectations, and that’s where Quality of Service (QOS) comes in.

QOS has been around as long as IP has. The IP packet header has a Type of Service (TOS) field for differentiating services, and over the years that field has evolved into the more sophisticated Differentiated Services Code Point (DSCP) field to better fit modern QOS classification strategies. And from the beginning it was understood that although IP provides connectionless best-effort delivery, some applications need reliable, sequenced, connection-oriented delivery. Hence TCP, which “fakes” the behavior of a wired-up point-to-point connection over IP.

QOS is really all about managing limited network resources. You don’t get extra bandwidth or faster delivery; you just get to decide what data gets first dibs at the available resources. High-Def video requires very prompt delivery. A web page can wait a bit longer, and e-mail can wait much longer still. Over the years, QOS technologies and strategies have become more and more sophisticated to deal with the diversity of applications using the network. Routers and switches have better and better queues and queuing algorithms, better ingress control mechanisms, and better queue servicing mechanisms. And the advent of Software-Defined Networking (SDN) introduces some new and interesting ways of improving QOE.

All of this growing sophistication brings with it growing complexity for network architects and engineers. There are a lot of choices and a lot of knobs, and if you don’t have the understanding to make the right choices and set the right knobs, you can do some serious damage to the overall quality of the network. Or at the least, you can fail to utilize your network’s capabilities as well as you should.

That’s where this book comes in. My longtime friends Miguel Barreiros and Peter Lundqvist have deep experience designing modern QOS strategies, and they share that experience in this book, from modern QOS building blocks to applied case studies. They’ll equip you well for designing the best QOS approach for your own network.