SIP Trunking — QoS at Its Simplest

Quality of Service (QoS) has been a contentious subject for ages. Even now, next generation network architects are devising sophisticated QoS frameworks in hopes that operators will be able to offer a range of service quality levels at a range of prices and use QoS to support VoIP versions of their bread and butter application — mobile voice telephony. Yet, despite decades of hype, the best-efforts Internet beat all previous approaches and remains a best-efforts network today — no QoS!

Why? The QoS frameworks are too complex. Real requirements for QoS are either non-existent or very simple. Most applications can work around delays and lost packets. For example, streaming video can be encoded to fit within available bandwidth and buffered at the receiving end. The exceptions are telephony and interactive gaming where latency of more than a few hundred milliseconds noticeably degrades the service. So it’s worth looking at how IP-PBXs are being connected over the Internet today; i.e., how VoIP QoS is handled in the real world.

The first thing to notice: the only problem is on access links. Once you get beyond the access network, every link in the Internet — local, regional, national or international — is carrying multiplexed traffic from many users. Multiplexing many, many, bursty flows results in relatively predictable volume. Traffic volumes vary by time of day, but these links don’t saturate, except as a result of poor engineering or forecasting on the part of an ISP, or failures in other parts of the network causing rerouted traffic. Either case generates a rapid response from any ISP that expects to remain in business. So “best efforts” in the Internet core means sub-millisecond delay variations and near zero lost packets.

LAN administrators know it’s easier to throw bandwidth at a problem than establish and manage a sophisticated QoS environment. But with access links, extra bandwidth is typically expensive or unavailable.

Increasingly IP-PBXs are being connected via SIP trunking so it’s instructive to see how SIP trunking handles QoS. The first and perhaps still the most common approach is to split VoIP and data traffic on two separate Internet connections, with the VoIP connection sized to preclude any potential congestion.

When access links are shared, the typical solution employs a QoS router at the customer’s end. This gives absolute priority to outbound VoIP packets and protects inbound VoIP packets by active traffic shaping; i.e., by signaling remote TCP hosts to throttle inbound TCP data flows so there’s enough capacity for inbound VoIP packets.

There are interesting lessons here. Real world QoS requirements are for just two classes, with VoIP bits getting absolute priority over best efforts bits. Very simple! Next generation network architects should take notice. IT

Brough Turner (News - Alert) is Senior VP of Technology, CTO and Co-Founder of NMS Communications (www.nmscommunications.com).

» Internet Telephony Magazine Table of Contents