Probe-based congestion management doesn’t have to be expensive
February 22, 2011 - Openet
In response to last week’s unfiltered blog post “When are probe-based congestion management solutions ideal,” I got a response from Tekelec that indicated probe-based approaches don’t have to be as costly as Gary Rieschick, senior product manager with Openet, indicated.“You can actually put a small number of probes in certain places in the network to do a solid job of measuring RAN congestion,” said Randy Fuller, head of strategic marketing with Tekelec. “If you don’t have a good source of data, you can use the probe as a network element that just spit out usage records to help you.”
He suggests that one probe at each GGSN or home agents can be sufficient. “That means tens of probes, rather than hundreds of them.”
In other words, if operators just eavesdrop in a minimal sense, without trying to measure throughput of every element in the network, the probe-based approach doesn’t have to be a necessarily expensive one. “As network elements talk and as signaling takes place, you just listen in to get what you need; it doesn’t have to be a multi-million-dollar thing,” added Fuller.