A report released last month by Washington-based Dittberner Associates, Optical Switching and its Impact on Public Networks, shows a sharp decline in the optical switching market both for all-optical (O-O-O) switches and for Optical in Electrical processing Optical out (O-E-O) switches.
The use of all-optical switches was stated in the report to reduce both capital expenditures (CAPEX) by 50 per cent and operational expenditures (OPEX) by 85 per cent, but adoption of these switches would be slow for the same reasons the switching market is declining.
“It’s declining primarily because of the unpredictability of traffic,” said Lilian Tau, vice-president of consulting and market research for Dittberner. “Large carriers deployed more bandwidth than they needed – this was done primarily to prevent massive congestion within networks and that was directly related to the Internet traffic which was growing quite rapidly.
“So this basically fuelled the massive capacity explosion that we’ve all seen in the past four or five years.”
He added this explosion initially worked in favour of the optical networking carriers and vendors but the result was they spent too much money for little return.
“A lot of aggressive ISP providers and service providers went bankrupt and don’t exist today,” he said. “The ones that are still active are finding themselves without any additional financing for network build-out.”
Despite this decline in the optical switching market, backbone traffic in industrialized countries continues to grow 50 to 70 per cent per year, however the market growth for optical backbone infrastructure equipment will not reach the level of the late ’90s during the current decade, mostly due to decreasing equipment prices, according to Dittberner.
“In the long-haul network you have the problem that the backbones of Internet and telephony that the traffic is growing, and you have to increase the capacity, but at the same time revenues are declining because the [costs] are getting cheaper and cheaper, and so much fibre has been installed in the past few years based on overstated assumptions of traffic,” said Dr. Lorenz Beug, vice-president, technology and network planning at Dittberner.
“So it is hard to sell anybody anything at the moment,” he added. “Also, the equipment prices are declining for a variety of reasons – improved technology, and people go to higher speeds…and this is not very interesting for the equipment vendors.”
Costs per port for both O-E-O and O-O-O switches are expected to decline about 50 per cent by 2010, while it appears the O-E-O market is declining faster, it is just an illusion, Tau said.
“The O-E-O switching market was larger to begin with,” Tau said. “The all-optical switches didn’t even exist until about a year and a half to two years ago, so the O-E-O switches were the first to appear on the market.
“That’s why the total market value was for the O-E-O switching. That’s why it seems like it’s declining faster but they’re declining at the same pace.”
However, there is one area where Dr. Beug and Dave Lively, senior manager of optical strategy at Cisco Systems Inc. see an area of growth: in metropolitan area networks (MANs) and regional networks.
“I think we are more optimistic in the short run for applications in the regional and metropolitan area network area,” Dr. Beug said. “Certainly for metropolitan area networks and regional networks you definitely need optical equipment and optical switches. And you very nicely employ all-optical switches but probably of different configuration from backbone networks, so that will be the same technology but in somewhat different configurations.”
Lively said he was not surprised at the results of the Dittberner report, and said the market is seeing a shift from the core to the metro.
“Enterprise customers are starting to build out into the metro area,” he said. “And are now tying them into high-speed Ethernet services.”
Lively also said another factor was the technology needs to be improved.
Indeed the Dittberner report said that all-optical packet switching was “far from practical” because several necessary components are not available yet. The technological breakthroughs aren’t expected until 2009 or 2010.