Summary of NGN Ventures Conference

At the closing of this week's NGN Ventures Conference in Burlingame, California, Co-Chairmen Dr. John McQuillan and David Passmore summarized some of the most notable themes of the event. McQuillan also announced that this was his last year as co-chairman of NGN Ventures.


In Last Mile networks, there are many different technical solutions. All are likely to succeed in some way, but each in specific environments. Last mile networks and systems are gearing up to deal with multimedia deliver, and regulatory uncertainty will continue to influence decision making. FTTH will happen in residential deployments before business, and it now looks likely that these will be PON.


In Metro Networking, we were reminded that the leased line business is still 50 times greater than Ethernet. Things are still moving slowly. Start-up equipment vendors face a major challenge in selling to an RBOC, which never want the risk of representing a majority of a key supplier's revenues. This category will be marked by big customers who buy startup technology from big vendors.


A problem with trying to eliminating OEO conversions from Optical Architectures is that there is still a big need for add/drops along network. In addition, it seems that the statements that transponders are the most expensive part of an optical network may be overemphasized as the key economic issue. An optical enterprise market is emerging, with plenty of CWDM equipment being introduced to target businesses.


The business model crisis facing the Service Provider industry is frightening. For existing service providers, to move from current networks to new networks will be extremely challenging. The network cost structure must be cut dramatically. In addition, an all new vendor approach may be needed, possibly along the lines of an integrator model.


Security is full of opportunity. Intrusion detection systems are evolving into intrusion prevention platforms. This requires an inline box that acts as a gatekeeper limiting packet access to the enterprise. Meanwhile, denial of service attacks needs to be tackled in the core, and cannot be done at edge. There appears to be a significant opportunity to consolidate the growing number of security boxes and applications to simplify management.


The Storage Networking market has clear needs and fast payback potential. The possible solutions are all very different, but all are part of deconstructing the old mainframe system. One problem is that there are far too many storage networking start-ups, and there are unlikely to be a lot of successful companies until many competitors fail. Service providers have not stepped up to the storage challenge. They say they want to go up the value chain, but they are not doing it. Enterprises are taking care of their own storage needs, and they may not outsource this.


Enterprise Networks are increasingly using VoIP, but it is primarily for trunking between PBXs. It will take time and better network QoS for VoIP to be implemented further. Data center consolidation is taking place but with the use of blade servers and new SANs, enterprises are facing the same problems of equipment proliferation and scaling as service providers. There is a trend for enterprises to take back some of the functionality they had been giving to service providers, particularly storage. Service providers are having trouble running their own businesses, and as enterprises are getting access to more and better technology, they are doing more on their own.


VoIP is the first bridgehead in the Session Controller battleground, but it is not clear how it will play out after that. Will service providers deal with this? Will it only happen at the enterprise level? This is probably an over-funded segment for the time being.


Already facing mobile voice substitution, ILECs are experiencing further problems caused by Next Generation Voice technologies. The barriers to entry for voice service providers is very low. You can create the equivalent of a central office for just $1 million. Meanwhile, it looks increasingly likely that customers will not buy voice services on their own. Enterprises want data with it, while residential customers want triple play voice-data-video. Voice will not be the same as telephony any more. It will be one of a set of services, giving voice a different character in the future.


Ethernet is going into the public network, but there is still a debate about the exactly how it will take place along the way. The biggest question is evolution vs. revolution: is Ethernet just a service that to be delivered over existing SONET/SDH, or should carriers start to build Ethernet as the metro infrastructure? Currently, the center of Ethernet activity is in Asia. It may be in the next year that we will start to see what happens in the US.


There is an incredible diversity in start-up approaches to Wireless technologies. However, the greatest challenge they face is the incumbents. The existing players will want to augment their wired Ethernet switches with the kind of technologies that the start-up players want to offer. This space may be determined more by incumbent switch vendors than the startups. While we know that WLANs will be big, it is not clear that there will be a great new company in the space.


The resiliency of the TCP/IP model is amazing, but it clearly is riddled with many flaws. There are many innovative ideas about addressing these problems. It may be very difficult to get service providers to make a big change to a new type of equipment, and there was much discussion about small appliances for enterprises. However, it is challenging to repeatedly call on scores of large enterprise prospects to get them to set up spot to spot appliance solutions.


In their concluding remarks, McQuillan and Passmore said that start-ups in today's environment are experiencing a great deal of tension over pursuing incremental vs. radical approaches. Start-ups with increment solutions may bring in short term revenue, but may not be long lasting or reach a large scale. Companies with radical approaches may have great ideas for where future networks should go, but they may not be able to survive while they await a more favorable economic climate.