Two weeks ago I was in New York for Telx's CBX conference, to be on a panel discussing video on the network and business to business video. During my time there I got to meet with Shrihari Pandit, CEO of Stealth Communications and learned about what they are doing. Stealth Communications is providing a number of disaggregated network functions and creating a revolution in how voice (and perhaps video) is connected.Stealth has created what they call the Voice Peering Fabric (VPF) which is a layer-2 network dedicated to peering voice traffic. The network is provisioned to support all the connected bandwidth, and only carries voice, so low loss and low jitter are guaranteed. The network appears in Stealth PoPs across North America and in London. Users or service providers who wish to peer voice traffic with other users connect to the VPF and just pay for the port bandwidth. All network usage is included in this one flat rate.
Stealth also has an ENUM service that resolves E.164 addresses (phone numbers) into URIs for making the right connections. Service providers list their supported phone numbers in the ENUM data base. An ENUM lookup of a phone number will then return the URI for the network connection of the service provider supporting that phone number, and a direct connection across the VPF can be made.
Recently Stealth has separated the ENUM database from the VPF and made it free. This means that any service provider, enterprise or other entity can list phone numbers in the ENUM data base and get URIs independent of their transport choices. So two Internet-based services could use the ENUM data base to connect without ever passing traffic through Stealth.
Within the ENUM service, Stealth has recently created the concept of groups. This means that users of the ENUM service can create groups that can only call within the group, thus providing call security. If an enterprise wants to use the ENUM service for connecting remote offices within the company they can specify an enterprise group. Likewise if business partners want to provide voice interconnect for managing a project or an ongoing relationship without opening voice communications to the rest of the world, they can create the appropriate group. Members can participate in more than one group to get the flexibility they need to meet specific business objectives.
Will this work for video conferencing as well? I think it will. The transport is the easy part. Video will require more bandwidth than voice, and we have to take a look at the burst characteristics to make sure there is sufficient bandwidth to cover them. But otherwise it looks a lot like voice.
The signaling is the interesting part, where we both need to evolve beyond H.323 and at the same time maintain compatibility with endpoints that don't have keyboards for typing URIs with alpha characters. Video really has three different approaches today: E.164 that evolved from the ISDN days; systems that are integrated with telephony solutions like Avaya and Cisco; and the new wave of personal video using presence for signaling.
By disaggregating the signaling from the transport, Stealth has opened up the possibilities for changing how we find the other guy. I think this kind of open environment will create rapid innovation and we will soon find solutions that fit our different modes of communications but still integrate with a common communications infrastructure.
