Metaswitch Networks announced this week that it was open sourcing its core IMS code, to enable operators and others to build and deliver cloud-based voice and data applications more efficiently and at massive scale.
The open source initiative, named Project Clearwater, is designed to enable the core functions of an IMS to run in a cloud environment.
Phil Harvey, Director of Corporate Communications at Metaswitch, said that the fact that Clearwater is designed for the cloud is its most distinctive aspect.
“The open source part is to an extent the secondary headline. The real interesting thing is that this is about designing from the ground up to get IMS core functionality into the cloud. Once you really think in that way you can realise the sorts of economics which are not possible using proprietary hardware. When you add in open source licensing that becomes an even more compelling offer,” Harvey said.
“Clearwater creates a massively scaleable IMS environment and you can only do that in the cloud. If we take the IMS core and use the same architecture as before we’d be missing the point. Telcos used to have the highest scale systems in the world and now Google and Amazon and Facebook do. So what we wanted to do was emulate their philosophical approach to programming, APIs and experimenting with code. And another component that’s interesting is the idea that instead of reinventing everything you’re doing, reuse as much code as possible. Our team looked at all different open source code out there and incorporated something like 50+ different pieces of sofware into Project Clearwater.”
Clearwater is not a complete service environment – it is an IMS core, providing SIP-based call control for voice and video communications and for SIP-based messaging applications. When deployed as an IMS core, Metaswitch said that Clearwater incorporates Proxy CSCF (Call Session Control Function), Interrogating CSCF and Serving CSCF, together with Breakout Gateway Control Function. It also includes a WebRTC gateway, and natively supports interworking between WebRTC clients and standard SIP-based clients, using SIP over WebSocket signaling.
To build services and applications, operators could deploy Clearwater as an IMS core in conjunction with other elements such as Media Resource Functions, Media Relays, as well as other service resources such as more advanced Telephony Application Servers.
“For us, we have made something where the economics works for carriers by taking the cost of deployment way down. But it requires some amount of effort to deploy – it’s not a completely packaged product. As an open source product carriers will need to take development in hand and tune it to their network. But the whole process is closer to where they need to be to offer IMS based services than if they are going down other routes.
“One thing I do want to make clear is that this is an alternative, not a replacement. We’re not trying to go in and have people ripping out their IMS core and replacing it. But for carriers and non-carriers that have thought about things like VoLTE, hosted business services, class 5 switch replacement, IMS has been a stumbling block in the way of those services.”
For us it’s about the components surrounding the IMS core that we sell…Getting carriers going with IMS in a more quick and direct way would benefit us.
So how does Metaswitch hope to make money from open sourcing its own IMS core code? Real-world deployments will require connectivity to other networks through the standard IMS interfaces for interconnecting with other IMS networks and with SIP-based media gateways for interconnecting with the PSTN over circuit-switched trunks.
“For us it’s about the components surrounding the IMS core that we sell. Clearwater has a built in Telephony Application Server but it is not fully featured. It offers basic functions like standalone call, voice business services, more mass market calling features. Something richer than that requires app servers, SBCs, media gateways to talk to the PSTN: we sell and offer all sorts of things surrounding the IMS core.
“Getting carriers going with IMS in a more quick and direct way would benefit us. We’re also going to take a business model to offer professional services to enhance the code, integrate, make software subs available with support commitments attached to them. The project is under GPL licence, anyone can download the code and we stand behind it.”
So how could the open source model affect others in the market. Does Harvey see it as a threat to other equipment providers, that operators could develop around and open source IMS core, rather than be dependent on vendor specific integrated offerings?
“Any vendor that has already got IMS assets in place can look at what we’ve done, take the code and use it to improve their processes. What we really hope is that they will roll their sleeves up and make some improvements and put that back out there as open source so we can all get telcos going and give them more momentum to offer services based on IMS. It doesn’t come out in a threatening way to anyone’s IMS core installed base. It’s about encouraging the industry to to start thinking in a new direction. Telcos are really fired up about NFV, are being very vocal about how to run networks and typical procurement and proprietary hardware is not in that mix. This is one of first steps to get the industry going.”
Recommended further reading: SBCs, NFV and the software operator