Jim Machi is senior vice president of product management and marketing for Dialogic, responsible for driving the overall product roadmap, programs and communications, and has been working to develop the company’s IP telephony strategy. Recognized by Internet Telephony Magazine as one of the “Top 100 Voices of IP Communications” and by WebRTC Expo as a “WebRTC Pioneer,” Jim recently sat down with SDxCentral to discuss the current state of network functions virtualization (NFV) and where he sees it heading.
SDxCentral: What impact are you seeing NFV have in the market?
Machi: There is a lot of hype in the market on NFV, but also a lot of excellent work being done trying to rationalize what effectively is a huge game changer for service providers and NEMs when it comes to building networks and delivering services to customers.
We’re already seeing NFV, and more specifically, virtualization, having a profound effect on how our customers are planning and deploying networks. Even more importantly, if you look at how things are shaping up, it’s also going to have a huge impact on the pace and rate at which new services are developed and rolled out — not to mention the costs involved. The barriers to entry for deploying infrastructure for both traditional and non-traditional service providers really change when you take hardware component out of the equation.
For Dialogic, the whole cloud aspect is affecting how we provide value to our customers. With software-based solutions, we are delivering to our existing customers the functionality they need over a private or public cloud. This model will enable new participants including OTT-type players, operators focused on connected devices and the Internet of Things (IoT), enterprises and other operators the ability to roll out services with significantly lower costs.
The bottom line is that we see NFV reinforcing a trend that we’ve been focusing on for several years, namely the migration of network functionality to the cloud, and we see it having a profound effect on the telecommunications and information technology market place.
Are all your customers interested in NFV?
Machi: Most of the operators we work with have mentioned NFV to us. But, NFV is more than just separating the network function software from the proprietary hardware; it also involves additional management and orchestration functionality that controls how application capacity is scaled out and up, how applications are chained together to support an end-to-end service and the interaction with the underlying virtual compute, storage and networking infrastructure. The scope of NFV includes all these moving parts come, but service providers don’t have to wait to really start to take full advantage of the benefits of a cloud infrastructure.
And not every one of our customers have the resources like the larger tier one service providers to be on the forefront of this technology turn, or is even ready for deploying a comprehensive orchestration capability. All operators have same need to stay on the leading edge of technology. They all want to lower the cost that’s currently entrenched in the traditional approach of deploying proprietary hardware, and all are interested to moving towards a software telco concept. So our development reflects the needs of our diverse customer base regardless of where they are on the NFV or virtualization deployment spectrum. They want to start out by moving select functionality in the control plane and data plane to software and that’s where Dialogic specializes.
What are Dialogic’s NFV and cloud infrastructure offerings?
Machi: Dialogic has a rich history of software innovation and is helping service providers and application developers reimagine how core network capabilities and value added services are deployed. Since the early 2000s, Dialogic has been actively moving digital signal processing (DSP)-based solutions to software and we supported this movement at the time when running thousands of channels of media processing on a COTS platform was not thought possible. Improving the performance of media functionality in virtualized environments to meet operator expectations is not a trivial task. Dialogic saw the possibilities the future held, and embraced this move to software early on.
In the context of NFV, Dialogic’s solutions are primarily virtualized network functions (VNFs) operating over a service provider’s virtualized infrastructure or as software applications running on COTS-based hardware. This puts Dialogic customers in an advantageous position since the applications are not extensively tethered to proprietary hardware platforms like some other vendors’ products. Dialogic’s virtualized offerings include the BorderNet™ Virtualized SBC, PowerMedia™ XMS media server, PowerMedia™ Media Resource Broker (MRB), and PowerMedia XMS transcoding capabilities, and Brooktrout® SR140 Fax Software. We’re applying the same transformational approach to our product roadmap that includes plans for virtualized load balancing and ControlSwitch™ System softswitch, to operate as VNFs. Here’s a proof point: As of 2015, Dialogic has deployed more than three million production ports globally of real-time, rich media processing on PowerMedia XMS media servers— all software-based and running on COTS servers.
We also plan to implement VNF management capabilities to manage and orchestrate our virtualized applications, and do so in an open manner by enabling APIs to interwork with other open third party VNF management and NFV management and network orchestration (MANO) solution vendors.
Your focus is on VNFs. Should service providers pay attention to what’s going on at this NFV layer?
Machi: That’s a good question because there’s a big difference between architecting an application to run on COTS and an application to run in a virtualized cloud environment, more specifically NFV. A lot of vendors are taking their monolithic software applications currently running on proprietary hardware and dropping them into a virtual machine and calling it a day. There’s more to NFV than that if you really want to take advantage of the cloud benefits of elasticity and flexibility. In addition, virtualizing real-time multimedia applications like media resource functionality is not a trivial task. If you want your service to have the performance and functionality needed, care must be taken in selecting a VNF designed to perform in cloud environments.
There are a few guiding principles when it comes to virtualized applications that operators need to consider as they start selecting vendors to help them move functionality to the Cloud Infrastructure..
First, automation, scalability, and programmability are not “nice to have” concepts when moving to NFV, but rather these should be “must have” goals. You want to automate as much as possible how application resources are scaled and where they are deployed.
Second, software modularity is critical to optimize VNF application performance and scalability and realize the full potential of a virtualized environment. Scaling up large monolithic application functionality takes time, and in order to respond rapidly to changes in demand you want to only scale up functionality needed as well as geographically where you need it in order to reduce the delay that can affect performance of real-time functions. With 5G, we start talking about sub-millisecond response time for applications, so speed is important and an intelligently architected VNF that is decomposed in a smart manner geared for virtualized environments is important.
Finally, VNFs should be architected for flexibility to allow operators to take advantage of technology advancements in cloud virtualization like containers without having to go back and retrofit their applications. Again, we’re talking about reducing the overhead of virtualized environments which allows applications to reduce the time it takes to scale from minutes to seconds down to milliseconds.
These are concepts that Dialogic is integrating into our technology as we work with our customers on their telco cloud initiatives.
What are the benefits you’re able to deliver to customers?
Machi: Naturally, two benefits of this move to software are opex and capex savings, but more importantly, you’re going to see increased flexibility. For service providers today there is a huge need to be able to turn up more services at a faster rate, so rapid prototyping and testing is important. Also, in order to address niche markets more effectively with targeted, personalized services, you’re going to need to be able to roll out more services at a lower cost. This is an area where our software-centric solutions help today, and we’re making considerable investment to help service providers accelerate their time-to-market even more.
A software-centric approach for infrastructure deployment checks all the above boxes. You can spin up infrastructure very quickly for rapid development and you don’t need to incur massive up-front costs like you do for hardware-based approaches. With agile development techniques you can reduce the cycle time of developing and fine tuning applications. The old approach involved sourcing test platforms, developing applications, testing, and then wide scale deployment without any guarantee that the service would be successful. A cloud-based approach allows you to develop and deploy in software at a rapid pace with reduced cost and lower risk of not getting it right.
Also, with our virtualized cloud-based offerings we’ve positioned ourselves to be able to add functionality faster compared to applications from other vendors tied to a hardware platform. For example, WebRTC-based services are based on next generation codecs. With a software-based architecture, we are able to implement these faster than a traditional method relying on firmware and DSP cards. Service providers can also leverage the elasticity of the underlying virtualization layer to improve availability and scaling out applications in the event of a failure. Flexibility, speed and lower operational costs truly tip the scales for our software-centric approach and our customers are responding well.
What role do you see open source playing with NFV?
Machi: We see open source playing a critical role in NFV. To make this transformational technology successful, we need open collaboration to build a common foundation for the future of networks. We are a member of the Open Platform for NFV (OPNFV) and our VNFs ultimately would be deployed on the open source reference platform development currently underway by the organization. NFV is becoming widespread in the market and will open the door for innovation, broader participation by companies looking to add value, and new business models as well. Industry-wide collaboration on an open NFV ecosystem will allow critical concerns to be addressed quickly and enable NFV to become mainstream.