Delivering 'PSTN Quality' Voice on the Internet
Telnyx CEO David Casem shares on his company vision to bring quality voice calls over the Internet.
I use the Internet regularly to record podcasts and interviews. Voice quality usually sounds pretty good until I play back the recording and I hear the packet dropouts and distorted speech. You may have encountered similar scenarios.
In my experience Internet voice is not as good as the PSTN in voice quality. I can't help but think, wouldn't be nice if you could find a way to talk over the Internet and connect to the PSTN and retain the voice quality?
The Internet was not designed for voice, yet we are seeing a growing demand for communication to be application-centric. The only network ubiquitous enough to facilitate application-centric communications is the Internet. As a result, end users must often make a choice between high quality calling and integrated communications.
At Telnyx, we have begun to shift this paradigm. By leveraging a purpose-built network, our customers can now offer mission-critical, carrier-grade voice services that are fully interoperable with the Public Switched Telephone Network (PSTN) over a common Internet connection.What industry trends are you predicting?
Calls are an important aspect of getting work done, especially for remote workers who need to stay in constant communication with an increasingly distributed workforce. Because building out high quality networks for remote workers is not economically practical, anyone collaborating or communicating with them is at the mercy of that worker's Internet connection.
Most real-time, corporate communication happens over MPLS networks, where the quality of voice is guaranteed and prioritized over other types of Internet traffic. But this kind of network is expensive to build and maintain, and often requires a circuit that can only be delivered by the Local Exchange Carrier.
Furthermore, organizations are looking to drive efficiencies by enabling unified communications. This trend was evident at Salesforce's Dreamforce conference this year. The number of Salesforce plug-ins/partners that are offering a solution that enables calling is staggering. While there will be a place for UC solutions like Skype for Business for quite some time, we believe most workers (both on-site and remote) will soon be using purpose-built tooling that incorporates communications operating in-browser and in-app.What are these workers looking for when working collaboratively?
Ease of use and quality are the two biggest drivers of adoption of tooling by remote workers. Tool adoption often directly correlates to ease of use. By enabling communications from within systems that workers are familiar with, there is a higher likelihood of adoption.
When you're collaborating with a remote worker through a voice or video call, the calls needs to be high quality to ensure productivity. When making a sales call to customer from your CRM software, the importance of being able to communicate clearly and effectively is even more significant.
There are three metrics that relate to the quality of a call: latency, packet loss, and jitter. The farther a user is away from a VoIP provider's Point of Presence (PoP), the more these metrics may increase, in turn increasing the probability of a suboptimal call. Some providers have attempted to circumvent this problem by staying out of the media path (often called direct audio). This just passes the buck to the customer's ISP. There is only one way to ensure call quality -- get the call off the public Internet as soon as possible.Do service providers support these requirements?
Many providers aren't supporting these requirements, and the problem is that most of them are resellers and not original providers. They have no network and have no control. Incumbent carriers have designed their solutions in way that keeps their legacy infrastructures relevant.How should the industry change?
This change needs to start with the providers. We're going to see more providers move away from legacy infrastructures. With open source software solutions like FreeSWITCH and Kamailio; cloud infrastructure from providers like Amazon, Microsoft, IBM, and Google; and modern software engineering principles like micro services; it's now possible to build a platform at a fraction of the cost of proprietary big iron solutions built by the incumbents. These savings can and ought to be brought to end users. Not only is this approach cheaper, it's better since it aligns with the Internet and brings telecom back into its rightful place at the forefront of the technology revolution.Does it have to be cloud based?
I think this is one of the reasons that the telecommunications industry remains so fractured, especially when looked at from a global perspective. Up until recently, in order to compete in the telecommunications space, a company would have to be carrier-centric. In the U.S. for example, you would have to partner with CenturyLink in the West, Verizon in the Northeast, and AT&T everywhere else. If you wanted to expand to the U.K., you'd need to set up shop near BT, and so on.
The reason one would have to do this is because building infrastructure is expensive. Telecoms have been managed as national monopolies for more than a century, but the cloud is changing that. By leveraging cloud providers like AWS, you can now offer "local infrastructure" on six continents. As you gain traction in a market, you can scale.
To ensure call quality, you must get the call off the public Internet as soon as possible. Too many hops stand in the way of high fidelity and ultra-reliable communication on the public Internet. The public Internet data packets travel based on the cheapest route (settlement-free peering) instead of most efficient (lowest latency). For example, it's entirely possible that a packet from New York to San Francisco could hit a router in Australia before arriving at its destination. While it might not be noticeable when browsing a webpage, sending an email, or even streaming a video on Netflix, it is obvious in any sort of real-time communication application.