When Disruptive Analysis first started writing about “The Future of Voice” about five years ago, the term “media server” only appeared relatively rarely. It mostly arose when talking about traditional telecom networks or enterprise communications, when discussing voicemail, recorded announcements and sometimes transcoding VoIP traffic (usually in the fixed-line world).
Now, the situation is very different indeed. WebRTC, mobile app-based communications, VoLTE and new forms of enterprise collaboration are emerging at a dizzying pace. Increasingly, they involve storage, processing or analysis of voice and video traffic, rather than just onwards transmission.
There are several fundamental drivers here, which are melding together to increase the scope and depth of media servers’ potential roles:
As well as the growth of “media handling”, there are also changes in network architecture that are impacting how and where media servers are deployed. While it is true that an increasing amount of media-processing capability is being embedded directly into endpoints, there are also many points in the network that it is becoming important as well.
In particular, the rise of cloud SaaS and PaaS has shifted the location of media-handling from small instances in corporate IT, to much larger centralised facilities online capable of dealing with multi-tenancy, rapid scalability, and often much faster-evolving use-cases and features. There may also be a desire for geographically-distributed media functions, to optimise for latency and operational efficiency. The addition of diverse connection types – especially mobile – further increases the need for cloud-communications platforms to handle media flows intelligently. This is also driving the shift of media servers towards software-based products and virtualised capabilities.
The main telco equivalents of this are IP Multimedia Subsystem (IMS) and Network Function Virtualisation (NFV). IMS is gaining some traction as LTE networks increase demand for IP-based VoLTE voice and sometimes video services. Together with broadband communications for cable and fibre networks, the industry is seeing an upswing in adoption of various standardised components such as Media Resource Functions (MRFs). This trend parallels another evolution vector in networks – virtualisation and the rapid emergence of NFV. Disruptive Analysis believes that NFV will be deployed earliest for elements and functions that are processing-intense and need rapid scalability and flexible distribution/location around the network.
Going forward, we may see other IMS-based applications beyond “plain telephony”, with some carriers discussing ViLTE, enterprise network-based conferencing and similar initiatives. Some operators also use media servers in RCS deployments, although Disruptive Analysis expects RCS to remain niche and little-used.
Carriers are also widely involved in IP communications services outside of IMS-based contexts. A number have already created standalone WebRTC cloud platforms and applications, aimed at both enterprises and consumers. We can expect to see vertical-focused deployments as well, using the technology in finance, healthcare, education, TV and other sectors.
The growth of “telecom apps” is also relevant here – there are various initiatives by telcos to expose APIs to developers for various realtime-media applications – including WebRTC/IMS hybrids and innovative communications formats which incorporate phone calls and media-processing.
Specific instances may well need richer, larger media-server resources – imagine a cable operator running a remote video version of a Reality-TV audition, via thousands of users’ set-top boxes or smart TVs. In the medical sector, “media” might extend from ordinary voice and video towards new forms of traffic – using microphones to measure breathing, sensors for temperature or vital signs, maybe even MRI or other scans and so on.
An interesting question is going to be the balance of general-purpose vs. specialised media servers, especially as the use-cases expand and fragment. As well as telco-grade products, we already see proprietary media-handling inside various applications and cloud platforms, as well as open-source and other alternatives. Some media platforms specialise in performing particular tasks very well – video-mixing, for example – while others are more flexible.
The likely answer is that we will see all types of media-server develop and grow in importance, although naturally with an ebb-and-flow in each sector and slice of the value chain. Virtualisation and scalability will be important considerations for the more generic servers, enabling them to both be adapted easily by developers, and slot into new frameworks such as telcos’ NFV architectures.
The next five years are going to bring further opportunities – we will likely see much more innovation around WebRTC in telcos, web/app domains, cloud and the enterprise. In addition to that we will see growth of VoLTE and other telco-based communications services, fragmenting from the historic telephony-only world. Perhaps the least-explored opportunities will come from new areas though – Disruptive Analysis has already covered the emerging domain of sentiment analysis in an earlier post. But we are also going to see a huge upswing in voice, video and data streaming from drones, sensors and the broader “Internet of Things” – where the ability to sort, discriminate and manage media may well be much greater, as they move beyond “calls” to much more complex forms of image and sound capture and manipulation.
Dean will be discussing these topics during a live Webinar on January 15 entitled “Media Servers - Powering Tomorrow’s Services Today”. Please click here to sign up and for more information. Reserve your spot today.