What is "voice in the cloud," with the operative word being “cloud?”

Voice and speech communications have existed and evolved for more than a century. From legacy plain old telephone service (POTS), public switched telephone networks (PSTN), and integrated services digital networks (ISDN), to Voice over Internet Protocol (VoIP) networks and speech applications, the trend for speech is to become more network-centric rather than to move towards the edge or client devices, particularly in-the-cloud, so to speak. With the omnipresence of IP networks and data centers, speech has become no different than text and video with respect to communications and networks. “Cloud” itself comes from the architecture diagrams to represent Enterprise PSTN telephony or IP networks. These days, “cloud” refers more to the enterprise IP network than the PSTN network due to the vast prevalence of IP networks for voice, video, data, and control applications and communications. 

With the advent of IP networks, VoIP application protocols such as Session Initiation Protocol (SIP) and H.323 became extremely popular in the past decade, with SIP gaining a lot more popularity due to its simple XML-type data structure (and hence obvious utility in the cloud), and being part of the admired IP multimedia subsytem(IMS) systems in wireless and cable infrastructures. More and more unified communications applications are exploiting the convergence of data and voice applications such as chat, IM, and presence with protocols such as eXtensible Messaging and Presence Protocol(XMPP) to provide a seamless, multimodal voice and data service to the end users. This convergence of voice and data service has given birth to new and improved services for the end users, such as visual voicemail service that could contain a transcribed text with the use of speech recognition engines. Other examples of voice services in the cloud, to name a few, can include access to presence information or location information by using voice inputs, posting voice messages on social media and networking sites. 

With prevalence of cloud-computing, every solution is becoming more of a XaaS, meaning “anything-as-a-service.” Started initially with SaaS, Software as a Service, things such as PaaS (Platform as a Service), and IaaS (Infrastructure as a Service) have become a common practice for hardware, software, and networking infrastructure providers. For instance, call center recording and analytics providers and dictation and transcription providers such as Transcription Service Operators, have all adopted the concept of hosted solutions or cloud-computing. From media industries to others, such as voice-to-text applications and services, it has always been the case that media/content such as voice, video, and data was produced and consumed “as-a-service” irrespective of software, hardware, or infrastructure. 

Voice over data communications has naturally trended the voice applications to become network-centric, with recording, storage, transmission, recognition, search, analytics, and other actions performed in centralized or distributed data servers in a hosted environment. Of course this leads to the more fundamental question as to the security of voice content (i.e. encryption methods for protecting user data and privacy) and methods of provisioning and authenticating the users using OSS/BSS (Operational Support/Business Support) systems in large IP network “clouds.”

Veeru Ramaswamy, Ph.D., is chief technology officer of Vianix. He can be reached at veeru@vianix.com.