Despite the hype over the past few years around the term "software defined networks" (SDNs), customers are just starting to understand its benefits. I've talked to people that recently attended the Gartner Data Center conference, and the feedback is consistent with what I've witnessed--the majority of organizations are still in the learning phase when it comes to SDN. However, what seems to be well understood now is that networks are complicated and could use better automation through centralized data abstraction. The same thing can be said for unified communications.
What the industry needs is some concept of "software defined collaboration" or SDC, to abstract the areas of complexity up into the software layer, ultimately making the environment more manageable.
Here are the primary areas of complexity that could be solved through an "SDC" model.
* Architecture and network. A SDC would enable a shift to integrated network management. Moving to a centralized IP architecture, versus the traditional distributed node-by-node implementation creates an environment that is significantly easier to manage. The deployment would look like other software applications (not surprisingly) where there is a centralized data center that distributes the voice over the corporate data network, augmented with regional locations on shared clusters or servers.
In this centralized deployment model, session management can be a service that is done at the aggregation layer. Additionally, the communications network will realize greater resiliency, including local break-out/break-in when the network is overloaded or fails.
* Devices. In this case, SDCs would provide holistic, integrated device management.
The need for this is fairly obvious. The number of devices per user is expected to grow somewhere between 400% and 700% over the next five years, depending on whose research you use. The exact number doesn't really matter. We have a few devices now and we'll have a whole bunch more over the next few years.
As the diversity of devices continues to widen, the specific device type needs to be linked to the services that can be supported. There's also the obvious bring your own device (BYOD) challenge of how IT managers can not only on-board a device (which MDM solves) but also then associate it with the user and the specific UC services (through SDC).
* Services. Services are typically managed individually today. Network managers need to be concerned with provisioning richer voice services as well as the many other UC services such as messaging, collaboration, video, mobility, presence and social media. A SDC would enable better integrated service management for faster provisioning and change management time, as well as multi-vendor interoperability. The SDC can enable cloud UC providers to offer broker services and new applications, content types, and corporate content distribution.
• Deployment models. The static nature of legacy communications makes managing mixed environments from a single pane of glass difficult, if not impossible. SDC could integrate the management of the migration process. This would enable a company to choose the best solution when evaluating vendors and deployment models. The benefit to this would be a single pane of glass for the following scenarios:
--Hybrid on-premise and cloud
--Mixed vendor environments
--Mixed TDM and IP environments
* Dial plans and central trunking. These are both areas of increased complexity in an IP world. Managing dial plans becomes exponentially more difficult when multiple vendors are used. Dial plans were simple to manage when the PBX was local to every site and calls went out to the local PSTN. Now organizations are faced with having multiple dial plans to accommodate multi-vendor environments. SDC would abstract this functionality up a layer and enable a mix of on-net and off-net calls.
With regards to SIP trunking, this is becoming an increasingly important part of IP deployments and requires an integrated dial plan across multi-layer architectures.
Both of these above points also alleviate the complexity with multi-country or regional dial plan requirements. Dial plans may seem like a mundane issue but they can significantly increase the complexity of IP environments.
* Back office integration. This is analogous to the "programmability" that's been promised by the SDN community. A SDC solution would drive advanced APIs and advanced workflow management to provide a single point of integration for the following:
--LDAP/AD (Active Directory) integration. Automatic on-boarding and off-boarding of users
--Accounting and business analytics and automated accounting of the suite of UC services
--Monitoring and performance of the UC infrastructure as well as syncing between assurance and fulfillment
--Mobile device management support, enabling the integration of device policy with service management
--Cloud service broker support to enable third-party services and packaged applications for the enterprise
--CEBP: UC integration with CRM, ERP and other enterprise applications
* Consumerization of IT. A successful BYOD deployment should be front-ended with a powerful but useful end-user portal. The SDC solution can provide the portal with real-time visibility and access to applications that will help them be more productive. Companies need to enable workers to use the tools they want but maintain the oversight and control of these applications.
While the term "software defined collaboration" certainly isn't mainstream, I have blogged about this fundamental need before but called it "specialty UC middleware". It's the abstraction of the functions from the physical boxes into a software layer. Vendors like Acme Packet and VOSS provide this functionality, as do products like Avaya's ACE and ALU's One Touch. So the concept of SDCs might be new but I do believe the capabilities are there today. The industry just needs to put it into action.
