Caas Architecture

CaaS (Communications as a Service) is beginning to flourish. As companies evaluate the cost, time, and flexibility advantages, CaaS becomes a desirable tool businesses can use to quickly respond to changing markets. A closer look at CaaS reveals that proper CaaS deployments are built upon the same basic principles used by good premise based implementations. The primary determining factors for selecting a CaaS methodology are:

  • Initial cost
  • Recurring cost
  • Functionality
  • Connectivity
  • Security
  • Administration
  • Survivability
  • Business Continuance/Disaster Recovery
  • Investment in Existing Equipment

Based upon these factors, a variety of CaaS solutions have been developed. As you would expect, each has strengths and weaknesses, largely determined by the software, hardware, and architecture of each approach. Currently, there are three primary types of CaaS offerings in the marketplace:

  • Hosted VoIP
  • Hosted TDM
  • Local Control/VoIP

Let’s explore the architecture of each solution. (Download - A New Approach to Communications as a Service (CaaS) Whitepaper)

Hosted VoIP

Hosted VoIP is by far the most common solution. In the Hosted VoIP model, the primary equipment and telco circuits are located at one or more facilities operated by the hosting company. For best call quality, the customer should use a connection that supports Quality of Service (QoS) to prioritize and protect voice traffic over their network connection. Although using a Multiprotocol Label Switching (MPLS) connection or other connection with an appropriate committed rate of throughput is recommended, there are specific situations where quality is not as critical. In those cases, establishing a VPN connection over a best-efforts connection (such as DSL or cable Internet) may be sufficient. Some configurations support almost no equipment at the customer premises, perhaps as little as a router with VPN connectivity and some SIP phones. More complete solutions usually include some local connectivity for local 911 service, a SIP proxy server, and onsite connections to the customer’s other applications.

Hosted TDM

For customers who are not ready to make the investment in VoIP but need more advanced communications abilities, a hosted TDM configuration could be a good fit. Often this is also a good choice when disruption to the business could be more expensive than the cost of the equipment. Other reasons to adopt hosted TDM include the need for quick time-to-market, lack of personnel with sufficient network or VoIP expertise, or a move to a new facility is in the near future or when investing in the current facility’s infrastructure doesn’t make sense.

Like the hosted VoIP model, interactions come into the hosted facility. However, the interaction is not routed to an agent via a network connection. The hosted system connects the incoming call to the agent via a bridged call to the agent. Typically, the agent works behind a traditional PBX system using DIDs. Usually this results in little or no extra hardware at the customer site.

There are some trade-offs for avoiding the expenditures of upgrading infrastructure. The first is incurring a higher monthly cost because of the charges for the bridge calls between your office and the hosting location. In addition, setup and administration costs will be higher because moves, adds, and changes must be made and coordinated between two phone systems. At times, the additional complexity may lead to difficulty when trying to locate a communications problem.

Local Control/VoIP

A third option is the local control/VoIP model. This choice gives a customer most of the advanced benefits of having a on-premise VoIP phone system even though the primary server is located off-site in a hardened data center.

This is especially helpful for customers who have or desire to create a robust local network supporting VoIP but need to add advanced functionality that their current system doesn’t support. Instead of replacing an entire phone system or creating a patchwork of existing features, the customer can connect to a more advanced system to provide greater functionality. While calls come into equipment located on the customer premises, call control is handled by the advanced phone system located in the hardened data center.

For customers with exceptionally strong compliance or security requirements or who rely upon high bandwidth applications, the combination of features, functionality, and flexibility are hard to beat. Since phones, network, gateways, proxy/media servers remain at the customer’s facility, they retain full access and control of those devices.

The customer’s telco lines also remain at the customer’s location. This works especially well when a customer has a preferred voice carrier, has existing contracts that must be fulfilled, or possesses many existing numbers they would prefer not to port or forward. They also avoid the costs of forwarding or bridging calls.

Should the WAN connection to the datacenter be interrupted, calls can still be handled by the on premise equipment, although with less functionality.

Local Control/VoIP (Extra Redundancy)

An additional variation of the local control/VoIP model would split the telco lines, placing some at the customer’s site and some at the hosted provider to provide seamless failover if local trunks ever go down. Likewise, a second, lower priority method of connectivity to the Data Center could be established providing even higher redundancy for data communications.

CaaS Featured Articles