SIP Carriers in the Contact Center

A large number of Interactive Intelligence folks were out at the ITEXPO event in LA this week giving presentations and I’m sure soaking in some nice California sun.  In reviewing the presentations and the agenda being delivered I was pleasantly surprised to see the amount of coverage on SIP based carriers.  It appears that SIP based trunking is finally starting to gain some real world traction outside of the early adopters.  This is being driven by the roll out of services from name brand carriers like Global Crossing and AT&T in an aggressive manner, often bundling these in as part of an MPLS build out to provide complete network services to their clients.  Over the past few months we have seen an ever increasing number of requests from contact centers inquiring to see how they can best leverage these new services to 1) save money and 2) provide a DR solution that doesn’t require duplicating all their gateways across multiple locations. While these services have shown some great promise in achieving these goals, it is important for organizations looking to leverage SIP trunking to make sure the carrier selected will meet their needs.

During a call with a client last week I was asked for the ‘capabilities’ and caveats of SIP carriers.  This is kind of like asking a veterinarian to describe the size and temperament of dogs.  In other words there is no one size fits all answer. Just like dogs can range from Chihuahuas to Great Danes, SIP carriers each can come with fantastic benefits (although I still haven’t gotten one to fetch a tennis ball) and also some limitations.  Today I want to focus on a few areas that you want to make sure you evaluate when selecting a carrier.  For some customers, these are show stoppers while for others these capabilities are no big deal at all.  These tend to vary by carrier so it is important to run thru them all with each carrier being evaluated.   For some real world data on carriers Interactive has tested with you can check out our testlab page on SIP trunking. 

1)      Fax –I am always amazed at how often businesses we work with still rely on faxing for everything from orders to HR processes.  As such, this tends to be a hot topic of conversation almost immediately when evaluating SIP carriers. In the VOIP world there are two types of fax, G.711 and T.38.  While both types can work T.38 is far and away the preferred method for VOIP types of deployments because it is much less prone to error.  Depending on the Softswitch chosen as the carriers core T.38 may or may not be supported. 


2)      ANI Restriction – Also known as Caller ID – Many outbound and collection customers need to be able to manipulate the outbound ANI when they are placing calls.  This is particularly true for teleservices firms representing multiple clients.  Many carriers support this and there are even some consumer based ANI spoofing services out there selling this ability as a service.  However, not all SIP carriers support this ability.


3)      Call Throughput – While not something most of the carriers are willing to disclose the system a large client is hooking up to a carrier could very well have more throughput capability than the carriers own core switch.  In the field we refer to it as melting a carrier when a large SIP based dialer pumps out a boatload of calls per second and ‘melts’ the carriers switch on the other end.  While this is never something anyone intends to do a large scale volume test using bulk call generators and the actual connections being proposed for the solution should be a requirement for any deployment.


4)      Transfers – Transfers sound really simple but in fact there are many flavors of a transfer ranging from a deflection to a blind transfer to a consult transfer etc.  The list is actually rather long when you start going thru all the types.  For each type of call transfer there is a corresponding set of messages that a SIP carrier must support to perform the function. 


5)      TCP/UDP – SIP communication can be transmitted using UDP or TCP.  Without getting to geeky if everything is going fine it doesn’t really matter but often times propeller heads like myself will prefer TCP because it can retransmit and recover from errors more quickly that UDP.  In addition there are different ways that many SBC’s (Session Border Controllers) for example behave when evaluating SIP packets.  Depending on the deployment type being chosen this may or may not make a difference but is certainly something that should be checked out.


6)      SRTP/TLS – In today’s increasingly security cautious business arena the ability to encrypt all kinds of data traffic is becoming important.  As such, many Interactive clients have chosen to encrypt their call traffic with SRTP/TLS all the way to the carrier’s core.  This is particularly true for sensitive customer traffic or for executive level communications. 

While this list may look a little intimidating to a non-engineer type none of these considerations are what I would treat as overall limitations to implement SIP trunking?  Why?  Mainly because each of the carriers has some subset or all of the capabilities listed above.  Just like no two families going to the local ASPCA shelter (please support them!) will pick out the same dog, each has a very good chance of finding the pooch right for them.  We are finding the same thing with SIP based carrier services.  With a little homework and leveraging of work already done in regards to integration and capabilities organization should be able to effectively create lower cost and enhanced disaster recovery services thru the use of SIP based trunking. 

-Tom Fisher