Reading an excellent presentation forwarded to me from Darryl Sladden, Product Manager for Cisco Unified Border Element (CUBE) that he gave at the 2008 Cisco Networks presentation, I want to share a couple of key points Darryl made:* SIP Trunking is a key ingredient in the network platform enabling UC * Customer should test, test, test before deployment of their first SIP Trunking solution * Adoption of SIP Trunks should be considered as a TRANSITION and not a REPLACEMENT scenario for TDM trunks
Read what Dave Michels wrote in Innovation: The Dumb PBX about virtual numbers and that many users have multiple numbers. The PBX manufacturers are guilty of being DUMB since they haven't developed their wares, namely PRI with advanced features that could easily adopt virtual numbers in the central office with the feature of TBCT. This remains an issue that the manufacturers continue to ignore. RIM already provides the necessary communications glue to bond desktop to Blackberry while for years the capability has sat under the noses of the PBX factory guys.
Hopefully, SIP trunks will be another segue to doing what PRI development never realized from the factory guys' lack of following and updating ongoing "standards." If the Telcos don't care to promote their features and the factory PBX guys don't care to update them, then who does care? I'd love to have the Blackberry Enterprise Server capability without having to buy a server--and TBCT offers at least a small part for the SMB/E. I understand the large enterprise IT guys wanting to control the wireless devices but how much billing/management information becomes redundant? Then, will SIP trunks develop further into customizable solutions that include the same glue? My bet is definitely yes because it means carrying more traffic. More traffic means more trunks and more trunks means more bandwidth. It's a funny thing, the Telcos failed to promote and educate folks about TBCT, virtual numbers and other advanced PRI features. A last tidbit--don't confuse SIP trunks with hosted PBX services and residential offerings. There's a huge difference between them. Just know that SIP trunkers want traffic.
Thinking about SIP Trunks: A Call To the Wild, Sorell Slaymaker points out another challenge. Conference traffic can literally eat up your bandwidth especially in large IPT deployments with a large "conference call." I invite those interested to think about this--we've always had a box when it comes to conference calling. There's a one to many and many to one relationship so why not have a box capable of managing those bandwidth hogs like conferencing? Even from a SIP provider--why send multiple calls to one location of the same thing? IF 3,000 people want to hear the conference call at HQ then only one connection using any CODEC should be enough to satisfy the need using just one channel. Let the "box" redistribute that one link to the 3,000 people internally instead of having the extra 2,999 outbound calls to the conference bridge. You can define the "box" anyway you want--I just hope you can visualize the point about preserving bandwidth.
Jamie Stapleton of CBSI forwarded an online calculator and don't forget, if you use VoIP encryption the numbers will change. Jamie also hinted at me to set my IP-PBX at all choices of G.711 which I will, but I've got other testing to complete first. As Darryl pointed out above--test, test, test is the best way and that's what we are doing. After reading Darryl's presentation again, I think we are seeing and will continue to see for the next 10 years the development of trunk side IP and it's going to be exciting. More coming on SIP trunks and more to ponder over the Telcos fate or future in delivering viable business wireline solutions. The Telcos have failed miserably to bridge business wireline to wireless communications traffic. They are myopically defined and fail to see that traffic is traffic and converging it matters more. The PBX factory guys are in danger too--will they continue to dumb down their wares or will they get serious enough to start using more intelligence in the network? Think about what I just said, pun intended, but the reality is routers are doing what PBXs once did and PBXs are dumbing down bigtime. It doesn't necessarily mean an end to the PBX or IP-PBX either but you'd better be nice to your router guys.
