TMCnet
ITEXPO begins in:   New Coverage :  Asterisk  |  Fax Software  |  SIP Phones  |  Small Cells
 
April 2007
Volume 10 / Number 4
Feature Articles
 

Deploying Peered IP PBXs:

VoIP Strategies for Businesses with Multiple Locations

By Tristan Degenhardt, Feature Articles
 

Businesses are constantly looking to new technologies to reduce costs. It’s expected of VARs and system integrators that any rollout of new technology should both save money and provide better services. VoIP technology is no exception to this rule, and reduced monthly communications bills are a big lure for businesses to adopt VoIP. Businesses with multiple office sites are particularly good candidates for using VoIP technology, by using the ability to connect phone systems at each location together over the Internet. The upside is quite compelling: free calls between sites, and no need for expensive services from the phone company such as frame relay. Instead, by relying on technologies such as SIP, all that’s needed is your existing ISP connection. SIP is an open standard and so IP PBX (News - Alert) systems that are SIP capable should be able to “talk” to each other directly, without intervention by a service provider or external carrier. Peering IP PBXs together allows them to work cooperatively.

Unfortunately, there’s been some reluctance to adopt VoIP by small and medium businesses (SMBs), either due to uncertainty about this new technology, or because of some anecdotal horror story of a failed implementation.

Hype and FUD are the enemies here. VoIP had been hyped to be the answer to everything. Free unlimited perfect phone calls? How can anything live up to that reputation? Simply mention VoIP now, and you’re sure to get groans from just about anyone. It evokes memories of expensive and failed deployments, poor quality calls and general confusion. Despite these horror stories, the benefits from VoIP are real, and by following a few simple rules, it’s actually quite easy to avoid becoming another horror story.




First, go for the low hanging fruit. Position your VoIP strategy and rollout to first address segments that are the easiest to implement, and provide the greatest gain. Where is the low hanging fruit in VoIP? Internal VoIP and peered IP PBXs. A lot more has been made of peering a PBX with a VoIP service provider that terminates calls out to the PSTN. While the VoIP service providers’ marketing departments would like you to believe that using their service is the end goal in VoIP deployments, many VoIP systems are successful without ever subscribing to these services. Internal VoIP involves using IP phones in conjunction with a local network to communicate with a hybrid PBX that may then connect to standard analog or T1 lines. Peering PBX systems together allows sending calls between them over the Internet without using a service provider, allowing locations to communicate for free.

Another rule to follow is to go for the incremental rollout. By taking small steps, businesses can feel more comfortable about the technology before jumping fully into a pure VoIP setup. A first step would be purchasing a hybrid VoIP PBX. Hybrid means that the PBX can link with standard analog or T1 lines, as well as VoIP devices. This PBX could connect to a business’ existing PSTN services without having to sign any new service contracts. After two sites were using VoIP PBX systems, they could then take the next step of peering them together. This step is the easiest, because it does not require additional equipment, and usually does not replace any existing peering setup. Only as a final step do companies need to explore commercial VoIP service providers, whose usage can be on a limited basis, and only where it is cost effective.

The final consideration when peering two locations together with VoIP is to ensure that the equipment is installed on a network that is ready for it.

Most VoIP failures can result from the network they were deployed on, not necessarily the VoIP equipment used. However, it’s easy to succumb to the marketing machine of network hardware vendors and believe that purchasing expensive network gear can solve this problem, but it’s not as simple (or expensive) as that. When peering two VoIP systems together, the best insurance one can take against network issues is to limit the number of network hops between the two sites— fewer hops, fewer points where problems can occur.

The best solution would be a pointto- point data T1 between the two sites; however, the cost of this can negate much of your potential savings from using VoIP in the first place. A more cost-effective approach is to ensure that both sites use ISP services from the same carrier, preferably one that has a national backbone. By staying on a single ISP’s network, you can ensure a minimum number of hops, and can avoid potential bottlenecks at peering points between different bandwidth providers.

Another consideration is ensuring that enough bandwidth is available to handle the desired number of simultaneous calls. A good rule to follow is that an uncompressed SIP call takes 90 Kbps (So your client needs to handle 10 calls inbound and outbound calls to and from the peer at peak times? They’ll need 900 Kbps of bandwidth available to handle just their phone calls). Compressed codecs such as G.729 can be used to increase the number of calls you’re able to fit over your Internet connection. This usually isn’t necessary except when dealing with small remote sites (or home workers) with low-end DSL connections.

What about that expensive network equipment? Most of it can be avoided. It’s tempting for network administrators to purchase expensive managed switches to implement link-level QoS, VLAN tagging, and a host of other largely unnecessary complications. In practice, simple unmanaged switches can service a VoIP network wonderfully, and it’s largely unnecessary to segment your VoIP equipment from other systems in a LAN.

One piece of equipment that does really matter in a VoIP deployment is the router; however, it’s a good idea to resist the temptation to purchase a router that’s loaded with features that you may not use. Many router manufacturers value features over stability, and often things like router-level virus filtering can inadvertently block valid VoIP traffic. The most egregious feature-laden routers even prevent you from disabling some functionality, which in the worst cases will put you on the waiting list for a firmware update to fix a feature that you don’t even use. It’s also not necessary to buy a router with built-in VoIP capabilities. Most SIP-aware routers can actually cause huge problems with modern VoIP equipment, which has outgrown the need for any router-level assistance.

One important feature that you will want to make sure is supported by your router, is QoS. Router-level QoS can ensure that voice traffic has priority over large downloads that may saturate your Internet connection. However, QoS is becoming a common feature these days, even in some very inexpensive home routers. All in all, the best rule when purchasing network equipment for VoIP is to keep it simple.

If it sounds like sacrilege to suggest that top-of-the-line network equipment is unnecessary, keep in mind that in most cases fiddling with all of those complicated settings is akin to being prepared to swat at flies when you’ve got an elephant in the room to take care of first. In a thousand locations, I’ve never encountered a problem with a VoIP deployment that needed to be solved with a fancier router, but I’ve seen plenty solved by actually removing unnecessary network equipment that was trying a little too hard to do the “right” thing with VoIP packets.

With all of the rules for a successful VoIP strategy in place, what sort of accomplishments are possible? Consider these examples of what can be done with peered VoIP systems:

Multiple Location/ Branch Office Least Cost Routing: If a call is placed to Los Angeles from a phone connected to the PBX in New York, the PBX in New York determines that the call would be better (more inexpensively) handled by the PBX it is peered with in Los Angeles. The New York PBX then routes the call using SIP (this leg of the journey is free) to the PBX in LA and from there, the LA PBX routes the call to the intended “local” LA phone number. Think of how many phone calls made by businesses with multiple locations fall into this type of category — for most, the answer is probably a fairly significant number because business contacts are usually clustered around the companies’ locations. Deploying a solution like this could have a huge impact on long distance bills.

Multiple Location Interoffice Calls: Peered PBX systems can dial between each peer seamlessly and for free (because the call doesn’t need to be terminated to the PSTN, there’s no carrier or service provider to pay). To call an extension on the Tokyo PBX from an extension on the Orlando PBX, a user would simply dial as if the Tokyo extension were in their own office. This scenario is the one that most customers are looking for when they’re moving to an IP PBX solution. Office-to-office calls can be an enormous expense and one that simply disappears with a PBX that supports peering.

Multiple Location Intelligent Routing: Similar to the least cost routing example, if a call comes in to your sales line in LA from a caller in NY, the LA PBX can hand the call off to the sales line on the PBX in NY, without having to prompt the caller with a menu asking which region they’re calling from or any other kind of manual intervention. If the peered IP PBX systems support this type of routing, there is a subtle, but very expensive-looking benefit in the customer service arena.

These examples show that there are several ways that a business can benefit from VoIP without rolling out a 100% VoIP solution. In each of the above scenarios, inbound and outbound calls to each location could be handled by POTS lines. This style of deployment allows for incremental shifts when the decision makers are ready and comfortable with taking the next step to using an ITSP for outbound calls, and the step after that, using an ITSP for inbound calls that will complete their transition to entirely VoIP.

Peering provides an excellent way to introduce VoIP to a client, with little risk that they will experience poor call quality, so long as adequate steps are taken to ensure that their network is ready to handle the additional traffic.

With a properly prepared network, good rollout plan, and some basic equipment, a VoIP deployment involving multiple locations can be incredibly successful in every measure. Clients should feel they are getting more for less, and instead of sharing a horror story with everyone they come in contact with, they can start spreading their good news: that rolling out VoIP across their locations has made a positive impact on their business.

Tristan Degenhardt is a co-founder and VP of Operations for SwitchVox. For more information, visit the company online at http://www.switchvox.com.

If you are interested in purchasing reprints of this article (in either print or PDF format), please visit Reprint Management Services online at http://www.reprintbuyer.com or contact a representative via e-mail at [email protected] or by phone at 800-290-5460.

 




Today @ TMC
Upcoming Events
ITEXPO West 2012
October 2- 5, 2012
The Austin Convention Center
Austin, Texas
MSPWorld
The World's Premier Managed Services and Cloud Computing Event
Click for Dates and Locations
Mobility Tech Conference & Expo
October 3- 5, 2012
The Austin Convention Center
Austin, Texas
Cloud Communications Summit
October 3- 5, 2012
The Austin Convention Center
Austin, Texas