Converged Campus Network Lessons
Convergence is alive and well; it's akin to a constant shuffling and re-evaluating until the right solution is achieved.
Convergence is still alive and well, at least in the education vertical. Specifically, it has taken my company since 2007 to weed out an affordable and viable paging alternative to the traditional paging companies that cost our schools, on average, $56,000 for a paging solution. We now have tested a SIP paging solution and moved it to a live campus for customer evaluation. The average cost to deploy this solution is $18,500. But paging isn't the only service or application we are putting on the network.
Network cameras and network DVR (Digital Video Recording) require bandwidth and power over Ethernet (PoE); and location, location, location is important for each camera. Then, as with cable plant, each camera must be suited for the environment: inside, outside, dry or wet, sunny or dark and whether or not it needs to be vandal-resistant. Next are the types of optics and the specific reason why a network camera is needed and what it is expected to capture.
WiFi remains big, with more and more deployments of Apple iPads vs. textbooks, and as means to cutting costs in K-12. But WiFi doesn't mean just one SSID (Service Set Identifier), but several with varying degrees of access. Then, WiFi is also touching M2M communications for HVAC and building systems controls.
Voice remains the top thing to add to the network, to cut costs by porting numbers over to SIP trunk services. In our recent porting, a school had 6 separate telephone bills--not including the numerous bills for long distance providers; we managed to reduce monthly costs by $200 and that effort eliminated one bill. We still need to extend fiber to other buildings before we can port remaining services.
Next, we did a weekend inspection of the inside cable distribution plant. This cable plant is the original plant installed by a Telco and it isn't without problems. Once we mapped out the distribution points, with butt set in hand, we tested every wire pair to determine what was on each cable pair. We found 4 unused lines billed at $224 monthly and issued disconnects for the customer. This eliminated one more telephone bill.
A couple of years ago we met with factory reps about an IP access/control system. We felt strongly about the product but uneasy about taking on any door work, since that's not our core strength. Recently we met with a local installation firm representing this product. We determined that the fit was good and began addressing campus requirements.
A key cost to door work is labor for doing the work at the door, and then cabling and system components. Even with IP solutions, we still end up needing two-wire connections for door phones. Distances and cost for installing signal wire can add up pretty quickly, and using legacy cable plant can minimize the cable distances and keep installation costs in check. The really cool feature of IP-based access/control systems is that if they integrate with Active Directory, moves/changes become easier.
Next, we documented the old cable distribution plant. Then we determined that we could eliminate two distribution points by installing new cabling between two points. Next, we re-terminated the cables at all intermediate distribution points and tested pairs. Now we have a means to effectively add door phones throughout the campus by installing a cable from each door to the closest distribution point (or Intermediate Distribution Frame).
The campus networks are fiber backbone networks with Cat5E cabling riser 150 feet or less to each endpoint. These newer LANs are not amenable to "pairs" or distributing 2500 dial tone, fax or POTS to different endpoints using the LAN. The old distribution system easily accommodates these needs. While there are IP door phone solutions, they are expensive compared to analog gear.
Several years ago I wrote: Why Would You Give Up Your Structured Wiring? Door phones, postage and fax machines seem to be the remaining devices that we need to use old cable distribution methods to connect dial tone. Installed cabling of any type is an asset. Sometimes the asset just needs a little attention to bring it back to life.
Our schools have Verizon FIOS (primary) and Comcast (secondary). Their SIP trunks route over FIOS and failover to POTS services terminated on FIOS. We decided to abandon the POTS via copper and have Verizon move the backup POTS lines to FIOS instead. Copper is just inherently more trouble, and while you can argue for diverse routes by using copper lines, Verizon and AT&T both are shedding outside plant by selling it off.
We are moving fax numbers over to services like Efax, Ring Central and MyFax. This allows multiple and simultaneous fax calls that route the fax to email and reduces the headaches suffered by callers reaching fax machines that are busy, out of paper, toner or simply need to be rebooted. Outgoing fax traffic is handled by connecting the fax machines to 2500 ports on the IP-PBX and routing out over SIP. I've heard of numerous horror stories but I can say that Panasonic has done an outstanding job on pushing fax calls in and out of their gateway. We simply don't have issues and complaints with the SIP dialer (gateway) and we've deployed in MPLS, FIOS, Comcast and T1 networks.
Convergence is alive and well; it's akin to a constant shuffling and re-evaluating until the right solution is achieved. Is the effort to converge all clean, tidy and all-IP? Not quite, but the difference is you end up with fewer vendors, it is easier to maintain, and it does provide value. Convergence is after all, a process, and good management dictates change that brings about benefits.