Now is the time to plan how back-end systems are going to migrate.
At the recent CDMA Developers Group Conference held in San Diego, one of the central topics was the migration of 3G networks to 4G, which for this conference anyway, meant both LTE and WiMAX (next-generation WiMAX). Most of the discussions, presentations and panels were about the airlinks and the advantages of OFDM for better data rates, better edge-of-cell data coverage and better capacity, which equates to a lower cost of delivery for the network operators.
Discussions also took place with some of the network operators that are planning to move directly from 2.5G in the form of CDMA2000 1X to LTE, and one network operator that announced it is planning to use LTE in as little as 1.4-MHz segments of spectrum. Engineers have been debating the relative advantages of LTE, EV-DO Rev A and B and UMTS/HSPA and HSPA+ for many months now. Many believe that below 10 MHz of spectrum, there is little if any advantage to deploying LTE in favor of an existing 3G technology. That debate, much like the debate about the merits of GSM and CDMA, could go on forever without either side giving in.
However, the most interesting aspects to come out of this conference for me did not have to do with the airlinks, which we know will be OFDM and IP-centric from the beginning, but rather the discussions around moving the data to and from the cell sites. Part of the issue is, of course, backhaul. Traditional wired backhaul systems cannot handle the amount of data that needs to be transported to and from these sites. The consensus is that fiber and microwave radio systems will be required.
But even more important than backhaul is the transition of the back end of the network from traditional switched and today’s IP systems to fully integrated IP back-end systems, which some refer to as IMS. The most sensible comment I heard from the experts on my panel and others who were speaking was that the time to work on the back end is now, during the deployment of the 3G networks and before the 4G networks are deployed.
The thinking here makes sense to me. If you can build a common IP back-end network system starting with your 3G network, deploy it and prove it out, then adding 4G to the network will be far easier than if you wait until you deploy a 4G technology to rebuild your back-end system. Some of you might say that this is stating the obvious, but sometimes the obvious is not so obvious to those involved in the day-to-day operation and upgrading of networks.
Today most 2G/3G networks are already a mixture of traditional switch networks (for voice, text and slow-speed data), and some form of IP back-end for 3G data services. However, many of today’s 3G IP back ends are not robust enough to add 4G to the same back-end system and this is where network operators can get into trouble. Adding 4G to a network means new radios because the airlink is new and different from either CDMA EV-DO or UMTS/HSPA (which is also based on CDMA).
If you are installing new radios at the sites at the same time you are building a new and different IP back end, the process will take longer and you will probably encounter more bumps along the upgrade road. However, if you spend the next year or so building the back-end network so it is able to handle not only today’s 3G services but also what is coming in the terms of 4G systems, the upgrade will be far less painful, time consuming and expensive.
Seybold heads Andrew Seybold, Inc., which provides consulting, educational and publishing services. For more information, visit www.andrewseybold.com.