Telecom's Global Warming

Over the past year, I've taken to calling mobile backhaul, “telecom's global warming.” Why? Well, the similarities are striking. Where we've been talking about the critical nature of each topic for years, only in the recent past (18 months or so), has significant hype built behind them. With hype in tow, a wide array of vendor “solutions” has emerged.

Yet, where sudden buzz is often the sign of a passing fad (recall the halcyon days of push to talk over cellular?) there's every reason to believe that the current focus on wireless transport is justified. Blame it on the iPhone. Blame it on netbooks, PC cards and USB dongles. Blame it on whatever you want. Mobile data traffic is in the rise across the globe. In the radio access layer, new base stations and new technologies (WiMAX, HSPA+, LTE) promise to help operators keep up with (and profit from) user demand for mobile broadband. RAN upgrades, however, assume the need for transport layer upgrades or you end up with the equivalent of an 802.11n access point connected to dial-up; the backhaul becomes a bottleneck to anything even closely resembling a satisfactory service.

Ultimately, however, while the idea that added capacity in the RAN requires added capacity in the transport layer is easy to grasp, looking at backhaul as nothing more than an issue of plumbing (upgrading the network with fatter pipes) ignores a number of key facts.

• TDM vs. IP. If the evolution from 2G to 3G to 4G is often described as a movement from voice to data to mobile broadband, it’s just as much about the move from circuit to packet to all-IP connectivity. Obviously, this has an impact on backhaul. Sure, mobile operators could run packet data from the radio access network (RAN) over costly T1/E1 connections, but this runs quickly into scalability and performance issues. And, did I mention it’s expensive? This leaves operators with various options: overlay existing circuit backhaul networks with a Carrier Ethernet (IP) transport layer for data traffic and base stations with IP interfaces; run packet and circuit traffic all over a Carrier Ethernet network thanks to interworking or circuit emulation services; run packet and circuit traffic over a Carrier Ethernet network thanks to native IP support in the RAN. In the long-run, all operators obviously hope to implement the last option. After all, one common backhaul network with native IP support is just simpler. In the long-run, however, all operators hope to replace their 2G and 3G networks with 4G – and we all know that’s not going to happen anytime soon. Given the diversity of cellco business models and technology upgrade plans, then, it’s fair to assume that all of these options will live together for a long time.