Mobile TV, Internet access, uploading and downloading videos to services such as YouTube, streaming music and other exciting stuff like this is set to make the mobile phone even more fun. But will the backhaul connections between base stations and the core networks be able to cope asks Peter Purton

How times change. It used to be that cellular networks were more than capable of handling the demands of even the most sophisticated handset. Today it is no longer the handsets but the networks that are becoming the bottleneck in providing new and exciting mobile services, particularly the base station core network connection, the backhaul.

In developing markets simply coping with the growth of subscribers is the biggest problem in cellular infrastructure, just as it was in North America, Europe and Japan, says Kai Sahala responsible for sales at Nokia Siemens Networks’ radio access networks division. “But now the biggest challenge in the developed markets is the growing usage of data. Operators report tens of per cent of growth each month,” he says.

Five years ago the typical connection between a base station and the core network was a single 2Mbit/s E1 line or maybe 1.5Mbit/s T1 line, maybe two for a high usage cell, says Gaby Janowicz, business development director at access equipment manufacturer RAD. It was a GSM world and all people did was make and take calls and send or receive the odd text message. “This changed dramatically with the inception of 3G,” says Janowicz.

3G started users down the road of ever bandwidth hungrier applications. And that means a stronger backhaul link if users are not to be left disappointed and complaining about service quality.

Exactly how much backhaul capacity is needed depends on the mobile network operator’s own business strategy. A mobile network operator pushing mobile TV, facebook or YouTube usage over mobile phones will need to provide more capacity than one pushing low cost calling and texting. But as mobile network operators follow their fixed network cousins down an increasingly IP service laden path, demands on the backhaul are expected to grow significantly.

Today we already need the equivalent of four times as many E1/T1s to support a base station as five years ago. In five years when 4G technologies such as LTE come on board, Janowicz estimates, backhauls will have to be able to cope with seven times higher traffic than today. Already a single HSDPA user could use the same capacity as over 1,000 GSM subscribers!

This represents a gigantic jump in only a few years and the ramifications on the backhaul are dramatic. Not so much from a technology point of view – everything is achievable at a price – but from an economic one.

Backhaul already accounts for around 75 per cent of mobile transport costs and up to 20 to 40 per cent of total operating expenses. Any significant rises in cost are untenable.

Many mobile network operators are already implementing optimisation and aggregation strategies, for example deploying cell-site gateways at co-located 2G and 3G towers that free up E1/T1 bandwidth and hand-off multiple E1/T1s to a larger pipe, in which the economics of E3 or even STM-1 make sense.  But this, along with statistical multiplexing of 3G data, only produces savings of around 50 per cent. If HSPA is introduced on the scale which has been promoted, this is barely a beginning.

According to analysts ABI Research, cellular backhaul, which today accounts for a CAPEX just in excess of $14 billion, is expected to reach $23 billion by 2012.

“The persistently high cost of backhaul – in both capital investment and operating expenditure – is a top-of-mind issue for mobile operators,” says senior analyst Nadine Manjaro.

For many European operators, the key to the solution is self-evident: Ethernet. Ethernet delivers about five times the bandwidth of E1/T1 lines for the same amount of money, is extremely flexible and is easier and less expensive to maintain and manage than legacy networks. Luckily for European mobile operators, there is ample low-cost Ethernet available over wholesale DSL and Metro Ethernet networks.

In the CDMA world, the ability to backhaul directly over IP is built in and, therefore, cable lines and DSL can be used for that purpose.  With GSM and UMTS, however, it was decided to use TDM and ATM as interfaces for base stations.  This means that if European mobile operators want to capitalise on the benefits of using inexpensive DSL, they need some means of emulating ATM over Ethernet.

According to research service Unstrung Insider, the first Ethernet-capable base stations with support for native IP RAN protocols appeared on the market in 2007, with a small number of forward-looking Asian operators deploying the technology in live, commercial networks during the second half of the year. These are still isolated early adopters, however, and not yet reflective of the industry mainstream.

But taking maximum advantage of IP/Ethernet economics requires not just an upgrade to mobile backhaul networks, but also the emergence of compatible cell-site equipment, notes Gabriel Brown, Unstrung Insider chief analyst.

“Cell sites can be Ethernet-enabled using specialized gateway equipment – a fairly common strategy today," notes Brown. However, in the near to medium term, operators will require base stations that support native IP RAN transport and present an Ethernet port to the network, if they are to drive the maximum benefit from packet-switched transmission, he adds.

In a recent report written by him entitled Ethernet-Capable Mobile Base Stations, key findings include:

-The first Ethernet-capable base stations, with support for both 2G and 3G IP RAN protocols, appeared on the market in 2007; more products from all the major vendors are scheduled for general availability in 2008.

-With traffic growth fastest in the existing 3G footprint (i.e., urban centers), a strong retrofit market for Ethernet transport modules is anticipated as operators upgrade existing base-station equipment.

-Huawei leads the market in terms of commercially available IP RAN capability and has full support for Ethernet across its base-station portfolio; Alcatel-Lucent, Ericsson, Nokia Siemens Networks, and Nortel will all respond strongly in 2008.

Meanwhile, a technology originally aimed at carrying different types of data traffic across an IP backbone has been adapted to support Ethernet based cellular backhaul. Pseudowire, as it is known, was developed by RAD as a technology called TDMoIP (Time Division Multiplexing over IP), which enables TDM type traffic, such as 2G GSM, to run transparently over IP. Since, this technology has been extended to ATM and has formally adopted by all the major standards organizations, including the IETF and the ITU-T, as PWE3 (pseudowire emulation end-to-end).

Pseudowire is the pre-eminent technology solution for transporting legacy protocols such as TDM and ATM over DSL, Ethernet and MPLS, says Eitan Schwartz, vice president, Pseudowire and Ethernet Access at RAD. And it already being used or adopted by the major European Tier 1 operators to enable their HSPA network to run over DSL, he adds.

The only thing stopping carriers from removing their E1s altogether is the issue of Quality of Service (QoS), says Schwartz. And this is being addressed.

“In essence, this amounts to operators asking whether or not they're happy for voice, as well as data, to travel over IP,” he says.

A key difference between employing a packet-based transport network such as Ethernet instead of traditional TDM and ATM is that there is no clock source for synchronisation. Inaccurate clocking across the network can lead to dropped voice calls and interference between adjacent base stations that affect QoS.

Two approaches are being considered among mobile operators. The so-called “Big Bang” approach, would transport all real-time voice and video traffic, along with data, over a packet-switched backbone network. The other, the “Step-by-Step” or “Hybrid” approach, would continue to let time-sensitive voice and video to be carried over E1s while less time-sensitive statistical HSPA data would be handed off onto DSL.

Most major operators in Europe are currently mulling over this decision and will most likely come to their conclusions over the next twelve months. As far as Schwartz is concerned this is likely to be the migratory approach.

By migrating to a partial IP system instead of a complete one, the operator achieves major cost savings without jeopardising QoS-sensitive applications, especially the old revenue workhorse – voice, he says.

“The mobile industry is finally grappling with the reality that backhaul is as fundamental to its service as the throughput of the radio network,” says RAD’s Schwartz.

“Buzzwords like WiMAX, WiFi and LTE may grab the headlines, but it is down in the trenches, where data clogs the pipes, that will make or break the mobile broadband industry,” says Schwartz. It is “smashing the backhaul bottleneck” that will open the way more than perhaps anything else to “the world of a genuine broadband mobile lifestyle,” he adds.

PANEL STORY
Mobile Backhaul Trends

The mobile communications industry continues to grow at a significant pace. Industry sources report the following figures:

-2.5 billion mobile phone users in 2006, growing to 3.6 billion by 2010

-2.3 million base stations/backhaul connections worldwide in 2006, growing to 3.3 million by 2009

-Two of every three towers already have more than one mobile operator on them

-Projected backhaul requirements per cell site: 2-16 T1s/E1s in 2006; 2-8 T1s/E1s and 10-30 Mbit/s Ethernet in 2008

-New data and video services drive demand for more bandwidth. In the U.S. alone, mobile operators’ transport costs are expected to skyrocket from $2 billion in 2006 to $16 billion in 2009.

-More technologies need to be supported: 2G/GSM, 2G/CDMA, 3G/UMTS, 3G/EVDO, HSDPA, WiMAX.

A closer look at these statistics reveals several warning signals. From 2006 to 2010, the number of mobile phone users is expected to grow by only 30 per cent, while backhaul expenses will skyrocket due to an exponential increase in bandwidth required for video and multimedia applications as well as the need to support multiple technologies.

Moreover, the Average Revenue Per User (ARPU) is likely to remain fairly flat, even for new data services, with competition from fixed line offerings keeping a lid on pricing. If mobile operators are to reach profitability targets, they must improve the efficiency of their networks by dropping the cost per Mbit/s of bandwidth.

Because backhaul is one of the major contributors to the high costs of building out and running a mobile network – estimated to be about 25-30 per cent of total operating expenses – it is critical that mobile carriers optimize their networks and/or find lower cost and more scalable cell-site backhaul solutions from alternative carriers.

To ensure customer satisfaction, these optimized, scalable backhaul solutions must be in place as new wireless applications and multimedia services, such as music downloads, mobile video and gaming, are rolled out.
Source: RAD