Cellular Looks to Converged Access Networks
Driving this change is a greater focus on operational efficiency and the expectation that new 3G networks will generate a surge in data traffic, finds the latest Unstrung Insider research report: Wireless Backhaul & 3G Network Economics.
The critical concern is that the prevailing Time Division Multiplexing (TDM)-based leased-line backhaul architectures, deployed to support 2G voice networks, and the overlay Asynchronous Transfer Mode (ATM) networks deployed for 3G, will not scale elegantly and cost-effectively enough to support future price/performance requirements. Thus operators are looking for ways to migrate existing networks onto converged 2G/3G packet-based transmission networks.
The aim is to reduce operational expenditures (opex) by combining two separate networks; to make more efficient use of bandwidth through the use of statistical multiplexing, grooming, compression, and quality of service (QOS); and to decouple the cost of running the network from incremental increases in the number of users or data traffic growth.
This last point is critical, since the profitability of data services will suffer if the cost of transmission increases linearly with network usage. Instead, the requirement is for backhaul networks in which the incremental cost of adding new users and more traffic is as close to zero as possible.
As they evaluate their backhaul options, operators are weighing "lease vs. build" approaches to access and metro networks (many already have wholly owned core networks), with a focus on how quickly capital investment (capex) in self-built transmission can generate a return on investment, compared to the ongoing opex costs associated with leased lines.
How the lease vs. build debate will play out, and how best to design and deploy the next-generation of mobile backhaul networks, is still far from clear. At issue are the wide varieties of local market pricing and regulation; the availability and pricing of leased lines and dark fiber; the uncertainty over how much demand for 3G data there will actually be; the requirement to support multiple generations of radio access technology, each with distinct base station interfaces; and the extent to which convergence of wireline and wireless operations will influence network design.
Despite these uncertainties, some key trends have emerged as carriers begin the migration to packet-switched backhaul networks. These include:
Access is Still King
Deploying and operating physical links with adequate throughput and QOS (especially for synchronisation) to the cell site remains the major challenge for mobile backhaul. The Insider report suggests an 80:20 rule can be applied to cell site capacity requirements, with 80 percent of sites likely to require 16xE1 equivalent (32 Mbit/s) or less in the coming five-year period. Given the wide geographic distribution of base station sites, this makes last-mile cost optimization crucial.
The three options are: copper (digital subscriber line (DSL), T1/E1, or possibly Ethernet); radio (microwave, satellite, 6GHz fixed wireless systems); and Fibre (Ethernet, ATM, or TDM), with each having distinct advantages and disadvantages.
Microwave is Booming
Operators worldwide are replacing, wherever economically feasible, the ongoing opex of leased-line backhaul with capex investment in self-built microwave. One example is Vodafone Group plc (NYSE: VOD), which expects to save 15 percent to 20 percent, over two years, of its £280 ($515) million annual opex on access transmission across Europe though greater adoption of microwave systems.
Microwave vendors aren't complaining; Despite unit price declines of 15 percent to 20 percent per year, many report annual sales growth of 20 percent to 30 percent in recent years, with no sign of a slowdown. Ericsson AB (Nasdaq: ERIC) stormed the market with its Mini-Link Traffic Node products, but Alcatel (NYSE: ALA; Paris: CGEP:PA), in partnership with Stratex Networks Inc. (Nasdaq: STXN)), Siemens AG (NYSE: SI; Frankfurt: SIE), and others, is fighting back with microwave aggregation products of its own.
The most popular products are still short-haul Plesiochronous Digital Hierarchy (PDH) microwave links capable of up to 16xE1, but some operators, especially in Europe, are starting to deploy higher-capacity Synchronous Digital Hierarchy (SDH) microwave, with vendors providing software keys for pay-as-you-grow capacity upgrades from 8xE1 to 32xE1 and beyond. Technological initiatives include the addition of Ethernet ports and introduction of adaptive modulation schemes.
Cell Site and Remote Aggregation Nodes Emerge
One implication of the move to packet-based cellular backhaul that combines 2G and 3G voice and data traffic, is that radio base stations deployed today typically support either TDM, ATM, or in some cases Ethernet interfaces. This creates a requirement for a small device at either the cell site or at a remote aggregation node that supports all three interfaces and can be used to combine traffic types for onward transport.
The advantages of such products are that 2G and 3G traffic can be sent over a single northbound link into the network; that any type of interface can be used; and that bandwidth optimization techniques can be applied. Vendors providing these types of products include Celtro Inc. , Carrier Access Corp. (Nasdaq: CACS), Cisco Systems Inc. (Nasdaq: CSCO), Eastern Research Inc. , NMS Communications Corp. , and RAD Data Communications Ltd. .
Psuedowires Find a Role
With a huge installed base of TDM and ATM equipment, pseudowires are seen as a way to enable a migration to a single optical Ethernet transport network. Almost all vendors have some kind of pseudowire story, yet Axerra Networks Inc. and RAD, which offer this capability in the access link, and Mangrove Systems Inc. , which offers a metro-area solution, stand out for making pseudowires a key feature of their market positioning.
Competing Approaches to the Aggregation and Metro Buildout
There is an economic incentive to build points of presence (POP) as near as possible to the cell sites (to save on leased-line and microwave costs), and operators are considering just how far out they should build their metro and aggregation networks, and in the case of integrated operators, how wireline and wireless backhaul can be converged.
Having multiple traffic types – TDM, ATM, IP – converge at the aggregation POPs presents operators with an opportunity to streamline the network from this point onward into the metro and core switching and data centres. The challenge is how best to do this. Due to the need to support an extensive installed base of TDM and ATM equipment, much of which is still at high book value on operator balance sheets, the focus is squarely on various types of multiservice transport equipment. In a nutshell, this means making critical decisions between architectures based around Multiservice Provisioning Platforms (MSPPs) for next-gen ynchronous Optical Networking (Sonet)/SDH networks, ATM-based Multiservice Switches (MSS), and IP/Ethernet-orientated Multiservice Switch/Routers.
Given that cellular backhaul is just one possible application for wireline transport products, almost all major networking vendors have some relevant positioning here, including Alcatel, Cisco, ECI Telecom Ltd. , Ericsson, Fujitsu Network Communications Inc. , Juniper Networks Inc. (NYSE: JNPR), Lucent Technologies Inc. (NYSE: LU), Nortel Networks Ltd. , Siemens, and Tellabs Inc. (Nasdaq: TLAB; Frankfurt: BTLA).
— Gabriel Brown, Chief Analyst, Unstrung Insider
The report, Wireless Backhaul & 3G Network Economics, is available as part of an annual subscription (12 monthly issues) to Unstrung Insider, priced at $1,350. Individual reports are available for $900. To subscribe, please visit: www.unstrung.com/insider.