For many businesses and consumers over much of the world, broadband services are becoming a necessity — an indispensable feature of our daily lives that extends well beyond the simple connectivity that legacy networks have provided. In light of this ever-increasing demand for data, the need for a mobile solution to bridge the connectivity gap between home and office has become an important market driver.
Over the years the wireless industry has sought to bridge this gap by improving the capabilities of mobile networks. The first cellular networks were designed to carry voice calls only using analogue, circuit-switched technology. With second-generation (2G) networks based on CDMA (code-division multiple access) and GSM (Global System for Mobile), the radio resources became dynamic, allowing sharing among several simultaneous calls.
The wireless experience
Packet-based data services were introduced with the so-called 2.5G networks, which included GPRS (General Packet Radio Service) and CDMA 1X networks capable of delivering peak speeds up to 100 kbit/s. As mobile networks moved into their third generation (3G), data services were enhanced with higher throughput and lower latency, while the range of technology options proliferated, with EvDO Rev A/B (Evolution Data-Optimized Revisions A/B), WCDMA/HSPA (wideband CDMA/high-speed packet access) and WiMAX all in the picture.
Now a new generation of mobile technology, Long Term Evolution (LTE), is arriving. For the first time in a mobile environment, LTE will offer broadband capabilities that rival those of the wireline access options such as DSL.
Fuelled by the success of 3G-enabled laptops, ultra-mobile PCs and must-have smartphone handsets like the Apple iPhone, the consumption of mobile data traffic has soared over the last couple of years. With the proliferation of these new wireless devices, and helped by flat-fee data tariffs, daily data traffic over 3G networks multiplied between three and six times in many mobile networks in 2008.
In addition to current services, end-users are consuming increasingly sophisticated multimedia services, including music sharing and downloads, peer-to-peer video and mobile Web 2.0 applications – in some cases with astonishing growth. Consider these statistics: 1 billion iTunes mobile apps were downloaded in nine months; there are 30 million mobile Facebook users who are, on average, 50% more active than web-only users; 30 Twitter mobile clients are already available. The trend is absolutely clear.
With the growth of wireless multimedia content and service delivery, new competitors are appearing in the wireless value chain. In particular, web players like Apple and Google are “mobilizing” popular in-house applications such as iTunes, Google Apps or Gmail, and encouraging third-party content and applications via open software development kits. Looking to the future, the growth in data traffic will be compounded by the increasing number and types of wireless devices, as well as machine-to-machine communication, for example, to monitor energy consumption in the home.
But there’s a catch: mobile operators are expected to deliver more for less. Traditional user-paid revenues are flattening, in part due to the flat-fee data tariffs that encouraged consumers to sign up for data services in the first place. In response, service providers are looking for ways to augment traditional subscription-based business models with sponsorship-based and converged applications (across multiple access media and devices) that can capture both user-paid revenues and non user- paid revenues such as advertising.
To address these changing market conditions, mobile operators need a new business model that can generate additional revenue from content providers, application providers and other new sources. They are also looking for a next-generation wireless broadband network that can support sky-rocketing demand for bandwidth and a high quality of experience, while reducing their total cost of ownership.
A whole new ecosystem
Depending on the operator’s assets (installed network, spectrum, etc) and strategy, there are several possible evolutionary paths to next-generation mobile broadband. To further complicate matters, a sizable majority of mobile operators will want to maintain services on their legacy networks, while simultaneously introducing next-generation wireless technology and combining this with an evolution to a packet-based core network.
Happily, there seems to be some level of agreement in the wireless world about what happens next. LTE, which is defined by the 3G Partnership Project in 3GPP Release 8, is becoming the common, standardized architecture for the bulk of the mobile broadband community. It has already been chosen by many Tier-1 operators — including Verizon Wireless in the US, Japan’s NTT DoCoMo and TeliaSonera in Scandinavia — as their mobile evolution strategy, and is actively being considered by many more (see box “LTE: Who’s Doing What and Where?”).
With theoretical peak download speeds of 173 Mbit/s, LTE is considerably faster than earlier generations of wireless broadband. But bandwidth is only one aspect: there are other compelling business and technical reasons for operators to adopt LTE. In particular, LTE promises to:
- boost network performance by offering large bandwidth, thanks to the highest spectral efficiency and better spectrum availability, and low latency with a roundtrip delay of just 10 ms in the radio access network (RAN).
- drive down the cost per megabit: both as a result of technical attributes, such as a flat network architecture, which requires fewer elements, and economies of scale that are anticipated because most Tier-1 operators have committed to LTE.
- simplify the delivery and monetization of IP-based multimedia (IMS) services and applications, thanks to a purposebuilt and optimized all-IP architecture.
- allow advanced network and service agility that can continuously integrate, blend and monetize new premium content and services, whether developed internally or jointly with the web community.
- drive revenue growth with a network capable of offering truly converged wireline and wireless services.
What technologies will allow LTE to meet these many and varied requirements at the same time? In the RAN, LTE will make more efficient use of radio spectrum, which is a scarce resource. It is designed to be deployed in a variety of bandwidths from 1.4 to 20 MHz. In other words, a mobile operator can reallocate some of its existing spectrum to LTE, while also deploying LTE in the new spectrum resulting from the digital dividend (the spectrum released by switching off analogue TV) or the 2.6 GHz band, which has yet to be assigned in many countries.
On the downlink (from the base station to the terminal), LTE will use orthogonal frequency-division multiple access (OFDMA), a technology well suited to achieving high peak data rates in high-spectrum bandwidths. Single carrier frequency division multiple access (SC-FDMA), which offers improved power efficiency and terminal battery life, will be used on the uplink.
Multiple-input multiple-output (MIMO) wireless techniques, which employ multiple transmit and receive antennas, will also be deployed. This results in numerous data paths effectively operating in parallel and — through appropriate coding/decoding — significant throughput gains.
However, LTE encompasses more than just the RAN. One of the main aims of the 3GPP standards work was to simplify the network by introducing a flat, packet-based architecture. This results in fewer elements, or boxes, in the network and reduces the hierarchy between mobile data elements (figure 1).
The Evolved Packet Core
The LTE network will therefore be all-IP, from mobile handsets and other terminal devices with embedded IP capabilities, over IP-based base stations, called evolved NodeBs, or eNodeBs, across a packetbased aggregation network to the network core (referred to as the Evolved Packet Core, or EPC), and throughout the application domain both IMS and non-IMS.
Unifying all communications around one protocol — IP — will enable significant infrastructure cost efficiencies and allow new business models and services based on transparent but trusted network-owned knowledge about users’ communications habits, preferences and location.
Underpinning all of this is the EPC, a new, high-performance, high-capacity all-IP core network for LTE, which combines what were previously two separate sub-domains — circuit switching for voice and packet switching for data — into a single IP domain with voice calls being transmitted using VOIP. This improves network performance through the separation of control and data planes and the introduction of a flattened IP architecture, while also addressing requirements to provide voice, advanced real-time and media-rich services with enhanced quality of experience.
As LTE cell sites are rolled out, it is expected that bandwidth requirements will range on average from 50 to 100 Mbit/s per site in highly populated areas. With this large increase in the traffic from each cell site comes the need for a highly capable mobile transport solution across the access, aggregation and core portions of the network.
To migrate and scale the backhaul and backbone network, network operators will need an end-to-end architecture — from cell site to core — that supports diverse evolution alternatives and strategies while providing a clear path to all-IP. A range of different platforms and technologies, including WDM systems, packet-optical transport and IP/MPLS, will all have a part to play in the final solution.
In the past the introduction of new services could be accommodated simply by over-provisioning network capacity. But with the higher aggregate capacities and bursty nature of the data services enabled by LTE and high-speed 3G technologies, this becomes an expensive and inefficient solution for mobile operators, who instead will need to ensure that adequate capacity in the mobile transport network can be provisioned quickly, where and when it is needed.
The use of WDM within mobile backhaul networks can help operators to accelerate time to service, simplify operations and improve overall performance to provide a lower total cost of ownership. In particular the use of new WDM systems that can be provisioned and reconfigured remotely can eliminate the need for frequent on-site interventions. With these systems, in-service capacity upgrades are possible, allowing the network to be reconfigured according to changing traffic demands — a concept we call “Zero-Touch Photonics”. And of course the beauty of WDM is that it can transparently transport any protocol — whether SDH/SONET, ATM, OTN, MPLS or Ethernet — to support multi-generational mobile traffic.
Looking to the future
As operators start to introduce LTE, these systems will often leverage existing 2G/3G cell sites. In cases where LTE is overlaid on either CDMA or GSM/W-CDMA networks, the backhaul network will need to scale to support the cumulative capacity of both technologies. Mobile operators investing in IP backhaul today would therefore be wise to choose a solution that will also support LTE in the future.
Conversely, the co-existence of multi-generational traffic means that the packet-based backhaul network will need to support a combination of TDM, ATM and Ethernet/IP traffic. This can be achieved using a multi-service transport solution like the Alcatel-Lucent Mobile Evolution Transport Architecture (META), which incorporates native TDM interfaces and uses MPLS-based pseudowire technology to transport legacy services over packet.
For clock recovery within the high-speed packet network, enhanced clock recovery mechanisms such as IEEE 1588v2 (also known as Precision Time Protocol) or clock transport mechanisms like Synchronous Ethernet can be used. 1588v2 will be particularly useful for wholesale providers of cell site backhaul, because it requires implementation only at the end nodes — and not in intermediate metro nodes that might be part of third-party leased services.
To meet the stringent quality-of-service requirements of real-time traffic, the IP backhaul network must integrate many of the qualities and attributes of switched networks: predictability, reliability and manageability. In our view, this can best be achieved by implementing MPLS-TP (MPLS-Transport Profile), which offers the engineering and management capabilities necessary to support all mobile services as well as business applications and consumer internet services over an IP/Ethernet network.
The increased bandwidth demands of LTE will also lead operators to evaluate new backhaul alternatives. With the traditional leased-line model ceasing to be a viable long-term solution (in other words, it’s too expensive), operators will look to build their own fibre, microwave and copper-based connections. Fibre-optic cable has clear advantages because it provides virtually unlimited capacity, low maintenance and has a long life expectancy. However, microwave can also provide high capacity, and because it is wireless it is less expensive to install. Therefore, we expect that microwave, fibre and copper will continue to coexist in LTE networks, although an increase in the adoption of fibre is likely.
An orderly transition
The increased speeds that LTE promises will change how mobile services are consumed, and are likely to encourage the development of new services that can take advantage of these speeds. The availability of mobile devices with enhanced features such as larger screens, higher resolution, faster speeds and longer lasting batteries will help to drive service demand by creating a more enjoyable end-user experience.
To make all of this possible will require an order of magnitude jump in bandwidth in the mobile transport network. This will drive the need to evolve from legacy circuit-switched transport networks based on SDH/SONET to a high-capacity, cost-effective, packet-based carrier-class transport mechanism. Europe’s mobile operators and their equipment suppliers are keenly aware that a well-thought-out optical transport strategy is not just desirable, it is one of the cornerstones of the next generation of mobile broadband.