T E C H N O L O G Y

W H I T E

P A P E R

Portfolio: White Paper 3

Customer solutions
As wireless makes the transition to data and video-dominated broadband, unprecedented demands are placed on operator networks. The lightRadio product family allows wireless operators to adjust to this new traffic mix, while recognizing the heterogeneity of operators existing assets and service scenarios. This paper links these new products with the capabilities of physical backhaul media and examines the potential for centralized baseband processing. This paper is one in a series authored by Alcatel-Lucent that discuss the current state of wireless networks and the benefits of transitioning to a lightRadio architecture that supports data and video traffic, now and well into the future.

Table of contents
1 1 1 2 3 4 6 7 8 8 13 14 14 1. Introduction 2. Wireless market overview 2.1 The impact of change and risk 2.2 Demand scenarios 2.3 Backhaul asset scenarios 3. lightRadio solution 3.1 Customer case 1: Abundant dark fiber in the first and second mile 3.2 Customer case 2: Abundant dark fiber in the first mile, little in the second 3.3 Customer case 3: Scarce dark fiber in both the first and second mile 3.4 Customer case 4: No owned dark fiber 4. Conclusion 5. Acronyms 6. Author

1. Introduction
Today, networks need to respond effectively to unprecedented change and growth. Based on a holistic view of the key objectives and constraints encountered by wireless operators, the lightRadio product family is designed to encompass all aspects of a wireless network, including wireline backhaul. This paper describes how lightRadio products, in conjunction with different forms of interconnect/ backhaul networks, can provide optimal wireless solutions. It provides a special focus on when and where centralized baseband processing can be advantageous, beyond the obvious scenario where the wireless provider already owns dark fiber between the base station and a centralized processing site. This paper discusses common customer requirements and maps them to optimal solutions in the context of lightRadio.

2. Wireless market overview

Planning for wireless networks has never been more difficult than today. On the demand side, planners must deal with unprecedented growth rates, along with changes in application use that drive dramatically different usage rates per subscriber. In addition, they must address regions with diverse subscriber density and behavior, as well as dynamic shifts in user location, activities and conditions, such as serving stadiums during sport events and providing communications for public emergencies. New devices, such as the iPhone and iPad, are not only driving changes in network technology but also the place, time, and intensity of network usage. On the business side, planners face changes in network geographical coverage, access to and leasing of spectrum, as well as competitive and regulatory changes. On the supply side, technology issues now include metro-cell infill and heterogeneous operation/networking, LTE-Advanced, standards revisions and technical innovation. These changes are occurring within diverse scenarios with respect to interconnections, siting, spectrum availability and Radio Access Network (RAN) or tower sharing.

Service providerS face many uncertaintieS and riSk factorS

Where will I have new spectrum rights? What will my mix of large and small cells look like and how will they work together? How will my backhaul network be structured to best optimise the way that the wireless and wireline assets work together? How many subscribers will I have in which locations? What devices will those subscibers use, with which radio technologies? What services and applications will they use, with what intensity of use?

2.1 The impact of change and risk Amid these changing conditions, wireless operators are facing a growing divergence in requirements and architecture between dense urban areas and other areas. In denser urban areas, macro cells are being divided into smaller cells to increase capacity. Metro cells (also known as pico cells), which incorporate WiFi as well as licensed band wireless, are also being introduced as a way to offload the macro network in hotspots or create LTE underlay networks. For both new macro sites and metro cells, the availability of cost-effective siting, power and backhaul are critical, and difficulties often

lightRadio Portfolio Customer solutions | Technology White Paper #3

arise in deploying a traditional rack-mounted baseband unit in a temperature-controlled enclosure at the base of a tower. At the same time, wireless operators are rolling out LTE, usually in a different spectrum band than previous 2G and 3G radios, which requires new radios and antennas. This step often creates additional issues with the increasing number of radios and antennas on a tower. The following sections of this paper describe key demand scenarios and requirements that networks must satisfy, as well as backhaul asset scenarios which fundamentally determine which solutions will be economically viable. 2.2 Demand scenarios Size and technology mix will vary dramatically, not only among wireless operators and regions, but within single operators as well. Technology mix In North America, CDMA, W-CDMA and LTE will be a common mix, evolving to LTE-A. Elsewhere, GSM, W-CDMA and LTE (evolving to LTE-A) will be most common. Macro scale In the 2013 timeframe, it will be rare for any 3G macro cell site to have fewer than 2x5 MHz carriers of W-CDMA or more than 6 carriers. Where LTE is present, it will generally be 10 MHz 2x2 multiple-input, multiple-output (MIMO) or larger, with configurations up to 20 MHz 4x4 MIMO possible. In 2013, a typical European macro cell site near the upper end of the size scale could have the following characteristics: GSM: 4+4 (900 MHz and 1800 MHz) W-CDMA: 1 5 MHz carrier MIMO 2x2 at 900 MHz and 3 carriers 2x2 at 2100 MHz LTE: 10 MHz (800 MHz) and 20 MHz (2.6 GHz) MIMO 4x2 Number of Sectors: 3 The standard for linking remote radio heads with baseband (digital) processing is known as Common Public Radio Interface (CPRI), which this paper will refer to as CPRI Interconnect, whether it is compressed or uncompressed. The large macro site forecast for 2013 would have around 30.7 Gb/s of raw CPRI traffic or about 11 Gb/s with compressed CPRI, using new Bell Labs compression techniques that have an industry-leading compression factor of nearly 3. CPRI traffic is symmetric (on the uplink and downlink) with a constant bit rate, and it is sensitive to latency and jitter. Thus, CPRI interconnect will be assumed to occur over a fiber-optic wavelength () that is not shared with other types of traffic or other base stations. Since the bulk of fiber deployment costs are in the physical network build, as more base stations are converted to fiber, the likely availability of non-shared s increases. On the wireless packet core network side of baseband processing, a macro cell site of this size would have a peak downstream data rate at an IP transport interface level (backhaul) of about 288 Mb/s. This 288 Mb/s is typically carried over IP/Ethernet, which is not a constant bit rate, or particularly latency sensitive, compared to CPRI. Its maximum latency, depending on application, is 10s or 100s of ms, whereas CPRI is roughly 5 ms upstream and 3 ms downstream. The 288 Mb/s can be carried through Ethernet switching/routed access and aggregation networks and combined with other traffic types. As the network evolves to small cells with higher peak rates, baseband processing near the radio is advantageous, because it allows shared backhaul networks to be shared among many base stations.

lightRadio Portfolio Customer solutions | Technology White Paper #3

Physically, a site of this size would have multiple radio heads. If single-band radio heads are used, then five different bands times three sectors results in 15 radio heads. Some operators are now looking seriously at 6 sectors for capacity growth; this would be 30 radio heads. Various advanced processing schemes are under consideration. Some of these schemes like coordinated multipoint transmission and reception (CoMP) offer the promise of substantial improvements in capacity and performance. Others become much more viable when baseband processing is centralized, because this facilitates sharing of antenna signals and other metrics among base stations. Other techniques, such as non-coherent CoMP/coordinated scheduling/ interference cancellation, may also offer substantial gain with less impact on processing resources and bandwidth. Metro scale Metro cells (also known as pico cells) are at the opposite end of the size spectrum from the macro case just described. When deployed in high-traffic urban hot-spots, they offer the potential to deload the macro network using WiFi, 3G and LTE technologies. They are single sector, typically only one sector of W-CDMA and at most 10 Mb/s of LTE, combined with 802.11n WiFi. Consequently, they serve only a small fraction of the users in a macro sector but can offer significant gains in effective capacity by offloading the macro network. The bandwidth to backhaul at an IP level is quite small, 10s of Mb/s, but WiFi access can take advantage of much larger backhaul links if they are available. 2.3 Backhaul asset scenarios Backhaul makes up a large proportion of most wireless networks OPEX and/or CAPEX, and higher bandwidths over fiber can enable centralized baseband processing. However, most sites are not currently served by fiber, and this situation will not change quickly. So new solutions need to cope with a diversity of base station situations. In 2010, about 55 percent of all base stations worldwide were served by microwave, 25 percent by copper and the remainder by fiber. However, this distribution is not uniform by region, and the largest percentage of copper connections is found in North America.
figure 1. Global distribution of backhaul by physical media type
Mobile backhaul installed connections by medium Air 8 7 6 Thousands 5 4 3 2 1 0 CY06 CY07 CY08 CY09 CY10 Period CY11 CY12 CY13 CY14 Copper Fiber Region: Tutto Backhaul type: Mobile

Source: Infonetics, September 2010

lightRadio Portfolio Customer solutions | Technology White Paper #3

figure 2. north american distribution of backhaul by physical media type

Mobile backhaul installed connections by medium Air 900 800 700 600 Migliaia 500 400 300 200 100 0 CY06 CY07 CY08 CY09 CY10 Period
Source: Infonetics, September 2010; annotations by Alcatel-Lucent

Copper

Fiber

Region: North America Backhaul type: Mobile

Copper (T1s) get squeezed out in favor of ber and microwave

But some copper converted DSL/ FTTN

Fiber grows

But so does microwave

CY11

CY12

CY13

CY14

In most parts of the world, wireless service providers do not own or operate the wireline assets for backhaul. Even if backhaul capacity is provided by a separate subsidiary of the same converged parent company, the wireless division will typically lease it at wholesale commercial rates. For some operators, the business is truly converged wireline/wireless, and there is no regulatory separation of divisions. For these operators, investments can be conducted on the basis of marginal cost rather than wholesale price, which makes centralized baseband processing much more attractive. This additional degree of freedom is a structural advantage of converged carriers.

3. lightRadio solution
The lightRadio solution is not focused simply on radio. It is also concerned with antennas, digital baseband processing, network-based controllers and management. Wireline and microwave transport networks are also considered, along with wireless components. In addition, the solution addresses the capital cost of equipment, the total cost of ownership, CAPEX and OPEX over time and how much will be required to maintain and augment the network and cope with the many forces of change. At the antenna end, lightRadio uses a multi-band, multi-standard active antenna array (AAA) with improved capacity, performance, coverage and energy efficiency. This antenna array will typically be configured as two vertical beams per frequency, per band, per technology, on top of horizontal MIMO, but other configurations are possible. Although active antenna arrays are not appropriate for all situations, and passive antennas can also be used in a lightRadio solution, the additional capacity and performance of AAAs make them a valuable part of lightRadio.
figure 3. 4x2 active antenna array with vertical beamforming
X X X X X X X X X X X X X X X X

OO

lightRadio Portfolio Customer solutions | Technology White Paper #3

Wireless operators often need to support many different technologies in multiple sectors and multiple spectral bands. For instance, the network loading example described earlier would require 15 remote radio heads. To meet these needs, the lightRadio solution is working toward two multiband radio heads (high and low bands) that would require just six radio heads to cover three sectors. This development can help reduce the cost of ownership, including site/tower lease costs and facilitate conformance to loading constraints on towers and masts. It also helps increase flexibility for deployment in new spectral bands, without additional site construction expense, and helps gain public acceptance of the towers. At the baseband processing level, lightRadio is built with digital modules that are used in every configuration, including metro (pico) cells. Common software is put onto highly programmable System on Chip (SoC) processors, and all hardware is remotely programmable to enable easy upgrades, addition of new features, changes of technologies and changes in baseband architecture. Diverse technology mixes, such as 3G and LTE, can be served on the same hardware. The lightRadio product family also supports intelligence and processing in multiple locations. Any given network can have a mix of architectures, and operators can change between these architectures without writing off assets.
figure 4. lightradio baseband deployment options
Multi-band multi-technology radio Multi-technology digital L1-L3 All-in-one: Baseband in remote radio head GigE, microwave, etc.

Multi-band multi-technology radio CPRI/ber Multi-technology digital L1-L3 Conventional: Baseband in separate baseband unit GigE, microwave, etc. Centralized processing Multi-band multi-technology radio 2x10 Gb/s over ber Digital L1, 2, 3

Multi-band multi-technology radio Digital L1 Remote site GigE, microwave, etc.

Split processing (research) Digital L2, 3 Metro point of presence

Split processing is the subject of ongoing research at Bell Labs. It offers some of the inter-basestation coordination, scaling and maintenance advantages of a completely centralized baseband, while using just slightly more bandwidth than having a baseband on each site. Centralized processing refers to putting the baseband (digital) processing at a distance from the radio head and antennas. The first incarnation of this approach, available today, is clustering. It simply puts a stack of conventional baseband units in a central location, connected by CPRI over fiber to each remote radio head. The next version is pooling, which treats digital resources as a pooled resource and allows lightly and heavily loaded base stations to be load-balanced against the pool. A further evolution of the concept is cooperative, which shares information among different base stations to improve capacity and performance. CoMP is a set of cooperative features that are made more effective and substantially easier and more cost effective to implement through centralized baseband processing.

lightRadio Portfolio Customer solutions | Technology White Paper #3

In many cases, a site may be shared among multiple wireless operators. It may also act as an intermediate aggregation point for other traffic or gather management and telemetry data from the site. All these functions are typically aggregated through a small service aggregation router. This extension of IP networking into the wireless network has been addressed in the Alcatel-Lucent Wireless All Around strategy, along with backhaul of IP traffic delivered over IP mobile backhaul networks, which is part of the Alcatel-Lucent High Leverage Network architecture. For simplicity, this white paper focuses on OSI physical layer concerns, which are the core of transport investment decisions, and typical IP components at each site will not be described in the following solutions. Where IP backhaul (post-baseband processing) is used, delivery over an IP mobile backhaul network is common. Where baseband processing has been centralized, the CPRI interconnection from the remote radio heads is assumed to be on a separate fiber pair or colour of light () with no electrical switching at intermediate points. The combined lightRadio and backhaul/CPRI interconnect will be described with reference to four customer cases, with progressively less fiber. These cases are summarized in Table 1, using the following definitions: First mile The span between a remote base station and the first building housing aggregation equipment, such as switches and routers (typically a central office building) Second mile The span between that CO/point of aggregation and the next higher level of aggregation such as a Metro point of presence
table 1. customer cases mapped by dark fiber availability
caSe 1 caSe 2 caSe 3 caSe 4

Abundant owned fiber Scarce owned fiber No owned fiber

1st and 2nd mile

1st mile 2nd mile 1st and 2nd mile 1st and 2nd mile*

3.1 Customer case 1: Abundant dark fiber in the first and second mile Abundant fiber makes possible centralized baseband processing that typically brings traffic from dozens of base stations back to a metro site or point of presence. This may be desirable for the following reasons: No equipment is required at the base of a site, which is an advantage where space is tight or where site rental or other costs are affected by the need for a baseband unit enclosure with temperature controls. Centralization of equipment makes it simple to maintain and upgrade digital processing equipment. Centralization allows use of new techniques, such as joint processing coherent CoMP, which require multiple base station antenna signals to be processed in one location. However, locating central processing at the first point of aggregation creates a scale issue. The hotel may not be quite big enough. Specifically, where centralization gains are expected to come from averaging the traffic between heavily and lightly loaded base stations, the central site must include sites with different demographics, most commonly from business and suburban areas with different time-of-day peaks. In addition to load-sharing issues, the potential gains from different forms of CoMP are greater when all the interfering base stations are within the same centralized processing cluster. Consequently, CoMP gains can be larger if the centralized processing cluster is larger. There is no fixed rule, but for planning purposes, we are assuming that a central processing cluster should aggregate at least 15 base stations to achieve most of the potential gains, while 30 or more is desirable.

lightRadio Portfolio Customer solutions | Technology White Paper #3

figure 5. rich fiber architecture

RNC

MME

Edge router/ SGW/PGW GGSN/etc. CPRI interconnect over ber Per RRH long haul, non-colored optic Large central processing cluster Max 40 km Aggregation point Higher level aggregation point

Remote site

3.2 Customer case 2: Abundant dark fiber in the first mile, little in the second More often, operators have abundant dark fiber in the first mile but much less in the second mile. It is generally better to handle this case as if fibers were scarce in the first mile, as well as the second (as in customer case 3). The cost of large numbers of optical interfaces at the remote radio head (RRH) end make it more expensive to use more fibers. If separate fibers from the RRH are desired, then Coarse Wave Division Multiplexing (CWDM) can be used at the remote site, as shown in Figure 6.
figure 6. Site-based Wdm architecture

RNC

MME

Compatible with: Large centralized processing cluster CWDM passive optical MUX CWDM optical MUX

Edge router/ SGW/PGW GGSN/etc. Large central processing cluster Higher level aggregation point

Per RRH long haul, colored optics Remote site

If separate fiber backhaul is desired right through the first mile for example, to increase robustness or reduce single points of failure then the recommended approach is Wave Division Multiplexing (WDM) at the CO level.
figure 7. co-based Wdm architecture

RNC

MME

Edge router/ SGW/PGW GGSN/etc. CPRI over ber Per RRH long haul, colored optics Remote site CWDM optical MUX Aggregation point CWDM optical MUX Large central processing cluster

Higher level aggregation point

lightRadio Portfolio Customer solutions | Technology White Paper #3

The choice of coarse or dense WDM depends on the number of signals to be carried and the number of available fibers in the second mile. If sufficient fiber is available, then multiple fibers carrying CWDM (which can typically support eight colors) will result in lower costs. If not, Dense Wave Division Multiplexing (DWDM) may be necessary. CWDM is less expensive, but assuming no fiber costs, both options are more expensive than the home-run fiber of customer case 1, due to the number of long-haul 10 Gb/s colored optics that are required at the RRH and cluster end. 3.3 Customer case 3: Scarce dark fiber in both the first and second mile This case is likely to be the most common availability scenario. Obviously any solution described in case 4 (no fiber available) will also work for this case. For configurations with centralized baseband processing, the lightRadio solution allows for both aggregation and compression of CPRI data such that most configurations can be carried over a single 10 Gb/s fiber pair (or two pairs with diversity), and all load from all base station sizes anticipated in the next five years can be contained within two fiber pairs. As noted in customer case 1, it is advantageous to make central processing clusters as large as possible. However, in some cases, the topology of available fiber may make hub Central Offices a better choice than a higher level Metro aggregation point. The aggregation and compression of IQ signals will be available through daisy-chaining RRHs to a compressed interface, where the composite signal is available. (Non-compressed is also available where the operator wants to preserve standard CPRI). Even if as many fibers as desired are available for use in this scenario, such as one pair per RRH, this approach provides a significant savings in the 10 Gb/s long-haul optics required.
figure 8. aggregated compressed cpri architecture

RNC Central processing cluster

MME

Multi-RRH long haul, non-colored optics Remote site

CPRI Daisy-Chain and compression in RRH 10 GigE/ber (long haul, non-colored)

Edge router/ SGW/PGW GGSN/etc. Higher level aggregation point

Alcatel-Lucent expects that this lightRadio solution to become the most common method of deploying centralized processing. 3.4 Customer case 4: No owned dark fiber When considering a macro cell with no owned dark fiber, owned means available to the wireless operator at marginal cost. Dark implies that it is a spare fiber pair available to be lit up, not part of any switched or routed network. In this case, aggregate IP traffic will not exceed 1 Gb/s, so packet microwave or leased fiber-based backhaul may be used.

lightRadio Portfolio Customer solutions | Technology White Paper #3

3.4.1 Leased backhaul

The first sub-case here considers leased backhaul of any type. This fiber leasing may be on a per-Mb/s (common), per- (rare) or per-fiber (dark fiber) basis.
figure 9. Leased backhaul architectures

RNC

MME

BBU Per RRH short haul, non-colored optics Remote site

Edge router/ SGW/PGW GGSN/etc.

Higher level aggregation point

RNC

MME

Edge router/ SGW/PGW GGSN/etc. Single short haul, non-colored optics

In the leased capacity case, either the all-in-one or distributed baseband processing options are appropriate, because these will minimize backhaul leasing costs. Alcatel-Lucent lightRadio reuses the same digital building blocks and software in different architectural configurations, so either product version or a mix may be used.
table 2. all-in-one BtS and conventional baseband units (BBu) advantages: a comparison
aLL-in-one BtS verSion conventionaL BBu verSion

Simple design, few components and zero enclosure footprint High availability which avoids single points of failure Low operational cost resulting from design for zero postinstallation site visits All-IP network No tower feeder cables (just power) when microwave is used for backhaul

Easy hardware upgrades High independence, because single point failure eliminates only part of site capacity Easy access for on-site maintenance Works with existing RRHs No cross-tower cabling required

3.4.2 Owned packet microwave backhaul

In this sub-case, owned packet microwave backhaul can be single-hop, multi-hop or more complex arrangements with rings and spurs.

lightRadio Portfolio Customer solutions | Technology White Paper #3

figure 10. Single and multi-hop microwave backhaul (conventional BBu top, all-in-one BtS bottom)

RNC IP/Ethernet/ber backhaul BBU Remote site

MME

Edge router/ SGW/PGW GGSN/etc. Higher level aggregation point

RNC IP/Ethernet/ber backhaul

MME

Edge router/ SGW/PGW GGSN/etc. Higher level aggregation point

Remote site

Also more complex topologies possible, e.g. rings

table 3. microwave and millimeter wave wireless backhaul options

Microwave Capacity per radio Max range Application 11, 18 and 23 GHz 400 Mb/s per radio, up to 4 radios ~Up to 16 km at 11 GHz Distance to or between macro cell sites separated longer (e.g. > 1km) and/or terrain that makes laying fiber more expensive Moderate bandwidth, (< 1Gb/s), baseband at site only 80 GHz 1 Gb/s ~1km Macro cell sites in dense urban environments (short reach) with good line of sight, where fiber-laying costs are prohibitive Moderate bandwidth, (< 1Gb/s), baseband at site only

Constraints

3.4.3 Owned fiber with Ethernet aggregation

This sub-case is distinct from other owned cases in that the infrastructure is typically organized to home many fiber pairs in access to carrier Ethernet switch/routers. These routers may aggregate many different traffic streams from business data users, and even residential users, as well as wireless base stations. In such networks, transport of processed IP traffic is viable, but transport of CPRIformatted antenna signals is not based on the ability to guarantee the required bandwidth, latency and jitter to which the traffic will be exposed. A conventional BBU (bottom of tower) and an all-in-one unit (with baseband combined with the radio head) are both viable for processing traffic before it reaches the network interface.

10

lightRadio Portfolio Customer solutions | Technology White Paper #3

figure 11. Scarce fiber architectures (conventional BBu top, all-in-one BtS bottom)

RNC GigE/ber Conventional BBU Per RRH short haul, non-colored optics Remote site Aggregation point or ber MUX node Carrier Ethernet switch router

MME

BBU

Edge router/ SGW/PGW GGSN/etc.

Higher level aggregation point

RNC GigE/ber Carrier Ethernet switch router

MME

Edge router/ SGW/PGW GGSN/etc.

All-in-one BTS Single long haul, non-colored optics Antenna + remote radio head (no baseband) All-in-one BTS (including baseband)

Passive Optical Network (PON) is a wireline broadband network enjoying growth among many operators. When base stations are near a PON infrastructure, it can provide effective backhaul for remote site-based broadband options. PON is not, however, suitable for macrocell CPRI interconnect, because centralized baseband has too much constant-bit-rate traffic. As noted earlier, a large macro base station could generate 30.7 Gb/s of uncompressed CPRI traffic, while typical PON systems reach their maximum at 10 Gb/s per PON segment. Since the whole point of PON is to share feeder capacity among multiple users, the required bandwidth per end point clearly must be much less than the PON segment capacity. On the other hand, metro cells (especially with baseband processing in the radio head) do meet this requirement. Where an operator has deployed a PON infrastructure, adding metro cells to the infrastructure is quite straightforward.
figure 12. pon architecture (all in one metro cell example)

ONT ONT ONT ONT Remote sites Aggregation point Higher level aggregation point Splitter GigE/ber OLT Carrier Ethernet switch router RNC MME

Edge router/ SGW/PGW GGSN/etc.

lightRadio Portfolio Customer solutions | Technology White Paper #3

11

3.4.4 Base stations served by copper pairs

In this final sub-case, the operator currently has a base station served by copper pairs, such as E1/T1. But fiber is also deployed in the access network to within a kilometer or less from the base station.
figure 13. copper/hybrid architecture
G. Vector/PhantomChannel bonded VDSL/copper pairs Fiber backhaul Carrier Ethernet switch router RNC MME

Edge router/ SGW/PGW GGSN/etc.

All-in-one BTS with backhaul shown Remote site

< ~1 km FTTN aggregation cabinet Aggregation point Higher level aggregation point

In this case, the current copper pairs can be converted to channel-bonded VDSL with vectoring and Phantom Mode to achieve quite reasonable performance for small and medium-sized base stations with two pairs or more per RRH. However, this option is not compatible with centralized baseband processing, because of limited available bandwidth, and traffic asymmetry (with more bandwidth in the downstream direction). If these were not enough an issue, latency is also a concern with the ~3ms of DSL latency being borderline for viable baseband processing performance.
figure 14. channel-bonded performance
Phantom Mode demonstrator April 2010 Line 3 Line 1 1000 900 800 Attainable bitrate (Mb/s) 700 600 500 400 300 200 100 0 First line + 2nd line + Phantom Mode + Vectoring With 4 pairs + Phantomon-Phantom 390 Mb/s (2 pairs 400 m) Line 4 Line 2 Phantom Mode 3-4 Phantom Mode 1-2 910 Mb/s (4 pairs 400 m) Phantom-on-Phantom

12

lightRadio Portfolio Customer solutions | Technology White Paper #3

4. Conclusion
The lightRadio solution has been designed to provide optimal solutions incorporating CPRI interconnect/backhaul. These new solutions meet wireless operators rapidly changing requirements, while minimizing TCO.
table 4. appropriate baseband architectures for different backhaul scenarios
aLL-in-one BtS conventionaL BBu SpLit-proceSSinG rrH centraLized proceSSinG cLuSter

Abundant 1st and 2nd mile dark fiber Dark fiber in 1st but not 2nd mile Scarce dark fiber in both 1st and 2nd miles PON Leased capacity fiber/ Metro Ethernet Microwave DSL/FTTN

Metro and macro Metro and macro Metro and macro Metro and macro Metro and macro Metro and macro Metro and macro Metro and macro

Metro and macro Macro with WDM CPRI aggregation and compression into 10 Gb/s X Not enough capacity for macro, possible for metro X Not cost effective X Bandwidth and latency not assured X Not enough capacity X Not enough capacity

The Alcatel-Lucent lightRadio product family provides the following key features and benefits to support optimal solutions: Versatility lightRadio products work with both fiber and microwave backhaul (whether leased or owned). They even enable DSL/FTTN copper backhaul for small cells, by leveraging industry-leading Alcatel-Lucent wireline broadband capabilities. Cost-optimized solutions for baseband centralization are supported as well. ALU is a recognized worldwide leader in such innovative Wireline technologies (DSL, GPON, OTN, Optical switching), and consequently completes, with lightRadio, a unique offer of end-to-end skills and solutions that can be leveraged by the operators to transform all their Wireless infrastructure, including backhaul, Wireless IP network, and packet core. Expanded reuse Remotely programmable baseband supports different GSM, W-CDMA, and LTE traffic mixes on common digital hardware, helping to avoid asset obsolescence and protect operators from shifts in the wireless technology in subscriber devices. Further, digital modules can be re-used across multiple baseband architectures, All-in-one, conventional, and centralized. Easier migration Software-configured migration is supported from decentralized to split processing to centralized baseband with no hardware changes on site. More efficient use of site resources Multi-band capabilities help reduce the number of pieces of equipment on a tower, without increasing the risk of a potential high-impact failure. Interoperable hardware to serve any location The lightRadio product family includes both multi-sector macro and single-sector metro (pico) offerings with identical features, seamless interoperability, and optimized hand-off and interference avoidance.

lightRadio Portfolio Customer solutions | Technology White Paper #3

13

5. Acronyms
AAA BBU BTS CDMA CO CoMP CPRI CWDM DSL DWDM E1 FTTN G.Vector GGSN GigE GSM LTE LTE-A MIMO MME PGW PON RAN RNC RRH SGSN SGW SOC T1 VDSL VPLS W-CDMA WDM Active Antenna Array Baseband (Digital processing) unit Base station (Antenna + Radio + Baseband) Code Division Multiple Access Central Office Coordinated Multipoint Transmission and Reception Common Public Radio Interface Coarse Wave Division Multiplexing Digital Subscriber Line Dense Wave Division Multiplexing European Multiplexed Carrier 2 Mb/s over copper pair Fiber to the Node ITU-T standard G.993.5 DSL crosstalk reduction method Gateway GPRS Support Node (for W-CDMA) Gigabit Ethernet Global Standard for Mobile communication Long Term Evolution Long Term Evolution Advanced Multiple Input Multiple Output antenna system Mobility Management Entity (for LTE) Packet Gateway (for LTE) Passive Optical Network Radio Access Network Radio Network Controller (W-CDMA) Remote Radio Head Serving GPRS Support Node (for W-CDMA) Serving Gateway (LTE) System On a Chip North American Multiplexed Carrier, 1.54 Mb/s over copper pair Very high bit rate Digital Subscriber Line Virtual Private LAN Service Wideband Code Division Multiple Access Wave Division Multiplexing

6. Author
Jonathan Segel, Alcatel-Lucent Bell Labs [email protected]

14

lightRadio Portfolio Customer solutions | Technology White Paper #3

www.alcatel-lucent.com

Alcatel, Lucent, Alcatel-Lucent and the Alcatel-Lucent logo are trademarks of Alcatel-Lucent. All other trademarks are the property of their respective owners. The information presented is subject to change without notice. Alcatel-Lucent assumes no responsibility for inaccuracies contained herein. Copyright 2011 Alcatel-Lucent. All rights reserved. CPG0591110117 (02)