It is not often that a technology lives up to its hype. Does your mobile phone always find a cell? Do you ever get 54 Gbps from your WiFi link? Or 4 Gbps from your ADSL provider? Multicarrier power amplifiers (MCPAs) claim much in terms of flexibility, adaptability, and the range of applications in which they can be employed. Is all this hype justified, especially from a technology with its roots in a 1920s patent? This article sets out to prove that MCPAs justify their following in the operator community and highlights a new range of MCPA-based systems that raise the game in base station flexibility.

Smooth Operator

A good MCPA is air-interface agnostic-it doesn't care what 2G or 3G system it is working with and can even handle both simultaneously. With operators now having to manage the transition from 2G to 3G systems and being increasingly reluctant to go down the "forklift upgrade" route, flexible and reusable hardware is now crucial to their businesses. This new generation of MCPAs can support multiple air interface protocols simultaneously-GSM/ EDGE/ W-CDMA/ CDMA/ CDMA2000/ iDEN/ OFDM. It can support any carrier type, any number of carriers, any carrier spacing, and any carrier power level, subject only to a maximum overall mean output power rating. This is an unprecedented level of flexibility in a complex base station component, although not all MCPAs offer this degree of flexibility. Andrew Corporation recently announced just such an MCPA-with variants for both cellular and PCS frequency bands-and a complete system rack, including filtering and LNAs, to provide a total solution.

This degree of flexibility is not just a '"nice to have" feature, it is essential in allowing a seamless migration to future networks, with no forklift required. The alternative, as shown in Figure 1, is to use a lossy combiner-or worse, to add new antennas, feed lines, or even new masts-to support the new air interface. It is easy to see why OEMs and operators are looking to MCPAs to solve this migration problem.

Well Connected

When looking to upgrade a site with an MCPA, quick and simple paramount. In a perfect world, an MCPA rack would connect directly to the RF output socket of the existing BTS (where the antenna feedline would otherwise have been connected). This has not been possible in the past, and some skilled BTS engineering has been required, on a site-by-site basis, to integrate an MCPA subsystem.

The new MCPA solution solves this problem, as it incorporates an input tray that provides attenuation, unduplexing (where required), and low power combining in addition to full monitoring and alarm functionality. This module has the flexibility to cope with any form of BTS output (duplexed or not) with any number of carriers at any power level.

 

Figure 1. Seamless air interface upgrading with an MCPA: (left) conventional approach, requiring a lossy combiner; (right) with MCPA

Saving Grace

Operators are zealous when it comes to saving capex and opex costs and rightly so since this leads to lower call costs for us all. MCPAs can play a part in this process in a surprising number of ways.

Capex savings come in many forms. First, when looking to add channels to an existing site, it is clearly important not to compromise cell coverage. The conventional approach to adding channels is to add a lossy power combiner to sum all the required channels (both new and old) to feed the existing antenna system (see Figure 2). To overcome these losses and thereby regain the original cell coverage, the individual carrier powers must be increased substantially (requiring many, very high power, single-carrier power amplifiers (PAs)) or, as shown in the figure, new sites must be deployed. The costs associated with the conventional and MCPA-based approaches to this problem are summarized in Figure 3. The substantial savings inherent in adopting an MCPA-based approach are obvious from this table.

Figure 2: Use of an MCPA to add channels without combining losses: (top) conventional approach requires additional base sties (or much higher power Pas); (bottom) MCPA-based approach doesn’t compromise BTS range

   

Figure 3: BTS cable savings with an MCPA: (left) conventional approach, (right) with MCPA

Second, it is possible to reduce the required number of antenna feed lines to one or two (if diversity is used) in a multicarrier base station (see Figure 4). This benefit is most notable if the feeder run from the base station cabinet to the top of the tower is particularly long. Savings in feed line costs can be substantial in this case-with even greater savings possible if the MCPA is remoted.

Figure 4. Extending cell coverage with an MCPA and TMA: (top) existing coverage, (bottom) coverage extension with MCPA. This solution improves power density by about 10% at a similar or lower price than competing solutions

A remoted MCPA can also yield substantial opex savings. This deployment option substantially reduces the RF losses following the PA and results in useful savings in electricity costs.

Power to the People

"Coverage is King!" is an oft-used maxim in the cellular industry, particularly in new network or system deployments (such as in 3G, at present) where site costs and zoning can be difficult. MCPAs are a simple and cost-effective method of extending cell coverage by boosting the transmit signal. They need to be used in conjunction with a suitable tower mounted amplifier (TMA) to ensure that the link remains balanced; however, this solution is much lower cost than using additional base station transmit sites to provide added coverage, particularly in rural areas and other locations where additional capacity is not required. This is illustrated in Table 1.

		Option 1: Investment Required in New Base Stations to Restore the Coverage Area (Note)				Option 2: Investment Required in a MCPA Solution to Restore the Coverage Area (Note)	Savings
No. of Tx Channels	Cell Location	No. of Combining Stages Required	Reduction in Coverage Area	No. of Add’l Base Stations Required	Investment in Deploying Additional BTS
8 per Sector	Urban	2	50%	1	$150,000	$40,000	$110,000
	Rural	2	75%	3	$450,000	$40,000	$410,000
16 per Sector	Urban	3	66%	2	$300,000	$75,000	$225,000
	Rural	3	12%	7	$1,050,000	$75,000	$975,000

Table 1: Cost savings when utilizing an MCPA to add channels to a BTS deployment

Basis:   1) Two antennas per sector available

            2)  $150,000 investment per new BTS deployment

Note:    Includes hardware and the associated implementation costs

Details:  Eight-channel example is based upon: 25 W (44 dBm) per channel desired power level at the antenna port for the required cell coverage. Two stages of combining would be needed to consolidate the output channels to two antenna feed lines, resulting in a 6 dB (75%) RF power loss. This loss would result in a 50% to 75% reduction in the coverage area (depending on whether the cell is in an urban or rural area). An additional one to three base stations would be required to restore the original area coverage. The same is accomplished at a fraction of the cost by using an MCPA solution.

Your Flexible Friend

"Capacity is King!" is the other maxim often voiced within the industry, and this is true as networks mature and quality of service becomes more crucial to winning and retaining customers. Spectrum is a finite (and costly) resource to an operator, and making optimum use of it is crucial to providing the best possible customer service at the lowest possible cost. The key to optimum spectrum usage is an ability to assign the spectrum flexibly, and this in turn means having total freedom in deploying channels across a base station network. Early base station designs used cavity (filter-based) combiners to sum the transmit signals to feed the antenna. This method was power efficient, as it wasted relatively little of the RF transmit power in the combining process, but very inflexible in terms of the carriers that could be deployed at a given base site (see Figure 5). Generally, a 5- to 10-carrier "dead zone" existed on either side of a deployed carrier in a typical GSM cell-site installation. More recently, lossy passive combiners have been employed in base station designs-these overcome the frequency allocation restrictions of cavity combiners but at a heavy price in terms of cell site efficiency and, hence, opex cost. MCPAs offer the flexibility of passive combiners (see Figure 6), with a power efficiency similar to (or often better than) filter-based combining, therefore offering an ideal solution.

Figure 5. Efficient spectrum utilization with an MCPA: (top) filter-combining leads to many unusable channels in a BTS, (bottom) an MCPA allows complete flexibility in channel allocation

Figure 6. Uniquely, the Andrew MCPA solution can be remotely located from the BTS.

Flexibility covers more than just the electronics, however; it includes, for example, the ability to support both 850 and 1900 MHz bands in the same subrack and the use of the same form factor for future upgrades (to improve power efficiency, for example). If cars were this flexible, the manufacturer would combine the utility of an SUV with the performance and agility of a sports car and the luxury of a sedan - and it would fir into a small garage.

Remotely Interesting

In theory, there is no reason why an MCPA must be deployed within a base station cabinet, particularly when it could be much more advantageously deployed closer to the antenna (for example, at the bottom of the tower itself or on a rooftop, with the remainder of the base station in a basement). This deployment method saves capex both in the PA itself (lower feed line losses mean that a smaller, cheaper PA can be used) and also in the reduced length of feeder cable required. As noted above, it also saves opex, due to the lower RF losses and, consequently, lower electricity consumption.

The base station signals required to feed the MCPA are sent by fiber-optic cable over distances that can range form tens or hundreds of feet to tens of miles. This deployment architecture is a form of "BTS Hoteling" [1] and is illustrated in Figure 6. It is also possible to deploy the MCPA as a high power repeater - a useful feature in sparsely populated regions where coverage is more important than capacity.

Split Personality

Some operators have a split allocation in the PCS band, perhaps for historical reasons or through spectrum acquisition. This is a problem for most MCPAs because the large instantaneous bandwidth required to cover these allocations typically necessitates a reduction in the useable output power. A new-generation MCPA can supply its full output power over 45 MHz of instantaneous bandwidth. It can even stretch to 60 MHz with a small sacrifice in output power. When configured with an optional subrack, up to four amplifiers may be installed with an integrated filter/LNA combination that provides a complete coverage enhancement solution.

Heaven on Earth?

Is an MCPA system the manna from operator heaven? The answer is "yes" as the widespread deployment of such systems testifies. The new generation solution from Andrew has, however, set the bar a few notches higher with a high-efficiency, fully-flexible MCPA and subrack to help ensure that operators continue to look like saints in the eyes of their customers.

Peter Kenington is the Director of Advanced Technology, Base Station Subsystems Group, Andrew Corporation.

Visit Andrew Corp. online.