Searching for a Home

"Base-station manufacturers don't want carriers to know about us," said Steve Blum, Celwave sales product manager. "Their solution for everything is microcells."

According to Breck Lovinggood, Andrew business unit manager, it's a matter of money. Adding repeaters in lieu of base stations can be a more cost-effective way to network build-out.

"We can extend network coverage for less than half the cost involved in cell division," he said.

Although Repeater Technologies' estimates are more conservative (maximum savings of 40% to 45%), these numbers still are substantial, especially for cash-strapped PCS licensees striving to build out networks. With build-out proliferation, it would seem that repeater sales are burgeoning. But major U.S.-based manufacturers have had trouble selling carriers on the notion that designing repeaters into their networks offers significant infrastructure cost savings. In fact, most manufacturers have yet to claim many sales in the United States. However, they confidently assert that repeaters soon will find market acceptance as PCS operators extend coverage.

"Really effective network repeaters are pretty new," said Hal Zarem, Ortel business manager for wireless communications. "Network repeaters work best in digital systems, and in the United States we've been slow in build-ing those out. Also, a lot of carriers find it difficult to conceptualize repeaters as integral network components. That's not how they've been used in the past."

SOMETIMES A GREAT NOTION Repeaters have a long lineage. They are common in commercial microwave networks and have been used in various commercial fixed-radio applications for decades. But their use in mobile radio and telephony networks is relatively recent, perhaps 20 years.

"It really started with the SMRs," said Dave Bolan, Repeater Technologies vice president of marketing. "That goes back to the 1970s. Use of repeaters in cellular came later, maybe 1987. The first deployments were quite specialized and were intended to plug holes in the coverage created by physical obstructions. A good example would be a long tunnel where the signal couldn't penetrate. In that case you'd put repeaters at either end of the tunnel."

Other spot solutions soon followed. Carriers placed repeaters in large enclosed and semi-enclosed areas such as shopping malls and performance venues and deployed them in ravines and canyon areas. Then, carriers used them along major thoroughfares in outlying network areas.

"Null filling is the usual term for this," added Andrew's Lovinggood. "It's generally the way repeaters have been used in the past, and still the way that most of them are used today. But there's another way to use the devices, and that's called cell extension. That's the approach that a lot of manufacturers are pushing now."

Cell extension is just what the name implies -- physical boundaries of the individual cell sites are extended. In best-case scenarios, carriers can double base-station coverage areas by adding a single repeater.

But repeater augmentation doesn't have to be limited to one repeater per base station. According to Repeater Technologies, 3:1 repeater-to-base-station ratios are feasible. In some cases, the repeater rather than the base station can become the real building block of the network. The areas defined by the individual repeaters become "virtual cells." Base stations then become the nuclei of enormous macrocells, while repeaters map out the micro and picocellular architecture that PCS and many cellular systems demand.

Although cell extension usually is not implemented systemwide, it isn't just a spot solution like null filling. For these applications, they serve as network repeaters deploying cell extension over a considerable portion of the network, specifically the portion where coverage is incomplete and network capacity is not taxed. Areas of relatively low potential subscriber density are good candidates for cell extension, but in some networks that could include most of a carrier's region.

Although building a network from simpler and more cost-effective elements seems irresistibly attractive, there is a catch. According to vendors, network repeaters and cell extension won't be useful in areas where increasing capacity is a concern.

"Repeaters provide increased coverage, not increased capacity," Lovinggood admitted. "They are not substitutes for base stations in a heavily loaded network."

According to Lovinggood, repeaters are most adaptable in rural settings. They will become attractive to carriers only after they have built out the urban core of the network and dealt with capacity in network portions with high subscriber densities. However, repeaters may be more welcome in certain digital systems.

"In CDMA systems, repeaters can actually add some capacity as well as increase coverage," he added. "They do so by equalizing the network and fostering better power control."

And several vendors are hoping that this new home will bode well. Many are announcing new products and enhancements to help fill in the technology gap.

For example, Allen Telecom offers a variety of products, including Extend-A-Cell 850 and Extend-A-Cell 800 dual-mode repeaters. Allgon's repeater systems now are available for GSM 1,900, CDMA and TDMA technologies. And Repeater Technologies is introducing over-the-air repeaters designed for CDMA at 800MHz and 1.9GHz. The repeaters include the Diversity feature, which maintains network capacity, preserves the RFER and retains the battery talk time for handsets.

The new role for the would-be lonely repeater is the result of an evolution in the design of the repeater itself and in the network's architecture. Mobile systems based on network repeaters would not have been practical in the remote past and were not even envisioned until a decade ago. Even then they had to wait for the general introduction of digital air interfaces.

RAPID EVOLUTION The first cellular repeaters were over-the-air, band-selective types. They picked up any signal within the entire allocated band, amplified it, and retransmitted it to only one base station via a directional antenna. Receiving antennas were directional as well, so they didn't pick up the transmission back to the base station and set up a feedback oscillation.

"There were a lot of problems with these early models," Andrew's Lovinggood said. "Because they amplified everything, they produced a lot of cross-modulation distortion."

But high distortion and narrow coverage was better than no coverage at all, and repeaters were deemed acceptable for null filling. However, repeaters needed improvements before they could be anything more than an expedient.

Refinements began to appear in the early 1990s, still within the AMPS cellular system and still largely within the overall application of null filling.

The first of these was the development of the channel-selective type where the repeater was tuned for only those channels specified in the channel-reuse pattern for a given cell. This cut down on modulation distortion and improved signal quality. But each separately tuned IF section and saw filter required for each channel received increased the cost and complexity of the repeater. Initial channel-selective repeaters used analog tuners, and with scores of channels being tuned simultaneously, the complexity of the RF circuitry was staggering.

"Analog channel-selective repeaters were never very successful," Ortel's Zarem noted. "The designs were just too expensive to justify."

Celwave's Blum suggested that analog systems in general didn't lend themselves to systematic use of repeaters. But today, manufacturers' goals are to leverage existing base-station infrastructure.

"In a digital system where you have wider and fewer channels and division according to time slots to boost capacity, the design of a channelized type of repeater can be greatly simplified," Blum said. "CDMA is really ideal in this respect because there are only a few channels over the entire band."

Now, repeater manufacturers concentrate on digital systems, and channel-selective repeaters predominate in this market.

But in order for the network repeater to emerge as a viable product, other evolutionary changes had to occur in the equipment design. Early repeaters faced the conflicting requirements of broad receiving-antenna coverage for picking up mobile signals and narrow directivity to avoid receiving the backhaul signal being transmitted to the base station. A sectorized receiving antenna only partially solved the problem.

A more satisfactory solution, though still imperfect, is found in channel translation where backhaul from the repeater to the base station is accomplished on a different frequency than that assigned to the mobile (a frequency to which the receiver in the repeater is not tuned). In some cases, that frequency is in the cellular band; in other cases, it is in the commercial microwave band. But microwave backhaul is increasingly uncommon because microwave transmitters and receivers are expensive, and you have to obtain additional licensing.

Many network repeaters sold today dispense with wireless backhauls altogether. Instead, they employ fiber-optic or coaxial-copper-cable linkages. These repeaters are fiber-fed.

Fiber-fed repeaters are immune from feedback and allow you more latitude of placement because they're untroubled by physical impediments that would obstruct high-frequency airwaves. On the other hand, they require cable laying, which is an expensive process.

Few repeaters on the market are equipped with diversity antennas for picking up mobile subscriber transmissions. Conceivably, you could add microprocessor-controlled smart antennas. This would allow the repeater to increase capacity and coverage. But cost considerations militate against elaboration in current generation products, and cost is ultimately the selling point of today's repeater.

DEFINING THE MARKETS The network repeater still represents technology in search of a market. Manufacturers talk about diverse applications; massive deployments within PCS, wireless local loop, andLMDS fixed-wireless broadband services; and growing use in in-building wireless PBX systems. But manufacturers didn't cite a single system-wide deployment within a public wireless mobile network in the United States, leaving no convincing proof. The commitment of so many major antenna manufacturers to the network repeater concept bodes well for the future of the category. As Bolan insisted, "We have a solution to a real problem. Competitive pressures will compel the carriers to listen to us."

In order to help you determine where in the network you can use repeaters as an advantage and what type of repeater will offer the best benefits, you need to understand their general design and function in a mobile network.

Wireless repeaters originated in vacuum-tube amplifiers, which AT&T installed in long-distance trunk lines early on. Spaced many miles apart, these ancient devices boosted the fading signal, traversing a run of copper cabling to the point where it could traverse miles out to the next repeater.

An analog process was adopted in long-distance microwave connections. The repeater included a receiver, transmitter and amplifier, which detected the signal, regenerated it and passed it along. Because microwave is strictly line of sight, repeaters were closely spaced, which meant a carrier often needed many repeaters.

These microwave repeaters are the direct ancestors of the modern cellular and PCS version, but they differ in many respects. For the most part, microwave linkages occur between fixed points in the network and involve highly directional antennas and rigidly controlled radiation along a single spatially defined channel over a single-carrier frequency. There are problems with interference and modulation distortion. But the environment that they operate in is more predictable and protected than their mobile counterparts.

A mobile repeater is a different animal altogether. A modern design of a mobile repeater resembles a base station in that it receives and transmits wireless messages from mobile units at a fixed point and provides network connectivity to mobile subscribers. Rather than occupying a simple point-to-point transmission path as the older microwave repeater did, it functions as point-to-multipoint, multipoint-to-single-point terminal, a condition that imposes new design considerations and constraints upon the repeater. Before, the transmission path was narrow and well-defined, and the bandpass was equally narrow. Now, the repeater must handle simultaneous signals at different frequencies in the band and must communicate with terminals at all points in its coverage area.

A repeater is purely an extension of an individual base station. It always connects back to the specific base station to which it is assigned, never to the switch or a base-station controller. It plays no part in implementing handoffs or, in the case of CDMA systems, in achieving power control. For these reasons, it cannot tighten or change reuse patterns, or truly augment network capacity. However, it can use existing capacity more fully.

"A properly designed repeater should be completely transparent to the network," said Dave Bolan, Repeater Technologies vice president of marketing. "It should have no effect on channel-reuse patterns or overall network architecture except that the individual cells can be made physically larger."

The repeater doesn't communicate with home location registers, intelligent network peripherals, modem pools, or other equipment involved in network management or provision of specialized services. A repeater basically is a transceiver. Except for certain diagnostic functions, it is virtually without intelligence and plays practically no role in routing calls. It cannot communicate with any terminals in the mobile network except the mobile subscriber units within its range or the single base station it serves.

Some repeaters are mere antenna boosters, but others incorporate considerable signal processing to reduce multipath distortion in the signal and improve transmission quality. Products differ considerably in the levels of channel isolation and signal-to-noise levels they afford. When evaluating product offerings, you should carefully consider price/performance ratios, as well as the signal-quality needs of the network.

Repeaters transmit radio signals at low power -- never more than 10W and often as little as 2W -- and incorporate relatively little circuitry apart from a class C RF amplifier and various DSP chips for signal conditioning and regeneration. Because they are enhancers and extenders of network performance rather than integral nodes, they do not demand the degree of redundancy, backup and failsafe provisions of base stations. A network absent even one base station is seriously compromised. But one network repeater failure will only slightly degrade coverage.

Consequently, repeaters don't usually need uninterruptible power supplies and bulky CPUs, so they are more compact compared to base stations. Some models barely exceed a cubic foot in volume. They also are relatively inexpensive. Repeaters run in the tens of thousands of dollars, much less than the 6-figure base stations.

Given the challenging role it fills, the modern mobile repeater does have a complex design. Indeed it has developed a considerable degree of specialization as well as numerous variants and subtypes.