The advantage of RLL is that it avoids the cost, bureaucracy, and delays associated with hard-wiring a telephone service. There is no copper to buy and string, no construction permits to be taken out, no trenches to be dug or telephone poles to be erected. Another significant benefit of RLL is that it allows service providers to incrementally invest in new markets. A limited number of sites can be connected to test the market, then as conditions warrant subscribers can easily be added or subtracted.

It is said that radio systems can be deployed in a single day. The speed and ease of radio installation has made it particularly attractive to countries that are just now installing communications infrastructure. The huge populations of some of these countries - most notably China and India - make RLL a potentially immense market.

System architecture uses a base station to relay telephone service to multiple subscribers, thus RLL can be viewed as a point to multi-point radio system. There is commonly a need for great dynamic range - often even greater than what is requested for CDMA type systems - and systems typically make extensive use of gain control in both the transmit and receive line-ups. This need comes from the fact that while some subscribers will be located very near to the system base station, others will be much further away; also, path losses will vary considerable from subscriber to subscriber due to the presence or absence of such fade sources as trees.

There are three main approaches to using radio for fixed access telephones:

The first is to use existing radio telephone equipment to create a radio-linked service. Everything from inexpensive 450 MHz analog TACS equipment to 900 MHz digital cellular (particularly GSM) equipment for high quality, secure transmission has been proposed. (900 MHz ISM is generally not a candidate, as it is thought to be "too trashy"; also it is only available in the U.S.) Proponents cite the ready availability and relatively low cost of cellular hardware as major advantages. (Motorola sells an adapter kit for some of their phones that is a wire that clips to the antenna to turn a "mobile" phone into a FWA phone!) Detractors point out that this approach gobbles up very expensive spectrum - while cellular services can get 30 to 40 cents/minute, fixed telephony rates are closer to 1 to 2 cents per minute. This approach may find acceptance, however, as a "stop-gap" means of deploying a system very rapidly, with the intent to later switch to hardware operating in a different frequency band. For Agilent Technologies product fit into this approach, refer to the individual cellular and cordless standards.

A second approach is to use proprietary band solutions in the 1.5 to 2.5 GHz range. Semiconductors are still readily available, drawing from Digital Cordless, PDC and PCN systems. Drawbacks include a lack of an international standard, including the changes in frequency from country to country, as well as the 2.5 GHz water absorption point in the spectrum. Unlicensed 2.4 GHz is typically avoided because due to interference concerns; however many of the semiconductor products used for ISM-2400 can be used for this type of RLL.

The third approach commonly proposed is to use solutions in the 3-4 GHz range and above. An example would be the Ionica system which operates at 3.5 GHz. Advantages include a relatively uncrowded spectrum and efforts to create a world-wide standard for RLL in this band. While the semiconductors are more expensive, it is believed that the simpler protocol and lack of need for hand-off will more than compensate. The use of high gain directional antennas, possible with stationary radio links, also helps lower the electronics parts cost.

• to Technology of Fixed Wireless Access-Columbia University