The lynchpin

Last month, we examined the transmitter portion of a land-mobile radio system. In this article we turn our attention to the receiver portion, including the study of noise and interference, which is very important to the understanding how the proposed system will function. Also important are the link budget, antenna systems, RF coverage tools and drive-test methods.

The receiver's ability to detect the wanted signal above the unwanted signals, such as noise and interference, determines the range of a wireless communications system. Whether the receiver is in the base station or in the subscriber unit, the ability of the receiver to detect the signals determines how well the system operates at the edge of the coverage area or in the core, in buildings or in tunnels, in basements or garages, in the forest, or on the water.

The measure of the receiver's ability to detect signals above the noise is called its sensitivity and is measured in microvolts, or in decibels referenced to a milliwatt (dBm). A typical value for receiver sensitivity is about -117 dBm. The measure of the receiver's ability to detect signals without interference from the adjacent channel is called selectivity; a typical value for an LMR system is 7.6 dB. Most of the receivers found on the market today have the ability to be free of interference on the adjacent channel (including the narrowband mode) by a factor of 70 to 80 dB down from the on-frequency carrier level.

Before mentioning the methods of predicting and verifying radio frequency coverage, a quick discussion of the components that make up a radio system is warranted. The antenna system is comprised of the antenna, coaxial transmission line and filters. Sometimes, in the receive path, there are low-noise amplifiers within receiver multi-couplers (to offset channel splitting) or within tower-top amplifiers.

The base antennas in LMR systems are either fiberglass collinear or aluminum exposed dipole antennas, depending on the frequency band and application. Antennas are measured in decibels over a dipole reference, or dBd; the higher the gain, the longer the antenna for a given band. These antennas provide the means to control the radiation leaving the transmitter and coming back from the subscriber units to the base receiver. By focusing the signals with high gain, or by providing fill-in to areas near or underneath the base station or on the subscriber, the range and coverage area can be controlled better to match the system's desired coverage area.

The addition of filtering in LMR systems allows the signals generated in the transmitters to broadcast clearly over the airwaves with minimal interference generated to other users in close proximity, either in terms of frequency or location. There are many types of filters, including band-pass, reject and pass-reject. The filters normally are manufactured from either aluminum or copper, though sometimes gold is used. The filters are placed in series, or in parallel to the radio frequency circuit or path, for both the transmitter path and the receiver path. The series path is for band-pass filters; as their name implies, these filters pass either a single frequency or multiple frequencies. Notch filters normally are placed in parallel to the radio frequency path and are used to eliminate a single frequency. Band-pass reject cavities are combination filters that allow a single frequency, or multiple frequencies, to pass while notching a particular frequency to prevent interference.

Additional items in the radio frequency path are lightning-surge arrestors, coaxial jumpers, connectors and low-noise amplifiers to boost the signals in the receivers.

The link budget is the analysis of all of the losses and gains in an LMR system from the base station transmitter to the subscriber receiver and from the subscriber transmitter to the base station receiver. (This was discussed in detail last month.)

Once the antenna system is defined and the link budget is calculated, the next step in the process concerns the use of a radio frequency coverage prediction program. There are many programs available today and the best programs use terrain, morphology or clutter, link budget, antenna patterns and multiple statistical analyses to calculate the probability of usable radio frequency signals at given thresholds within the area under consideration.

These programs have three major components; the link budget, the path loss and the statistical analyses.

Link budget: The link budget, which contains data for the transmitter power, antenna system and receiver specification — including the estimated effects of noise and interference — will produce an estimated volume of radio signal throughout the study area. The path loss takes into consideration the terrain in typical block or tile sizes of 30 meters (or perhaps 3 arc seconds). The terrain is layered with the clutter or morphology loss values. These losses are estimated, using definition files from resources such as TSB-88 or measured empirical data from previous predictions of the radio signal losses, for different types of blockage, such as dense urban buildings, urban buildings, suburban buildings, dense forest, light forest and water.

Path loss: The path-loss portion of the radio frequency coverage programs take into consideration the height of the antennas, the curvature of the earth — which is dependent on the path length — free-space losses — which also are dependent on the path length — and any diffraction losses. The path-loss calculation adds the losses of free space, diffraction losses caused by terrain, the earth curvature based on path length, and the clutter-loss values associated with the different environments.

Statistical analyses: The statistical analysis of the estimation of radio coverage has many features, but there are three primary elements to consider: area, location and time. Area calculations are derived from TSB-88; the number of tiles that are at or above the design threshold divided by the total number of tiles yields the percentage of area covered. This is represented either as area covered or bounded area covered (BAPC).

Percentage of time is based on the amount of signal that fades over time in any given location, based on the velocity of the subscriber unit – either vehicle-mounted or a handheld – and is referenced, using sources such as TSB-88, as digital audio quality signal, or DAQ. A DAQ of 3.4 (or equivalent measurement of a Bit Error Rate of 2%) yields a 17.7 dB design-faded margin.

The excess fade margin will be required during the Coverage Acceptance Test Plan (CATP) phase because the minimum requirement for 95% probability of achieving a DAQ of 3.4 is equivalent to a 97% probability of achieving that DAQ over the defined area. Percentage of location, from TSB-88, is characterized with a 50% design margin in the radio-coverage-prediction tool, with a 13.2 dB margin to ensure a 30-meter tile reliability of 95%, assuming a 8 dB standard deviation.*

The location confidence correlates a single point to a 30 meter by 30 meter bin size, and assimilates all of the bins in the area. The 8 dB standard deviation establishes the difference between drive-test data and the predicted signal level across the tile, and is used to establish the quality of the data taken.

The base transmitter power and the receiver sensitivity must be measured upon system installation to insure that the transmitter and receiver do meet the specifications for which the system was designed. In addition, all system components also are measured to confirm that their losses and gains are meeting the engineering specifications. This includes measuring the antenna return loss and the coaxial line sweep(s) to confirm that there are no installation issues to contend with later on.

The final part of the system installation is to measure the noise floor at the site to confirm that the receiving part of the system meets the level for which it was designed. If the noise floor is higher by even a slight amount (over 2 dB), the range of the mobiles and portables getting back into the system will be seriously reduced. There are many systems in use today that do not work properly because of the high noise floor at the site, or noisy areas in the field where the subscriber units are located. With the move towards narrowbanding, this problem will exacerbate, and will be a huge problem for the systems affected by this issue.

In future sections we will discuss in more detail problems associated with noise and how to mitigate them.

Ira Wiesenfeld, P.E., is a consulting engineer who has been involved in the radio communications business since 1966. He is a senior member of the IEEE and has been a licensed amateur radio operator since 1963. He can be reached at [email protected].

Robert C. Shapiro, P.E, is the senior manager-systems engineering for PlantCML, an EADS Company. He serves on the TIA TR8 committee as the TSB-88.4-C task-group chair and is a senior member of the IEEE. He can be reached at [email protected].