Interference & RSSI

AMPS radios are capable of working over a large range of input signal. Close to a base station, signal levels are in the region of -40dBm. Although this would not be the design aim of the system, the phone still will have acceptable quality at -113dBm on the edge of a cell.

Where the signal level is too low for the phone to receive service, the obvious solution would seem to be to increase the base station's power. This is not necessarily the best answer, however, because not only does the phone need to receive a signal from the base station, but also the base station must be able to receive the signal from the phone. Because the majority of today's phones are portables with an output power of 0.5W, the weak link in the chain often is the phone.

Even in cases where the weak link is not the phone's output power, increasing the base-station power still is not necessarily a good idea because one carrier's received signal strength identification (RSSI) is another carrier's interference. Because of frequency reuse throughout the network, the fundamental balancing act of RF planning is giving good coverage while maintaining low interference. Base stations in busy areas are densely packed. Because it is not uncommon to have base stations as close as 0.5 mile or 0.25 mile from one another, the 416 channels that are allocated to AMPS carriers soon have to be reused.

Signal-Level Prediction All radio planning is done by predicting acceptable RF signal levels in the cells' required coverage area while making sure co-channel interference is at an acceptable level in the same area. This becomes a problem toward the edge of a cell where the signal is becoming weaker from the intended cell but increasing from a neighboring cell. This is referred to as the carrier-to-interference (C/I) ratio and usually is targeted at about 18dB. This means that the signal from the real cell is 18dB larger than the interfering signal coming from the neighboring cell (or cells). This planning usually is done by computer simulation tools, which generally give acceptable results.

This downlink coverage and interference prediction then is tested by performing drive testing, where a modified cellular phone is used to make and/or receive calls repeatedly. The vehicles are fitted with positional recording equipment so that maps can be produced of the drive-test routes and the call quality analyzed in the area of interest.

The most widely used method to measure coverage and C/I on the drive test is when the vehicle is fitted with two receivers. One receiver measures the "real" signal coming from the target base station, and the other measures the signal strength coming from the suspected interfering cell. This second channel, which is offset in frequency from the first so that the two receivers have unique signal strengths to monitor, is powered up manually at the second cell site to guarantee that it always is available when the real channel is in use by the other receiver. The two signal strengths then are recorded. The post-processing should show that the "real" signal is 18dB larger at all locations for the C/I ratio to be acceptable.

A problem with this method of RF planning, however, is that the software tools only are able to predict the downlink interference. They are not able to predict uplink interference because of the magnitude of simulating signals coming from all of the possible combinations of two phones in any position within the two cells' coverage areas. The simulation would have to try every combination of every power level coming in from both phones from every location where the two phones are expected to be. Sometimes phones are in a location where simple theory says they would not possibly be. Common examples are when the hand-offs have not occurred at the right time; the hand-off or operating parameters of the base station have not been set correctly; or the phone is being used high up in a multistory building where the propagation is good.

The only practical way of predicting and analyzing uplink interference on a base station, therefore, is to measure it with a radio receiver at the base station over several days. Identifying the interference and the "typical" signal strengths coming in from the real mobiles using the cell site can lead to some interesting findings. For example, the mobiles are coming in "too hot" or "too cold," or the "power window" is too wide because the base site is not correctly controlling the phones' output power to keep it within an acceptable range. This is usually 10dB to 15dB, which is centered on about -85dBm. (See Figure 2.) This range depends on the parameters with which the cell site has been configured, but no testing ever has taken place to ensure that what the system was designed for is what is actually happening.

Regular Maintenance It is important to continue drive testing on a regular basis, but you should perform base-site maintenance regularly as well. Routine maintenance is one of the first casualties when aggressive system enhancements are taking place and manpower is short. This maintenance should not be limited to the obvious things such as channel output powers and receiver sensitivities, but should include antenna VSWR testing. Antenna degradation is difficult to pick up from statistics and has a huge effect on the performance of a cell. The degradation takes place over a long time and is difficult to detect.

The phones are never systematically tested, and faults usually are detected by the customer instead of by the network operator, but a limited amount of testing can be performed on the phone by equipment at the cell-site while the phone is on air making calls. This can aid in the early diagnosis of phone problems.