T&M Breaking Through the Static

Wireless frequencies are your most valuable resource, apart from your staff of skilled RF engineers. How you manage and maintain your allotted spectrum determines not only subscriber call-capacity -- and hence, your revenue potential -- but also the grade of service you can offer. Noise, an RF-signal source's arch enemy, will degrade call quality and reliability, causing dropped calls, failed handoffs and impaired voice quality.

In a competitive environment, marred service translates into dissatisfaction and customer churn. You must monitor noise and interference measurements closely with an eye on the air interface to ensure successful system operation.

The ability of a cellular signal's contents to survive the harsh RF noise environment all comes down to a measure of the prevailing signal-to-noise ratio (S/N). The definition of noise in the electrical sense is any unwanted signal that is present at the output of a system or at any part within the system. Noise appears in various forms, including co-channel, adjacent channel, broad or narrowband interference. Each type exhibits its own particular nuances and affects the system at various levels and in a variety of ways.

Co-channel interference occurs when two or more sites or subscriber units are operating on the same channel and interfere with one another. With frequency reuse, the same channels are used by other cells and thus are referred to as co-channels, which offer greater capacity to the subscriber population. A greater spatial separation is of benefit here -- the farther away the two channels are, the lesser the interference. However, the constraints of achieving maximum capacity require that frequencies are reused in densely populated metropolitan areas, often as close as 10 miles away.

Adjacent channel interference occurs when two or more sites use neighboring channels. Despite each channel being bandwidth limited, a token amount of energy will break through the channel filter's spectrum and can breach the neighboring channels' bandwidth. Despite filter design and performance improvements, this scenario is never perfect, and maximum spectrum efficiency is never possible without the expense of a lowered signal quality.

Natural noise such as thermal noise (omnipresent at any stage in electrical circuitry), atmospherics and static as well as local noise from unsuppressed ignition sys-tems are broadband in general and greatly contribute to the noise floor for any communications system.

SIGNAL-TO-NOISE DEFINED In cellular system engineering, particularly in the optimization stages, noise appears mainly in the form of interference -- co-channel or adjacent channel -- and S/N becomes known as the carrier-to-interference (C/I) ratio. During a system's life cycle, the measurement of noise is of paramount importance. A new site in the system may reuse an existing frequency, an unlicensed user may be using that channel for clandestine purposes, or noise of some other form may have reared its ugly head.

S/N or C/I indicates the comparative level of the two values. As an example, a C/I of 21dB would mean the interference is 21dB lower than the wanted signal. Suppose a signal of 10mW is measured and an interfering signal level of 0.5mW is present, the S/N can be calculated as:

10 log(10) 10/0.5 = 13dB

CRUCIAL TO CELL COUNT The quantification of noise or interference in a system is routine practice among all RF engineers. Depending on the system technology (FDMA, TDMA or CDMA), certain rules determine the actual measured S/N. One of the primary concerns you face before system build-out is establishing the cell count, a factor you will use to determine build-out cost as well as the return on investment. The cell count is largely based on a predetermined S/N that allows for signal reliability and acceptable voice quality. This S/N will thus determine receive-level thresholds for both up and down links. Coupled with future frequency reuse requirements, the S/N also will determine the cell size and thus the cell count.

The minimum S/N is established by determining the acceptable voice quality for a given C/I value, based on vendor input and RF design services' recommendations. Choice is critical; we all want a low C/I, but voice quality and call reliability decrease with it. If the C/I is too high, the cell count increases, along with build-out costs.

Once you have established your system's minimum S/N, the design must be based around it. Of equal importance, it should be maintained for a live system. Co-channel and adjacent channel interference are the main causes of degraded quality and reliability in a cellular system and should be your main concern in terms of system design and optimization.

IN THE FIELD During the initial design or expansion phases, you must perform drive tests to ascertain the potential interference issues in your market's footprint. You must carry out a clear-channel test to ensure that no channels are being used illegally or interfered with. You should perform this test within a site's predicted footprint or coverage area and analyze the results for noise levels above a predetermined threshold.

A scanning receiver capable of measuring the RF energy of a particular technology's frequency and bandwidth will reveal the extant level of signal and interference present. Ideally, you should repeat this test more than once over a given period before assigning frequencies to that site.

On a live system, you can measure S/N or C/I in the true sense because the modulated carrier data, when post processed, enables an almost perfect S/N calculation, particularly in the case of digital systems. Consider a generic TDMA system reusing a given channel 20 miles away. Despite the natural and destructive signal decay over the distance the signal has traveled, a measurable amount of the signal will exist and is a potential source of interference to co-channel users. The external interfering signal will have an effect on the wanted digital signals' bit error rate that can be translated to a C/I ratio.

Some test-gear manufacturers offer equipment for this type of measurement for an assortment of wireless access technologies. Accompanying software can provide the ability to map out the system's performance, coverage and health.

Noise levels in a wireless communication system demand constant attention. If you choose to ignore the need to closely monitor your S/N and C/I, you probably will become aware of noise and interference problems first through reports of churning customers, as opposed to early feedback from your RF engineers.

Drive tests are critical to assessing a system's health. When you drive the system, you can analyze performance characteristics easily and use them to identify areas that pose a threat to good performance. Drive testing at a minimum requires a scanning receiver/test phone, a GPS receiver and a laptop computer. The scanning receiver, which ideally doubles as a test phone, receives real-time signal/interference power levels, as well as data messaging from a live system. The laptop collection software reads this information in conjunction with GPS location data and stores it in a data file.

Post-processing software can provide a mapped view of the test route with the measured S/N or carrier strength overlaid. This method provides an instant, at-a-glance snapshot of the performance and quality of a site or market.

Where frequencies are tight and capacity is in demand, combating interference can be your worst nightmare. In an ideal situation, all sites would have a low center of radiation and have a small footprint. This certainly would assist frequency reuse, but it also would increase the cell count greatly, offsetting the economic viability of it all.

Despite the pitfalls of a reduced trunking efficiency and added infrastructure equipment when sectoring cells, this may be the better way to reduce interference. By sectoring a site, what were previously omnidirectionally transmitted channels are now split into at least three sectors with a beamwidth of 120 degrees. This offers the advantage of using those frequencies again, very much closer to the site, as seen below. Side-lobe attenuation of a directional receive antenna at the site also reduces interference from a local co-channel, offering greater immunity. Reducing the radiation center of an existing site, if it is higher than 200 feet, also may reduce interference, as would downtilting the antennas.

To offset the loss of coverage from reducing the radiation center, you could construct a nearby fill-in site with a low radiation center to bridge the gaps.