Identifying the Culprits

Optimizing wireless networks can be challenging when pilot pollution or adjacent and co-channel interference are present.

Wireless carriers consider the effects of interference at every stage of the network lifecycle. Network optimization involves drive testing the network to ensure coverage, to determine dropped-and blocked-call statistics and to troubleshoot the causes of subscriber complaints.

Interference can be external or internal. External interference can include illegal transmitters and harmonics or spurious emissions from other legal transmitters. Spectrum analyzers are often the best tools to measure external interference since they have broad frequency coverage (9kHz to 26.5GHz), excellent sensitivity and dynamic range. Many analyzers also offer sophisticated digital signal processing (DSP) capabilities to demodulate and measure complex digital signals such as CDMA. When used with a portable directional antenna, the spectrum analyzer can direct the user to the source of the interference.

Alternatively, some receiver-based drive-test systems also offer spectrum-analyzer capability over the cellular or PCS uplink and down-link bands of interest with the added benefit of logging the time-and-location-stamped measurements to a PC when equipped with internal GPS receiver modules.

Some types of interference are generated internally. Adjacent and co-channel interference can occur in analog, TDMA, or GSM networks because limited spectrum requires tight frequency reuse. Pilot pollution occurs in CDMA networks when too many pilot signals are received by the mobile phone. Internal interference also may include multipath components, which cause signal reflections, reducing the voice or data quality of a wireless call.

For any wireless network, the optimum drive-test solution requires the capability to decode the base-station-identification parameter associated with each technology. Drive-test products composed of a phone and digital RF receiver integrated into a single platform provide a solution to identifying the cause of internal interference. The phone can identify what the problem is while the receiver tells why the problem occurred and from which base station it originated.

For CDMA networks, the most common form of interference is pilot pollution. Each base-station sector is assigned an identifier called a PN offset, which is a timing offset based on the GPS even-second clock. Since each base station assigned to a particular frequency carrier operates at the same center frequency, the PN offset is used to distinguish base stations from one another.

When a CDMA phone searches for the strongest base-station signal, it identifies the PN offset of each signal it receives. It only looks for PNs for which the network tells it to search. This list of PNs, the neighbor list, constantly is changing since it depends on the phone's current location. Pilot pollution occurs when the CDMA mobile phone's rake receiver receives more than three (four for newer phones) pilot signals having approximately the same Ec/Io relative power levels.

The phone's three rake-receiver fingers allow it to enter 3-way soft handoff with three base stations, providing robust handoffs and minimizing the chance of a dropped call. Alternatively, the rake fingers can use multipath signal reflections to its advantage, holding a call even in low signal-level conditions.

Receiver-based drive-test tools can identify pilot pollution. The receiver can measure all 512 possible pilots. Simultaneous measurements of Ec and Ec/Io of each pilot are displayed on the PC screen. Whenever more than three significant pilots are present, pilot pollution is easily seen. (See Figure 1.) Visible or audio software alarms further automate the process. When the receiver is used with a companion CDMA mobile phone in the drive-test system, some products can link the phone to the receiver to tell the receiver the PN increment and carrier frequency that the phone is using in the network. The phone logs FER values (that relate to voice quality), dropped calls, and blocked calls. When alarms are set using Boolean expressions to notify the operator when high FER and pilot pollution are simultaneously present, for example, this type of drive test solution is indispensable.

Although these receivers are often called PN scanners, not all receivers are alike. When used with companion collection software, some receivers also make CW power, channel power, and spectrum analyzer measurements.

Interference In TDMA and GSM networks, the main sources of interference are adjacent and co-channel interference. (See Figure 2.) In both networks, subscriber calls are assigned to frequency channels. Since frequencies are reused tightly to increase network capacity, co-channel interference can occur when a phone receives two signals from different base stations using the same frequency. Each time the network design and frequency-reuse plan is updated, the likelihood of co-channel or adjacent channel interference conditions increases.

A common reuse plan is 7/21 - each frequency is used once in every cluster of seven cells. Since each cell often is split into three sectors, each frequency is used once within the 21 sectors comprising the 7-cell cluster. Each frequency then is reused again many times throughout the cell clusters of the operator's network. Automatic frequency-planning software tools can simplify the RF design engineer's job of determining how to reuse these frequencies in a manner that minimizes the possibility of interference. Pointing the antennas of cells using the same frequency away from one another and using natural barriers to minimize interference are two common techniques. The technique of using alternating channel assignments for two base stations using the same channel set also can prove effective.

To keep track of the frequency channels that belong to each base-station sector, cell-channel-planning tables are used, composed of rows and columns of channel numbers (or frequencies). Each column is a channel set representing a base-station sector and the channels assigned to it. A 7/21 network has 21 columns, each containing many rows of channel numbers. Each channel set is repeated again in the next cluster. Knowledge of the channel table combined with the right kind of drive-test equipment enables engineers to solve adjacent and co-channel interference problems.

When co-channel interference is present, phone-based drive-test tools alone are unable to distinguish the primary and secondary signals in a co-channel interference situation. The solution is to use drive-test tools consisting of an integrated phone and digital receiver.

Identifying Interference in TDMA Networks Determining the source of adjacent and co-channel interference in IS-136 TDMA networks requires an integrated phone and DSP receiver-based drive-test tool. Initially, the phone determines where a high bit-error rate (BER) is present. A DSP receiver-based drive-test tool then scans a predefined set of channels that are suspected of causing high BER on the serving channel BER. The DSP receiver must be able to decode the digital verification color code (DVCC) of each base-station sector it measures.

Measuring adjacent-channel interference is straightforward using a receiver with DVCC-decode capability. Using its 30kHz IF filter and DSP processing, the receiver tunes to the serving channel and its adjacent channels, measures the power in each channel and displays the DVCC of each signal. Drive-test tools that can link the phone's serving channel to the receiver's tuned frequency can keep the receiver tuned to the serving channel automatically. High BER will occur if one or both of the adjacent channels is equal to or higher in amplitude than the serving channel. The interfering base station is identified easily by noting the DVCC of the adjacent channel signal having the high power level. The test is performed quickly without the need to take the suspected interfering base stations off the air, thereby maintaining maximum network capacity.

Co-channel interference always has been a difficult problem to solve. If the co-channel interferer is less than 17dB below the serving channel, high bit errors will occur. The first step is to measure the BER of the serving channel using a phone-based drive-test tool. Next consult the channel-planning table to see which channel list includes the serving channel. Then a receiver scans this list of frequencies looking for an interferer with the same channel number but different DVCC.

Since the co-channel interferer is lower in amplitude than the serving cell, it will be masked by the serving signal on the drive-test display. A preferred technique involves one of three inferred methods to identify the interferer: wait for a clear channel (a channel that is in the channel list of the interfering cell but not the serving cell); wait for an idle channel at the serving base station (both cells have overlapping channel lists but one channel goes idle during a period of time); and force an idle state at the serving base station. For example, using the first method: If the amplitude and DVCC of a clear channel within the channel list of the interfering base station is measured by the scanning receiver, its amplitude relative to the serving channel can be determined. This is a C/I measurement. The C/I will be accurate if all channels in a sector are set to transmit at the same power level. This practice is followed in nearly all cases, except when downlink power control is implemented. The co-channel interferer can be determined by reading the DVCC value from the receiver-based drive test display.

Identifying GSM Interference An integrated phone and DSP receiver-based drive-test tool also is used to solve adjacent and co-channel interference in GSM networks. It must be able to decode the base-station identity code (BSIC) of each base-station sector it measures. If the receiver also can decode the BSIC of the co-channel interferer in the presence of the serving channel, then time spent troubleshooting can be reduced.

In a typical scenario, a phone-based drive-test tool is used to identify areas of poor RxQual, a parameter related to bit-error ratio that can affect voice quality. Using post-processing software, it may be determined that while RxQual is poor, RxLev (received-signal strength) is good. Consequently, interference rather than poor coverage is suspected.

The first thing to check is adjacent-channel interference. If the adjacent-channel levels are found to be within reason, then co-channel interference is suspected. Previous approaches required all co-channel cells to be turned off and the network to be driven again to see if RxQual improved.

The GSM standards state that a wireless phone should operate correctly for a wanted primary signal in the presence of an unwanted secondary signal that is 9dB or more below the wanted signal. Ideally, the C/I ratio is measured using a receiver-based drive-test system.

However, measuring co-channel signals is not that simple. The total signal power at the phone's antenna can include combinations of direct path signal, short multipath components, long multipath components and co-channel components.

Reflections from objects can cause multipath. The reflected portions of the signal are out of time phase with the original signal causing destructive interference and affecting the RxQual. Short fading components may have reflected from nearby objects or long-path from more distant objects such as hills.

Figure 3 on page 31 shows two base stations transmitting on ARFCN 200. The mobile phone and measuring receiver in the vehicle receives signals from both base stations. Each of these signals could have a primary (direct) and multipath component. Identifying the wanted signal (primary carrier) ratio to noise (interfering carrier and background noise component) signal can help determine if co-channel interference is the source of the problem.

A receiver-based drive-test tool can decode the BSIC of the base station causing the co-channel interference. This parameter is known as the secondary BSIC, since it is the signal causing interference to the primary serving channel. Also, the receiver makes simultaneous measurements of other signal and interference parameters.

When drive-test tools are used with companion software tools, operators can have more confidence in managing and optimizing their networks throughout each stage of the network lifecycle. And these solutions will be scaled and expanded to grow with them as their needs emerge for new 2.5 and 3G network technologies.