Receiver Performance

As wireless deployments become more dense, cell sites often have to accommodate base stations on competing networks or base stations that operate near high-powered transmitters from diverse services on different frequency bands. The other transmitters may interfere with the operation of a base station's receivers unless you take appropriate precautions. One of these precautions is to examine receiver specifications that sometimes are ignored. The blocking specification is helpful in determining whether a site is suitable.

DIGITAL RECEIVER CHARACTERISTICS Digital receiver performance characteristics often are described by noting the receiver's ability to recover the modulation intelligence from an RF carrier, injected at low levels, into the antenna port in the presence of different kinds of interference. One aspect of digital radio that distinguishes it from analog radio is that information is imparted to a carrier by altering (modulating) one or more characteristics (phase, frequency or power) of an RF carrier in discrete steps. Each carrier state is a symbol. In GSM, information is carried one bit at a time in a special form of frequency-shift keying called GMSK (Gaussian minimum shift keying).

Because all bits or symbols on the GSM air interface are not equally important, some are more carefully coded, or protected, from radio channel hazards than others. Most voice data, for example, is less carefully coded than signaling data. This bit apartheid scheme means that symbol recovery errors have to be calculated in accordance with the class of the bits represented by the modulation symbols.

The bit error rate (BER) considers every recovered symbol when the probability of errors is calculated. The residual bit error rate (RBER) method takes a more general view. RBER is considered on speech or user data frames that are not marked by the receiver as bad or corrupted frames. This is because some traffic bits, owing to their relative importance in properly reconstructing the original transmitted voice sounds or data traffic, are coded so that the receiver can take advantage of decoder mechanisms that can, within certain limits, discern the proper state of symbols arriving in error from the state of other recovered symbols. Because these kinds of powerful mechanisms hide the individual significance of arriving symbols, the BER procedure cannot be used. The frame erasure rate (FER) procedure maintains a count of frames that contain unresolved errors when they are decoded by the receiver. The FER method of describing receiver performance is similar to the RBER method, except that it can apply to the critically important signaling bits as well as to voice or user data bits.

TESTING DIGITAL RECEIVERS GSM base stations have two parts: some transceivers and a controlling entity. A transceiver is called a base transceiver system (BTS). The controller is called the base site controller (BSC). Together, the BTS and the BSC are called the base station system (BSS). GSM infrastructure makers perform specific tests on a BTS to describe how well a BSS will perform in the presence of interference: co-channel rejection, adjacent-channel rejection, inter-modulation (IM) rejection and blocking1. Although they usually are specified, these on-channel and off-channel tests seldom are performed in the field on a live base station because the characteristics primarily are a design or installation consideration.

Co-channel rejection and adjacent-channel rejection specifications are indications of a receiver's robustness against interference from signals in the same or the adjacent channels, respectively, as the receiver tries to recover the intelligence from an on-channel GSM signal. The IM rejection specification judges the apparent linearity of a receiver by measuring its ability to reject intermodulation products caused by non-linear elements within the receiver itself. A blocking measurement investigates the behavior of a receiver in the presence of a strong transmitter operating on any frequency.

BLOCKING The ability of the receiver to receive a desired GSM signal in the presence of a strong interfering signal on any frequency is described by its blocking specification. This becomes an important test when a base station is installed close to a broadcast TV transmitter or in a crowded radio site next to a high-powered paging transmitter. The interfering transmitter may be on a different frequency from that of the base-station receiver and may even be modulated in a different way than the victim receiver is designed to demodulate. The blocking specification is a special case of the more general spurious response characteristic of a receiver.

BLOCKING CHARACTERISTICS GSM equipment deployed in the 900MHz and 1,800MHz bands is specified by the European Telecommunications Standards Institute (ETSI). ETSI publishes test procedures and equipment specifications called European Telecommunications Standards (ETS). The 1.9GHz version of GSM systems deployed in North America and Chile relies on a series of standards that are closely aligned with their ETSI counterparts and which are edited by the committee T1P1. ETS 300 609-1 (also called GSM 11.21) describes tests that should be performed on a base station system to determine how well it conforms to GSM's system design goals. The level of an interfering signal that blocks the receiver's ability to recover any useful GSM signals is the blocking specification.

There are two steps to finding the blocking level of a receiver. The first is to discover all of the off-channel interfering frequencies on which the receiver is likely to suffer blocking. The second step is to inject a high-level interfering signal at all the frequencies discovered in the first step, together with a typical on-channel GSM signal of about -95dBm.

The receiver should be able to recover the desired GSM signal in the presence of any interfering signal as high as about +8dBm, which is frequency modulated +/-100kHz at a 2kHz rate. The actual test levels can deviate slightly from the typical values in this article. The applicable ETSI or T1P1 specifications list all the on-channel GSM and off-channel interfering FM levels for various types of base stations and operating conditions.

The first part of the procedure, finding all of the likely frequencies for a blocking interfering signal, is the most tedious and expensive part of the test. The base-station manufacturer starts by disclosing all n of the intermediate frequencies (IFn) used within the receiver and the frequency of the local oscillator (LO) applied to the first mixer. Frequency hopping is turned off during the tests. Interfering signals, fi, at integer multiples of 200kHz (GSM's channel spacing) are considered between fLO and 4.0GHz. This can become an overwhelming number of frequencies, but the GSM specifications suggest some shortcuts and alternative searching schemes to find the appropriate frequencies on which to tune an interfering signal. The second phase of the test is to inject a powerful interferer at all of the sensitive frequencies, fi, to make sure the receiver is not blocked for interfering levels above values specified in a table of prescribed maximum levels, ranging from about -30dBm to about +8dBm. The receiver is declared blocked when the RBER exceeds 2%. The use of the RBER method implies the low-level, on-channel signal is modulated with actual user information. Indeed, the blocking tests call for a full-rate traffic channel for the on-channel GSM signal and that only the class II bits are examined. The maximum level of the interfering signal depends on how far away fi is from the desired on-channel GSM signal.

Any receiver can be completely blocked by an off-channel interfering signal that is strong enough. The receiver's third-order intercept specification closely tracks its blocking specification. A receiver's third-order intercept level, and its resistance to blocking, can be enhanced by shielding, front-end filters, attention to linearity and careful selection of IF. Because all of these considerations will draw a compromise from some other desired performance characteristic and increase costs, a blocking specification far in excess of what the GSM specifications recommend should be considered an extravagance that is not likely to be of much benefit to a GSM carrier. The network engineer should make sure the extra expense is justified and that the costs can be recovered. Base-station receivers that experience blocking can sometimes be compromised in their off-channel performance because of faulty installations or system designs. Performing field tests of all the other interfering scenarios will disclose the cause of the problem.