Selectivity and blocking tests measure a receiver’s ability to receive the wanted signal (to achieve≥95% of the maximum throughput of the reference measurement channels) at its assigned channel frequency in the presence of interfering signals in adjacent channels and beyond. As usual, this requirement must be met when the transmitter is set to 4 dB below the supported maximum output power. A low SINR is assumed. A summary of the selectivity and blocking tests for the LTE UE in a 5 MHz channel is illustrated in Figure 21.15. The requirements generally scale with bandwidth.
5MHz channel bandwidth
5 MHz Adjacent
channel interferer Narrow-
band blocker 29 dBc 31 dBc
Second adjacent channel interferer
Third adjacent channel interferer 38 dBc
50 dBc
RF band edge
≥30 MHz
≥60 MHz ≥85 MHz Spurious response
requirement -44 dBm
-30 dBm -15 dBm
In-band Out-of-band
Out-of-band blockers
Wanted signal
31 dBc -25 dBm
Large adjacent channel interferer
Wanted signal for large adjacent test
Figure 21.15: Selectivity and blocking requirements for a 5 MHz UE.
As with UMTS, the selectivity and blocking requirements include the case of a close Continuous-Wave15 (CW) blocking signal, some cases with modulated interferers in the first three adjacent channels, and some OOB blocker requirements with a spurious response allowance [5].
21.4.5.1 Adjacent Channel Selectivity
Adjacent Channel Selectivity (ACS) is a measure of a receiver’s ability to receive a wanted signal at its assigned channel frequency in the presence of an adjacent channel interfering signal at a given frequency offset from the centre frequency of the assigned channel, without the interfering signal causing a degradation of the receiver performance beyond a specified
15A ‘continuous wave’ signal is an unmodulated tone.
limit. ACS is predominantly defined by the ratio of the receive filter attenuation on the assigned channel frequency to the receive filter attenuation on the adjacent channel.
For LTE, ACS is defined following the same principles as UMTS and requiring similar performance capability up to a 10 MHz bandwidth, but is more relaxed for 15 and 20 MHz bandwidths. The LTE ACS is defined for each bandwidth using a modulated LTE signal as the interferer, and is only defined for low SINR conditions.
In order to check the ability of the receiver to handle the full required dynamic range, the ACS requirement is specified for two cases – a small adjacent channel interferer power and a large adjacent channel interferer power. These cases are explained in more detail below.
ACS with a small adjacent interferer
In this ACS case, the wanted signal is, like in UMTS, 14 dB above REFSENS (given in Table 21.8) and therefore takes a different absolute level for each bandwidth. The Carrier- to-Interference Ratio (C/I) is set at−31.5 dB for bandwidths up to 10 MHz and at -28.5 dB and -25.5 dB for 15 and 20 MHz respectively; the ACS is quoted as being 33 dB including a 2.5 dB implementation margin for bandwidths up to 10 MHz and relaxed to 30 dB and 27 dB for 15 and 20 MHz respectively.
Up to a bandwidth of 5 MHz, the bandwidth of the interferer is the same as the bandwidth of the wanted signal. Above 5 MHz, the bandwidth of the interferer stays at 5 MHz, which means that the RF test equipment does not need to be able to generate a wide bandwidth interferer. For bandwidths above 5 MHz the interferer is positioned at the near edge of the channel. The consequence of this is that the interferer power is concentrated at the edge of the adjacent channel at which the filters used in the receiver have least selectivity. The digital filters will have a sharp cut-offbecause they are designed for OFDM, but the analogue filters in the RF front-end of the receiver could have much less attenuation at the near edge of the channel, which will push up the dynamic range at the ADC. To compensate, the ACS is relaxed by 3 dB and 6 dB for the 15 MHz and 20 MHz modes respectively as mentioned above.
Level diagrams for ACS requirements for two bandwidths (5 and 20 MHz) are shown in Figure 21.16.
The gap between the edge of the wanted signal and the edge of the interferer is a function of the used bandwidth of both signals. Given that both use a full allocation of RBs, this gap can be calculated as 1.25 MHz for 20 MHz bandwidth reducing to a minimum of 300 kHz for 3 MHz bandwidth. Below this bandwidth the channel usage reduces, so the gap increases.
Note that in percentage terms, a 1.25 MHz gap adjacent to a 20 MHz channel is actually smaller than a 300 kHz gap adjacent to a 3 MHz channel (6.25% compared to 10%). The channel filters should be designed to scale with bandwidth, so it is this percentage ratio which determines the filter roll-offrequirements. Comparing the 20 and 5 MHz modes we see that the percentage has roughly doubled but the specification is relaxed by 6 dB, so the filter roll-offrequirements should be similar.
ACS with a large adjacent interferer
The large adjacent interferer requirement uses an adjacent channel interferer power of
−25 dBm. Similarly to the case of the small adjacent interferer, the SIRC/I is fixed at
−31.5 dB for bandwidths of 10 MHz and below (i.e. the wanted signal power−56.5 dBm)
REFSENS=-100dBm
REFSENS=-93.9dBm
20MHz channel bandwidth
5MHz channel bandwidth
Pinterferer=-54.9dBm Pwanted=-86dBm
C/I=-31dB
C/I=-25dB 5MHz
Pinterferer=-55dBm
5MHz channel bandwidth
ACS
Wanted Signal
Adjacent Channel interferer 14dB
SNR+IL +margin
5MHz channel bandwidth
Pwanted=-79.9dBm
Wanted Signal
Adjacent Channel interferer 14dB
12.5MHz
Figure 21.16: Adjacent channel selectivity illustrated for Case 1 [5].
and−28.5 dB and−25.5 dB for 15 and 20 MHz respectively (i.e. the wanted signal power is
−53.5 dBm and−51.5 dBm). The margin between the wanted signal power and REFSENS varies for each bandwidth in the range 41.0 to 50.2 dB, which is well above the noise floor and therefore not of much significance (see [5, Section 7.5, ACS Case 2]).
In practice, the toughest test for the receiver is likely to be somewhere between the small interferer and large interferer ACS tests. Assuming that the dynamic range of the receiver is limited, there is a point at which the interferer power first becomes high enough to require that the front-end gain needs to be reduced. This gain reduction will degrade the noise figure of the receiver at a point at which the wanted signal power is not very high. The gain control algorithm used by the receiver must be well planned to avoid such problems.
Adjacent Channel Interference Ratio (ACIR)
The ACS and ACLR (see Section 21.3.2.1) together give the Adjacent Channel Interference Ratio (ACIR). The ACIR is the ratio of the total power transmitted from a source to the total interference power affecting a victim receiver, resulting from both transmitter and receiver imperfections.
It follows that
ACIR 1
1
ACLR+ 1 ACS
ACLR and ACS have been extensively used for coexistence studies.
21.4.5.2 Narrowband Blocking (in Adjacent Channel)
The blocking characteristic is a measure of the receiver’s ability to receive a wanted signal (to achieve≥95% of the maximum throughput of the reference measurement channels) at its assigned channel frequency in the presence of an unmodulated unwanted interferer on frequencies other than those of the spurious response or the adjacent channels, without this unwanted input signal causing a degradation of the performance of the receiver beyond a specified limit. The blocking performance applies at all frequencies except those at which a spurious response occurs.
The narrowband blocking specification is a severe test of the receiver’s ability to reject 3rd order intermodulation products resulting from cross-modulation of the transmitter leakage which appears around the narrowband blocker. The frequency of the unwanted cross- modulation product depends only on the narrowband blocker frequency and not on the frequency of the transmitter, or any other modulated blocker.
The LTE ‘narrowband’ blocking performance requirement uses a CW interferer very close to the wanted signal at an offset less than the nominal channel spacing. For such small offsets nearly half of the transmitter leakage will appear in-band.
The CW blocker is positioned at approximately 200 kHz from the near edge of the adjacent channel. For example, for a 5 MHz bandwidth the offset is∼2.7 MHz, which is 450 kHz from the edge of the wanted signal. The offset of the blocker from the edge of the wanted signal reduces with bandwidth, reaching just 350 kHz for the 3 MHz channel bandwidth. Below 3 MHz bandwidth the channel occupancy reduces, which compensates for the reduced gap.
For this test, for bandwidths of 10 MHz and below, the power of the wanted signal is set at the REFSENS level plus a bandwidth-specific offset (22, 18, 16, 13, 14 and 16 dB for 1.4, 3, 5,10, 15 and 20 MHz respectively – see [5, Section 7.6.3],); the blocker power is set to
−55 dBm independently from the bandwidth. Compared to the ACS test, the gap between the wanted and interfering signals is between 30 and 50 kHz less, which makes the narrowband blocking test a little more demanding than the ACS test.
In UMTS the wanted signal power is 13 dB higher than the REFSENS and the interferer power is set to−56 dBm. The most important difference between the UMTS narrowband blocking specification and the LTE case is that the blocker for UMTS is actually a GMSK16 modulated signal, not CW. A GMSK signal is a little narrower than QPSK (although clearly not as narrow as a CW signal), and the modulation has a constant envelope so there is no PAPR variation which could increase non-linear distortion. On balance, it can be concluded that the UMTS and LTE narrowband blocking specifications are similarly demanding.
21.4.5.3 Non-Adjacent Channel Selectivity (In-Band Blocking)
Non-adjacent Channel Selectivity (NACS) is a measure of the receiver’s ability to receive a wanted signal at its assigned channel frequency in the presence of unwanted interfering
16Gaussian Minimum-Shift Keying, as used in GSM.
signals falling into the receive band beyond the adjacent channel or at less than 15 MHz offset from the edge of the receive band. The interfering signals are modulated and occupy the same bandwidths as specified for ACS. The LTE specifications refer to this test as ‘in-band blocking’, although, unlike the other blocking tests, NACS does not use CW signals.
There are two requirements to be met, the first with an interferer of−56 dBm17 in the second adjacent channel or further (referred to as Case 1), the second with an interferer of
−44 dBm18in the third adjacent channel or any larger frequency offset up to 15 MHz out of band (referred to as Case 2). Furthermore, a specific requirement which applies to assigned UE channel bandwidth of 5 MHz and for Band 17 (referred to as Case 3) is also provided [5].
Unlike the ACS tests, NACS does not need to be repeated at higher signal levels, so it does not test dynamic range to the same extent. However, the wanted signal level is much lower, at just 6 dB above REFSENS for bandwidths up to 10 MHz, and 7 dB and 9 dB above REFSENS for 15 and 20 MHz bandwidths respectively. Consequently the C/I ratios are much lower, falling for example to−50 dB for the third adjacent channel.19 The total filtering requirement will therefore be of the order of−60 dB at three times the bandwidth.
A summary of the NACS requirements for LTE and UMTS UEs is shown in Tables 21.9 and 21.10. For UMTS, the wanted signal power includes 21 dB spreading gain.
Table 21.9: Non-adjacent (N±2) channel selectivity.
System LTE UMTS
Bandwidth of
1.4 3 5 10 15 20 3.84
wanted signal (MHz) Own signal power
6 6 6 6 7 9 3
above REFSENS (dB)
Power of −56
interferer (dBm)
Frequency offset ±2.8 ±6 ±10 ±12.5 ±15 ±17.5 ±10 of interferer (MHz)
Bandwidth of
1.4 3 5 5 5 5 3.84
interferer (MHz)
21.4.5.4 Out-of-Band Blocking
The LTE OOB blocking tests measure the receiver’s ability to receive a wanted signal at its assigned channel frequency in the presence of unwanted interfering signals falling outside the receive band, at 15 MHz or more offset from the edge of the band.
The wanted signal power is at the same level as for the in-band blocking test (at 6 dB above REFSENS for bandwidths of up to 10 MHz, and relaxed to 7 or 9 dB above REFSENS for bandwidths of 15 MHz or 20 MHz respectively).
17Same as for UMTS.
18Same as for UMTS.
19This example is for a REFSENS of−100 dBm for Band 1.
Table 21.10: Non-adjacent (N±3) channel selectivity.
System LTE UMTS
Bandwidth of
1.4 3 5 10 15 20 3.84
wanted signal (MHz) Own signal power
6 6 6 6 7 9 3
above REFSENS (dB)
Power of −44
interferer (dBm)
Frequency offset ±4.2 ±9 ±15 ±17.5 ±20 ±22.5 ±15 of interferer (MHz)
Bandwidth of
1.4 3 5 5 5 5 3.84
interferer (MHz)
The blocker is a CW signal with a power of−44 dBm at 15 to 60 MHz offset,−30 dBm from 60 to 85 MHz offset and−15 dBm from 85 to 12750 MHz offset.20These values are all the same as UMTS.
The actual offset of the blocker is the specified offset from the band edge plus half the wanted signal bandwidth plus the RF guard band, which is 2.4 MHz. Hence, for example, for a 5 MHz bandwidth, the−44 dBm blocker is at a minimum offset of 15+2.5+2.4= 19.9 MHz, which is close to four times the bandwidth.
As in UMTS, there is an additional requirement for RF bands 2 (DCS1800), 5 (GSM850), 12 and 17, comprising a−15 dBm blocker coming from the transmit band, which is at an offset of just 20 MHz for bands 2 and 5, 12 MHz for band 12 and 18 MHz for band 17, see Table 21.1. This latter requirement is clearly far tougher than all the other blocking specifications.
A summary of the selectivity and blocking requirements is shown in Figure 21.17. This includes the specified 2 or 3 dB margin added to the C/I requirements.
When normalizing all frequency offsets to the wanted signal bandwidth, it can be seen that the 20 MHz LTE bandwidth requires the most severe filter frequency response relative to its bandwidth (and also to the digital sampling frequency). Additionally, the LTE 5 MHz selectivity requirement is relatively tougher than UMTS.
21.4.5.5 Spurious Response Specifications
Frequencies for which the throughput does not meet the requirements of the OOB blocking test are called spurious response frequencies. Spurious responses occur at specific frequencies at which an interfering signal mixes with the fundamental or harmonic of the receiver local oscillator and produces an unwanted baseband frequency component. Spurious responses are measured by recording when the OOB blocking test is not passed.
At the spurious response frequencies, the LTE receiver must still achieve the required throughput with a −44 dBm CW blocker and a wanted signal level set to the power level specified in the OOB blocking test.
20This blocker test is defined for UEs supporting all bands a part from 2, 5, 12 and 17.
-80.0 -70.0 -60.0 -50.0 -40.0 -30.0 -20.0 -10.0 0.0
0 10 20 30 40 50 60 70 80 90 100
Frequency offset (MHz)
CombinedACS and blocking requirements (dB)
1.4MHz 5MHz 20MHz UTRA
Figure 21.17: Selectivity and blocking requirements.