3GPP51.010-1Mobile Station (MS) conformance specificationPart 1: Conformance specificationTS
In some laboratory environments, test equipment may not be available to measure and/or maintain the base station simulator burst time alignment to the accuracy required by the "unassisted" test environment described earlier. In such cases, a test configuration of the type shown in figure A6-2 may be employed for E-OTD performance validation. In this configuration, an LMU of known accuracy is used to measure the real timing difference (RTD) between the serving and neighbour base station simulators.
The "assisted" test environment should require a relatively short time (< 3 000 s) for the test environment to obtain N measurements for the purpose of calculating the RMS90 timing error of the MS. In many cases, the predominant component leading to uncertainty in RTD between the two base station simulators during the measurement period will be phase jitter, which should follow a Gaussian distribution. The timing offset or RTD between the two base station simulators is reported by the LMU. Even if an RTD measurement from the LMU is made in synchronism with an OTD measurement from the MS, some means of verification must be available to assure that the standard deviation of the RTD is kept within a range that allows the laboratory to confirm that this uncertainty does not significantly affect the results of the OTD reported by the MS. Test labs and LMU manufacturers are expected to have some means of verifying the standard deviation of the RTD reported for the test environment (including phase jitter introduced by the RF channel simulator), and declaring that this uncertainty component will be negligible relative to the 100 nanosecond RMS90 best-case requirements of 3GPP TS 05.05 (V8.7.1), Release 99, annex I.
Figure A6-2: "Assisted" E-OTD Test Environment
Almost all active RF channel fading simulators introduce some intrinsic propagation delay, even when set to an RF channel delay of zero. In some cases, this delay may be to great to compensate for using a passive delay as shown in figure A6-1 and figure A6-2. In such cases, the intrinsic delay of the channel simulator shall be included in the calculation of RTD for the test environment. Any phase jitter contribution from the RF channel fading simulator must be taken into account when evaluating the standard deviation of the test environment’s RTD.
Simulated Geometric Time Difference
Once the RTD of the test environment is known, any additional time delay added to the fading simulator RF path will simulate the effect of distance between the MS and its neighbour cell. 3GPP TS 05.05 (V8.7.1), Release 99, annex I does not specify a value for geometric time delay, in part because the MS could be equidistant from the three base stations required for E-OTD calculation in a real network. Also, in an actual network, the geometric delay an MS must contend with can vary from 0 sec to over 50 sec, however, the 3GPP TS 05.05, annex I specification is only concerned with measurement error. Consequently, the test procedures described in this subclause require that no additional time delay will be added to simulate a geometric time difference.
The serving base station simulator must be configured to include the neighbour base station simulator in its BA list. During interference tests, the interfering signal generator shall not be included in the serving base station simulator’s BA list.
Interfering BCCH Signal Generator
The interfering BCCH signal generator shall provide a continuous GMSK signal, modulated with a pseudo-random bit sequence. This signal generator shall not be frame-synchronized with the serving or neighbour base station simulators.
RRLP Measure Position Request
The RRLP Measure Position Request sent from the serving base station simulator shall include the field values listed in table A6-1.
Table A6-1: RRLP Measure Position Request Field Values, Positioning Instructions Data Element
This field is optional
Value=Multiple sets allowed
This field is optional
The following Measure Position Request components will be used when relevant to the test:
– E-OTD Reference BTS of Assistance Data;
– E-OTD Measurement Assistance Data for System Information List Element;
These two assistance data elements are necessary in the Measure Position Request to facilitate the execution of certain physical-layer E-OTD tests. For example, 3GPP TS 05.05, Annex I requires that in some instances, an E-OTD capable MS must support a specified timing measurement accuracy when the neighbour BCCH is below the device’s reference sensitivity. Without assistance data, it may be impossible to execute the necessary physical layer validation tests because the MS upper protocol layers would be incapable of decoding the BSIC of the neighbour BCCH.
MS Mode During Measurement
The MS under test shall make the requested measurements while in the dedicated mode on a TCH.
Automation of E-OTD Measurements
If at all possible, the laboratory environment used to verify E-OTD accuracy should be capable of supporting automated measurements. A minimum of 250 trials shall be utilized.
Terminal Unit RRLP Message Monitoring
The test lab shall have some means of logging the Measure Position Response message transmitted by the MS. This can be accomplished utilizing a suitable Um interface analyzer monitoring the MS RRLP messages on the uplink, or through a base station simulator capable of reading and logging RRLP messaging. The device used to capture the received RRLP messages should be capable of logging the MS observed time difference measurement to a flat ASCII file, with each of the reported OTD values in decimal.
E-OTD Measurement With 8PSK Modulated Bursts
3GPP TS 05.05 (V8.7.1) Release 99, annex I requires that an E-OTD-capable MS must support E-OTD measurements when the serving and the neighbour base stations are transmitting 8PSK modulated bursts. This test plan verifies that the timing error of an E-OTD-capable MS is maintained regardless of whether the serving and neighbour base station simulators are transmitting either GMSK or 8PSK in time slots 1 through 7.
In order to minimize the effects of "outlying" data points, the timing difference measurement accuracy of an E-OTD MS shall be calculated as an RMS value of 90% of the measurement trials with the least error. For example, if N=250 measurement trials, the trial error results x1 through x250 shall be sorted in ascending order of error. A subset M that includes 90 % of the trials in set N (M=225 trials in this example) shall be established. In this example, the subset M will include the 225 trial results with the least error from set N. The RMS error is then calculated from the data points in subset M according to Equation A6-1 below:
(Equation A6-1) RMS90 Calculation
Annex 7 (informative):
General rules for statistical testing