Method and system for testing base stations of a mobile telecommunications network

Abstract

A method and system for testing of base stations of a mobile telecommunications network having a plurality of cells. A base station at its antenna is connected to a testing system by a radio frequency cable. Mobile terminals of a cell are emulated. The mobile terminals transmit data and sends/receives calls within the cell via the base station. A separate channel emulator is provided for each emulated mobile terminal.

Claims

1. Method for testing of base stations of a mobile telecommunications network comprising a plurality of cells, comprising the steps of: connecting a base station at its antenna to a testing system by a radio frequency cable; emulating mobile terminals of a cell, the mobile terminals transmit data and send/receive calls within the cell via the base station; obtaining channel quality indicators, in each iteration, by utilizing a pair α.sub.1, α.sub.2 from a previous iteration, and testing N adjacent pairs α.sub.1, α.sub.2, best of the N pairs is used as a base pair in a next iteration, an exponential effective signal to interference and noise ratio mapping (EESM) expression $SIN{R}_{\mathrm{eff}}^i=-{\alpha }_1{\ln \left(\frac{1}{N_{\mathrm{bl}}}\underset{n=0}{\overset{N_{\mathrm{bl}}-1}{.Math.}}\exp \left(-\frac{SIN{R}^i\left[k\right]}{{\alpha }_2}\right)\right)}^n$ using N pairs of values α.sub.1, α.sub.2 associated with N channel quality indicators, the EESM expression converging on a correct value in response to calculation with correct pairs of values of α.sub.1, α.sub.2; wherein i is an Orthogonal Frequency Division Multiplexing (OFDM) time index, k is a subcarrier index, N.sub.bl is a total number of subcarriers to be averaged, α.sub.1, α.sub.2, are calibration parameters; and wherein a separate channel emulator is provided for each emulated mobile terminal.

2. Method according to claim 1, wherein base stations of eNodeB type for Long Term Evolution networks are connected to the testing system.

3. Method according claim 1, further comprising the step of implementing a plurality of multipath uplink channels based on a single frequency-time transform processor.

4. Method according to claim 3, wherein the single frequency-time transform processor is a fast Fourier transform type processor.

5. Method according to claim 3, wherein the multipath uplink channels are finite impulse response filters whose complex coefficients vary with time.

6. Method according to claim 3, further comprising the step of applying the multipath channels to a frequency domain prior to a frequency-time transform.

7. Method according to claim 3, further comprising the step of frequency multiplexing the multipath uplink channels before a signal is transmitted by the mobile terminals.

8. Method according to claim 1, further comprising the step of emulating a plurality of variable distances between the base station and the mobile terminals based on a single Fast Fourier Transform processor.

9. Method according to claim 8, wherein each variable distance simultaneously emulates a propagation time and weakening of a signal in accordance with a decrease in a power of the signal.

10. Method according to claim 8, further comprising step of emulating the plurality of variable distances in a form of a phase ramp in a frequency domain.

11. Method according to claim 1, further comprising the step of emulating a variation in Doppler conditions based on a single set of multipath channel implementations, a Doppler speed being induced by a sub-sampling of the single set of multipath channel implementations, and a channel implementation being associated with a delay spread which varies based on a selected sub-sampling.

12. Method according to claim 1, further comprising the step of symmetrically transposing channel implementations to ensure a continuity of a channel to allow a continuous loopback.

13. Method according to claim 1, further comprising the step of compressing a Signal to Interference plus Noise Ratio in accordance with a Doppler effect, an estimated noise, a fading, and a receiver noise factor.

14. Method according to claim 1, further comprising the step of emulating based on a Look-Up Table of an emulation of Signal to Interference plus Noise Ratio compression for calculating a downlink block error rate and a downlink channel quality indicator.

15. Method according to claim 1, further comprising the step of emulating a downlink multiple-input multiple-output Long Term Evolution channel by interacting with a returned channel quality indicator, a rank indicator, precoding matrix index and block error rate information.

16. Method according to claim 1, further comprising the step of coherently propagating the following parameters: a transmission power, a fading during transmission, a channel quality indicator, a pre-coding matrix index, a rank indicator, a Doppler, and a downlink block error rate based on a delay line type channel model and on a distance profile.

17. Method according to claim 1, further comprising the step of implementing an uplink power dynamic in a respective digital/analog partition of a dynamic specific to the mobile terminals and a dynamic specific to a cell.

18. Testing system for testing of base stations of a mobile telecommunications network comprising a plurality of cells, comprising: a radio frequency cable to connect a base station at its antenna to the testing system; an emulator to emulate mobile terminals of a cell, the mobile terminals transmit data and send/receive calls within the cell via the base station to obtain channel quality indicators, in each iteration, by utilizing a pair α.sub.1, α.sub.2 from a previous iteration, and to test N adjacent pairs α.sub.1, α.sub.2, best of the N pairs is used as a base pair in a next iteration, an exponential effective signal to interference and noise ratio mapping (EESM) expression $SIN{R}_{\mathrm{eff}}^i=-{\alpha }_1{\ln \left(\frac{1}{N_{\mathrm{bl}}}\underset{n=0}{\overset{N_{\mathrm{bl}}-1}{.Math.}}\exp \left(-\frac{SIN{R}^i\left[k\right]}{{\alpha }_2}\right)\right)}^n$ using N pairs of values α.sub.1, α.sub.2 associated with N channel quality indicators, the EESM expression converging on a correct value in response to calculation with correct pairs of values of α.sub.1, α.sub.2; wherein i is an Orthogonal Frequency Division Multiplexing (OFDM) time index, k is a subcarrier index, N.sub.bl is a total number of subcarriers to be averaged, α.sub.1, α.sub.2, are calibration parameters; and wherein a separate channel emulator is provided for each emulated mobile terminal.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) The invention will be more clearly understood with the help of the description, given below as a purely explanatory example, of an embodiment of the invention in reference to the Figures, in which:

(2) FIG. 1 illustrates the system according to the present invention in one embodiment;

(3) FIG. 2A represents a modulator and an external channel emulator according to the prior art;

(4) FIG. 2B illustrates a modulator comprising a channel emulator according to the present invention;

(5) FIGS. 3 and 4 represent the step for compressing the SINR (Signal to Interference plus Noise Ratio), in the context of the method according to the present invention;

(6) FIG. 5 illustrates the frequency-multiplexing of a plurality of multipath channels in the context of the method according to the present invention; and

(7) FIGS. 6 and 7 represent the system according to the present invention in one embodiment.

DETAILED DESCRIPTION OF THE EMBODIMENTS OF THE INVENTION

(8) The system 100 and the method according to the present invention make it possible to emulate several hundred LTE-compatible mobile terminals 30, 31, 32.

(9) According to the invention, the method makes it possible to laboratory test base stations 10, 11, 12 of a mobile telecommunications network 20 comprising a plurality of cells 40, 41, 42.

(10) The method according to the present invention comprises the following steps: connecting a base station 10 at its antenna 50 to a testing system 100 by means of a radio frequency cable 60; emulating mobile terminals 30, 31, 32 of a cell 40, said mobile terminals 30, 31, 32 transmitting data and sending/receiving calls within said cell 40 via a base station 10.

(11) There is a separate channel emulator for each emulated mobile terminal.

(12) The emulated mobile terminals 30, 31, 32 stimulate a load in the base station 10: they send calls and transmit data, particularly by means of a Voice over Internet Protocol (VoIP). La base station 10 is thus stimulated in the laboratory as it would be in a real environment.

(13) The emulated mobile terminals 30, 31, 32 are contained in a chassis 80.

(14) The method seeks to emulate the reality of the radio environment. It simulates buildings and the movements of the mobile terminals 30, 31, 32 (for example inside an automobile or carried by a walking pedestrian).

(15) The system 100 and the method according to the present invention make it possible to simulate: a multipath channel model a (wave) propagation model a Doppler model (related to mobility).

(16) The system 100 and the method according to the present invention make it possible to represent a realistic environment, taking into account the speed of the mobile terminals 30, 31, 32, their distance from the base station 10, and the environment (buildings, etc.)

(17) FIG. 1 illustrates the system 100 according to the present invention: a base station 10 communicates with emulated mobile terminals 30, 31, 32.

(18) The system 100 according to the present invention makes it possible to laboratory test base stations 10, 11, 12 of a mobile telecommunications network 20 comprising a plurality of cells 40, 41, 42. This system 100 comprises means for: connecting a base station 10 at its antenna 50 to said testing system 100 by means of a radio frequency cable 60; and emulating mobile terminals 30, 31, 32 of a cell 40, said mobile terminals 30, 31, 32 transmitting data and sending/receiving calls within said cell 40 via a base station 10.

(19) FIG. 2A represents a modulator and an external channel emulator according to the prior art.

(20) FIG. 2B illustrates a modulator comprising a channel emulator according to the present invention.

(21) FIGS. 3 and 4 represent the step for compressing the SINR (Signal to Interference plus Noise Ratio), in the context of the method according to the present invention.

(22) FIG. 5 illustrates the frequency-multiplexing of a plurality of multipath channels, in the context of the method according to the present invention.

(23) FIGS. 6 and 7 represent the system according to the present invention in one embodiment.

(24) In the system according to the present invention, there is a separate channel emulator for each emulated mobile terminal. Thus, for each emulated mobile terminal, a different channel is detected. The system according to the invention accounts for the possible presence of buildings (and associated signal reflections) and the distance from each terminal mobile to the base station. This makes it possible to represent a realistic environment.

(25) In one embodiment, the same IFFT-type function is shared by all of the emulated mobile terminals.

(26) In one embodiment, the same FFT-type function is shared by of the emulated mobile terminals.

(27) Uplink

(28) In one embodiment, a plurality of channel responses is applied to the frequency domain, taking advantage of the fact that in LTE, the resources used by the users are on orthogonal frequencies, and a term-by-term multiplication is applied prior to the application of an IFFT-type function.

(29) The method implements a plurality of multipath uplink channels based on a single FFT (Fast Fourier Transform) processor.

(30) This is done by “mapping” the various channels.

(31) In one embodiment, for each user, the responses to the initialization of the system are stored, in the presence of a Doppler effect.

(32) The instances of the channel that correspond to the slowest speed to be emulated are stored in memory. The storage is performed so as to provide sufficient channel statistics.

(33) In a particular implementation, fifty coherence times are stored.

(34) In one embodiment, the Doppler effect is emulated thanks to a sub-sampling of the channel. This sub-sampling is performed based on the mobile terminal speed that has been parameterized.

(35) In one embodiment, the method according to the present invention ensures the continuity of the channel by symmetrically transposing the channel implementations.

(36) In one embodiment, the method according to the present invention implements an uplink power dynamic in a respective digital/analog partition of the dynamic specific to said mobile terminals (30, 31, 32) and the dynamic specific to a cell (40).

(37) Downlink

(38) In one embodiment, information returned by the mobile terminal to the base station is used so that the latter can make decisions. The channel is then emulated by disregarding the physical channel and by emulating the measurements returned to the base station (eNodeB). From the channel implementations, the method seeks to model the parameters expected to be returned by the mobile terminal to the base station (eNodeB). Among these parameters are BLER (Block Error Rate), CQI (Channel Quality Indicator), PMI (Precoding Matrix Index) and RI (Rank Indicator). These parameters are directly dependent on the channel, and make it possible to model said downlink channel.

(39) In one embodiment, the method according to the present invention is implemented based on an LUT (Look-Up Table) for the emulation of the compression of the SINR (Signal to Interference plus Noise Ratio) in order to calculate the downlink BLER (Block Error Rate) and downlink CQI (Channel Quality Indicator).

(40) In one embodiment, said method also comprises a step for compressing the SINR (Signal to Interference plus Noise Ratio), taking into account the Doppler effect, the estimated noise, the fading, and the receiver noise factor.

(41) This method makes it possible to transpose standard block error rate curves under AWGN (additive white Gaussian noise) channel to the block error rate under multipath channel and Doppler conditions. The innovation of this method stems from its implementation described below, and also from the fact that it combines the Doppler effect and the multipath channel in the step for compressing the SINR (see FIGS. 2 and 3).

(42) In a first step (FIG. 3), the SINR is compressed in order to account for the Doppler speed.

(43) The SINR compression induced by the Doppler effect may be expressed according to the following formula [1] in the particular case of an SISO transmission:

(44) $SINR=\frac{P_S\times A_{\mathrm{pathloss}}}{\frac{1}{{.Math.H_{00}.Math.}^2}\left(P_N+P_{\mathrm{ICI}}\right)}=\frac{P_S\times A_{\mathrm{pathloss}}}{\frac{1}{{.Math.H_{00}.Math.}^2}\left(P_N+\frac{P_S\times A_{\mathrm{pathloss}}}{A_{\mathrm{dop}}}\right)}=\frac{{.Math.H_{00}.Math.}^2}{\left(\frac{P_N}{P_S\times A_{\mathrm{pathloss}}}+\frac{1}{A_{\mathrm{dop}}}\right)}$

(45) Where A is expressed as follows:

(46) $\mathrm{SIR}=\frac{1}{\frac{{\pi }^2}{6}{\left(\frac{f_{\mathrm{doppler}}}{•f}\right)}^2}=A_{\mathrm{dop}}$

(47) Ps is the power of the signal, PN is the power of the thermal noise, and H00 is the subcarrier channel gain. This expression may be extrapolated to the case of a MIMO transmission using the following formula:

(48) $SIN{R}_i=\frac{P_S\times A_{\mathrm{pathloss}}}{\underset{j=0}{\overset{\mathrm{Nr}-1}{.Math.}}{.Math.Q_{\mathrm{ij}}.Math.}^2\left(P_N+\frac{P_S\times A_{\mathrm{pathloss}}}{A_{\mathrm{dop}}}\right)}=\frac{1}{\underset{j=0}{\overset{\mathrm{Nr}-1}{.Math.}}{.Math.Q_{\mathrm{ij}}.Math.}^2\left(\frac{P_N}{P_S\times A_{\mathrm{pathloss}}}\frac{1}{A_{\mathrm{dop}}}\right)}$

(49) Where
Q.sub.v×Nr=inv(HW)=((HW)).sup.H(HW)).sup.−1(HW).sup.H=(W.sup.HH.sup.HHW).sup.−1(W.sup.HH.sup.H)

(50) W being the precoding matrix defined in the specification 3GPP TS36211.

(51) In order to reduce the complexity induced by these methods, the present invention proposes to pre-calculate and store in an LUT (Look-Up Table) the following terms:

(52) $\underset{j=0}{\overset{\mathrm{Nr}-1}{.Math.}}{.Math.Q_{\mathrm{ij}}.Math.}^2$

(53) For each multipath channel implementation.

(54) In a second step, the method known as “Exponential Effective Signal to Interference and Noise Ratio Mapping” (EESM) (described in documents [2][3][4] cited above) is used to compress the SINR.

(55) $SIN{R}_{\mathrm{eff}}^i=-{\alpha }_1\ln \left(\frac{1}{N_{\mathrm{bl}}}\underset{n=0}{\overset{N_{\mathrm{bl}}-1}{.Math.}}\exp \left(-\frac{SIN{R}^i\left[k\right]}{{\alpha }_2}\right)\right)$

(56) where i is the OFDM time index, k is the index of the subcarrier, N.sub.bl is the total number of subcarriers to be averaged (in time i and in frequency k),

(57) α.sub.1, α.sub.2 are the calibration parameters.

(58) In order to reduce the complexity of the second step, the present invention proposes the following method: Replacing the calculation of the Log exponential sum by the Jacobian algorithm with the additive correction term stored in an LUT. Averaging the channel by reducing the calculation of the current SINRs for each subcarrier and each time index to one SINR per subcarrier and time index subset. Calculating the SINRs simultaneously by using one or more “threads” per mobile terminal, The number of mobile terminals processed each millisecond is equal to two or more times the number of CPUs available for the system.

(59) In one embodiment, the method according to the present invention comprises a step for emulating an LTE (Long Term Evolution) MIMO (Multiple-Input Multiple-Output) downlink channel by interacting with the returned CQI (Channel Quality Indicator), RI (Rank Indicator), PMI (Precoding Matrix Index) and BLER (Block Error Rate) information.

(60) To calculate the (Precoding Matrix Index), a prior art method based on the calculation of the mutual information [2][3][4] described below is used:

(61) $\begin{array}{cc}{W}_{\mathrm{selected}}=\arg {\max }_{W_i\in \mathrm{codebook}}\underset{k=1}{\overset{K}{.Math.}}\underset{n=1}{\overset{N}{.Math.}}{I}_{\left(k,n\right)}\left(W_i\right)& \left(1\right)\\ I_{\left(k,n\right)}\left(W_i\right)=\log 2\det \left(I_{\mathrm{Nlayer}}+\rho W_i^H{H}_{k,n}^H{H}_{k,n}{W}_i\right)& \left(2\right)\end{array}$

(62) In order to reduce the complexity of the calculation, this implementation proposes to pre-calculate and store in an LUT (Look-Up Table) the following elements:
I.sub.(k,n)(W.sub.i)=log 2det(I.sub.Nlayer+ρW.sub.i.sup.HH.sub.k,n.sup.HH.sub.k,nW.sub.i)=log 2(1+ρ.sup.2(e.sub.0e.sub.3−e.sub.1.sup.2−e.sub.2.sup.2)+ρ(e.sub.0+e.sub.3))  (3)

(63) where e.sub.0e.sub.3−e.sub.1.sup.2−e.sub.2.sup.2 and e.sub.0+e.sub.3 may be stored as real scalar values.

(64) With:

(65) $W_i^H{H}_{k,n}^H{H}_{k,n}{W}_i=\left[\begin{array}{cc}{e}_0& e_2\\ e_1& e_3\end{array}\right]$

(66) Furthermore, in order to avoid using the K*N log 2 calculation expressed in (1), the arguments of (3) are factored using the expression log.sub.2(x*y)=log.sub.2(x)+log.sub.2(y). This makes it possible to replace evaluations of log 2 functions with multiplications.

(67) The EESM expression

(68) $SIN{R}_{\mathrm{eff}}^i=-{\alpha }_1\ln \left(\frac{1}{N_{\mathrm{bl}}}\underset{n=0}{\overset{N_{\mathrm{bl}}-1}{.Math.}}\exp \left(-\frac{SIN{R}^i\left[k\right]}{{\alpha }_2}\right)\right)$
uses N pairs of values α.sub.1, α.sub.2 which are associated with N CQIs (Channel Quality Indicators). This expression converges on the correct value if it has been calibrated with the correct pairs of values of α.sub.1, α.sub.2.

(69) In the prior art, we proceed as follows: in order to obtain the correct CQI, the pairs °, ° are tested in decreasing order of their associated CQI value. We search for the highest value for which the SINReff guarantees that BLER is less than 0.1, so in the worst case, it takes fifteen attempts, corresponding to the fifteen CQIs of LTE, to find it.

(70) In the context of the present invention, we proceed as follows: in each iteration, the pair α.sub.1, α.sub.2 from the previous iteration is used and the N adjacent pairs α.sub.1, α.sub.2 are tested. The best of these N pairs will be used as the base pair in the next iteration. Thus, the calculation complexity is limited to the testing of (N<<15) pairs and the algorithm is expected to converge on the optimal pair.

(71) The invention is described above by way of example. It is understood that a person skilled in the art is capable of producing different variants of the invention without going outside the scope of the patent.