WO2009154527A1

WO2009154527A1 - Improved uplink measurements in a cellular system

Info

Publication number: WO2009154527A1
Application number: PCT/SE2008/050735
Authority: WO
Inventors: Magnus Johansson
Original assignee: Telefonaktiebolaget L M Ericsson (Publ)
Priority date: 2008-06-19
Filing date: 2008-06-19
Publication date: 2009-12-23

Abstract

A method (200) for use in a cellular communications system (100) in which a cell (110) in the system is associated with a controlling node (130) which serves to control traffic to and from users (120) in the cell. Users in the cell are enabled to make transmissions (205) to the controlling node on a plurality of uplink channels, and the controlling node transmits (210) control commands to the users on downlink control channels for parameters in the user's uplink transmissions. The controlling node measures (215) one and the same parameter in uplink transmissions from a user in both at least a first and a second uplink channel, and combines (255) these measurements in order to determine (260) one or more of said control commands for the user in question.

Description

IMPROVED UPLINK MEASUREMENTS IN A CELLULAR SYSTEM

TECHNICAL FIELD

The present invention discloses a method and a device for improved uplink measurements in a cellular system.

BACKGROUND

In certain cellular communication systems, such as, for example, LTE, Long

Term Evolution, and WiMAX, Worldwide Interoperability for Microwave Access, there is a lack of dedicated data channels for user terminals in the cells of the systems, which means that all data transmissions from the user terminals in a cell to the controlling node of the cell, i.e. "uplink transmissions", are sent on shared channels. Control channels, on the other hand, may be dedicated to a certain user terminal.

The principle of shared uplink data channels, in turn, means that user terminals in the cells in the system need to be scheduled and controlled both in time and in frequency in order to be able to receive and transmit data; this scheduling and control of the user terminals in each cell is performed by the controlling node of the cell by means of transmitting control commands for various uplink parameters to the user terminals in downlink channels.

In order to exemplify the notion of different uplink channels in, for example, an LTE system, the following LTE channels can be mentioned:

« PUSCH, Physical uplink shared channel. This channel carries data and control, and is shared between all UEs in the cell.

• PUCCH, Physical uplink control channel. This channel carries control information only, and is UE-specific. Multiple PUCCHs can exist in parallel. • PRACH, Physical random access channel. This channel is used by UEs that are not synchronized and which need to access the eNodeB.

• Demodulation reference signals. These are UE-specific reference signals (pilots) associated with PUSCH or PUCCH, and are used by the eNodeB for channel estimation of PUSCH or PUCCH.

β Sounding reference signals. These are UE-specific reference signals (pilots) not associated with PUSCH or PUCCH, and are used by the eNodeB for a wideband frequency selectivity estimate of the channel.

As mentioned, the eNodeB performs control of the UEs in the cell, by means of transmitting control commands in downlink channels. Among the UE uplink parameters which are controlled by the NodeB in an LTE system, two parameters which can be mentioned as examples are uplink power control, i.e. the output power of the user terminal, and the so called time alignment of the user terminal.

The purposes of the respective control mechanisms for these parameters are as follows:

• Uplink power control: adjust the UE's output power so that the SINR,

Signal to Interference and Noise ratio, in the eNodeB is sufficiently close to a target SINR. The target SINR in turn depends on the targeted MCS, modulation and coding scheme, and the QoS, Quality of Service, requirements of the service carried by the uplink signals. The uplink power control will affect the PSD, Power Spectral Density, in the eNodeB. • Time alignment: adjust the UE's timing so that the TO, Timing Offset, of the uplink signal is sufficiently small in the eNodeB receiver, i.e. so that the transmissions from the UE are received within a receiver "time window" in the eNodeB. The time alignment of the transmissions of the user terminal is thus an "offset" in time which the user terminal applies to its uplink transmissions with respect to a nominal time when the transmissions should be made.

With respect to these two parameters, i.e. uplink power control and time alignment, it will be realized that it is advantageous if the PSD in the eNodeB is kept as close as possible to a "target PSD", and that the time alignment should be such that the TO is minimized. For example, some reasons for keeping the PSD as close as possible to an intended value are the following:

» If the PSD is too low, the throughput in the system will decrease, since either the MCS needs to be more robust, or the PER, packet error rate, will increase,

• If the PSD is too high, the UE power consumption will be unnecessarily high, which will shorten the UE battery lifetime, and will also cause unnecessary interference in the own cell and in adjacent cells.

Regarding the TO, a small TO will result in a lower degree of interference, which is of course beneficial.

The controlling node of a cell, the eNodeB in an LTE System, carries out measurements on the transmissions from the UEs in the cell in order to properly determine parameters such as output power and timing advance which the controlling node then signals to the user terminals. However, the principle or principles used or proposed at present for such measurements often yield results which are not satisfactory.

SUMMARY It is an object of the present invention to provide an improvement in the way that a controlling node in a cellular system determines control parameters for user terminals by means of measurements on transmissions from the user terminals.

Such an improvement is offered by the present solution in that it discloses a method for use in a cellular communications system in which a cell in the system is associated with a controlling node which serves to control traffic to and from users in the cell.

According to the method of the invention, users in the cell are enabled to make transmissions to the controlling node on a plurality of uplink channels, and the controlling node transmits control commands to the users on downlink control channels.

The control commands are for parameters in the users' uplink transmissions, and according to the method of the invention, the controlling node measures one and the same parameter in uplink transmissions from a user in both at least a first and a second uplink channel and combines these measurements in order to determine one or more of said control commands for the user in question.

Since, according to the invention, the controlling node combines measurements of transmissions from more than one channel in order to determine control commands for the user which made those transmissions, an improved degree of accuracy is obtained, as compared to previous solutions. In one embodiment, of the invention, the controlling node also measures a second parameter in uplink transmissions from the user in both at least the first and second uplink channels and combines these measurements in order to determine at least a second control command for the user in question.

Also, in a further embodiment of the invention, the measurements from the different channels are weighted differently when said control commands are determined.

Further embodiments as well as further advantages of the present invention will become evident form the following detailed description.

Along with the inventive method, the invention also discloses a transceiver for use as a controlling node which functions according to the method.

BRIEF DESCRIPTION OF THE DRAWINGS

The invention will be described in more detail in the following, with reference to the appended drawings, in which

Fig 1 shows a schematic view of a system in which the invention may be applied, and

Figs 2 and 3 show flow charts of a method of the invention, and Fig 4 shows a block diagram of a transceiver of the invention.

DETAILED DESCRIPTION

The invention will be described in the following using terminology from an LTE system, Long Term Evolution, but it should be pointed out that this is merely in order to facilitate the reader's understanding of the invention, and is not intended to limit the scope of protection sought for the present invention in any way. For example, it is perfect possible to apply the invention to a system which uses WiMAX, Worldwide Interoperability for Microwave Access, as well as to an LTE system. Before the present invention is described in more depth than in the previous chapter, a brief introduction to the system for which the invention is intended will be given, in order to facilitate the understanding of the invention:

Fig 1 shows a schematic view of a system 100 in which the present invention can be applied. As indicated in fig 1 , the system 100 comprises a number of cells, one of which is shown as 110 in fig 1. Each cell can comprise a number of user terminals, "UE"s, one of which is shown as 120 in fig 1.

A cell in the system 100 will be associated with a controlling node, a so called eNodeB, shown as 130 in fig 1. A function for the eNodeB 130 is to control traffic to and from the UE 120 in the cell 110.

As mentioned previously in this text, an objective of the present invention is to obtain an improvement in the measurements of the eNodeB which in turn enable the eNodeB to send proper control commands to the UEs for various uplink parameters.

A basic principle or method 200 which the present invention uses in order to achieve its objective will now be described with reference to the flow charts of figs 2 and 3. Steps which are options or alternatives are shown with dashed lines.

As indicated in step 205 ("UL Tx"), according to the invention, the eNodeB controls, step 210, ("Control"), the traffic to and from users in the cell, and, measures, step 215, ("Measure, CH>1 "), one and the same parameter in uplink transmissions from a user in at least two uplink channels.

The eNodeB combines, step 255, ("Combine") its measurements in order to determine, step 260, ("Determine"), one or more corresponding uplink control commands for the user which made the uplink transmissions. In one embodiment, as indicated in step 220, ("> 1 Parameter"), the eNodeB will measure more than one parameter from a user in the uplink channels in question and use the measured parameters in order to determine corresponding control commands for the user which made the uplink transmissions.

Regarding the parameters which are measured by the eNodeB in the uplink channels, a wide variety of parameters can be chosen within the scope of the present invention, but in one embodiment, step 230, ("G_UL"), the uplink gain, G_UL, of a UE is measured, by means of which the UE can be controlled with respect to its output power in the uplink transmissions, so that the PSD, the Power Spectral Density, in the eNodeB is maintained at a constant and desired level.

In addition, in another embodiment, shown in step 225, ("TO"), which can be combined with the embodiment in which G_UL is measured, the TO, Time Offset, of the uplink transmissions from the UE is measured, in order to arrive at timing advance commands to the UE which will minimize the TO. The timing advance commands are thus commands to the UE for a time offset which should be applied to the UE's uplink transmissions in order to minimize the TO in the eNodeB.

As will be described in more detail below, in one embodiment of the invention, which is indicated in step 235 ("Weight"), the measurements from the different channels are accorded different weights when the control commands are determined.

The weighting is suitably such that the different measurements are given different weights according to one or more of the following factors, as indicated in step 240:

β The bandwidth of the measurement, ("BW"), • The duration of the measurement, ("T"),

• The SINR, Signal to Interference Noise Ratio, of the measured channel, ("SINR"),

• The interval of time since the latest previous measurement in the channel, ("ΔT").

In a particular embodiment of the present invention, step 245, ("Filter"), the measurements are filtered in the eNodeB in order to arrive at the control commands which should be sent to the UE on which the measurements were made. In one embodiment of the "filter embodiment", shown in step 250, ("Kalman"), the filter which is used is a so called Kalman filter, which as such is a known type of filter, and which will thus only be described briefly later in this text.

As mentioned previously, the invention can be used in a number of different cellular systems, but in one embodiment, the invention is applied in an LTE system, as indicated in step 265. In such a case, i.e. in LTE applications, the two or more uplink channels on which measurements can be made include the following:

• PUSCH, Physical Uplink Shared Channel,

• PUCCH, Physical Uplink Control Channel, β SRS, Sounding Reference Signals.

In another embodiment, the invention can instead be applied in a WiMAX system, Worldwide Interoperability for Microwave Access, as indicated in step 270.

Returning now to the "filter embodiment", the measurements made on uplink transmission from a UE can, as mentioned above, in one embodiment be filtered in the eNodeB in order to arrive at the proper command parameters for the UE. The filter function which is used can be one of many known filter functions, but in a preferred embodiment, the filter which is used is a Kalman filter.

If a Kalman filter is used, the function will vary slightly depending on the measurement parameter which is filtered, G_UL or TO, or both. For this reason, a short description of a general Kalman filter will be given below, followed by a brief description for both possibilities, i.e. a Kalman filter for G_UL and a Kalman filter one for TO.

Kalman Filters

The Kalman filter can be used for dynamic systems such as the one for which the present invention is intended, and which can be expressed according to the following:

x_k = AX_k-i + Bu_k + w_k-1 (1)

z_k = Hx_k + v_k.i (2)

where:

x is the present state of the variable which is filtered,

A is a state transition model for x from one state to another in the model,

B is a control-input model, u is a control vector, w is the process noise with covariance Q, z is the observation of x,

H is the measurement model, v is the measurement noise with covariance R,

The variables x, u, w, z and v are in the general case vectors, while A, B and

H are matrices. The models A, B and H can vary over time, i.e. they can differ between state k and state k+1 , k+2....k+n.

The Kalman filter is suitably implemented as a predictor-corrector. First, the prediction is updated based on knowledge about the measured process, and then the result is corrected based on the measurement, as exemplified by the following:

Time update ("knowledge based prediction"):

Ξ_k = AX_k-1 + Bu_k π_k = AP_k-1A^T + Q

Measurement update ("correction"):

K_k = π_kH^τ(Hπ_kH^τ + R)^"1

X_k ⁼ Ξ_k + K_k(z_k - HΞ_k) p_k = (I - κ_kH)π_k

X_k is here the estimate of the process x, P_k is the error covariance. Ξ_k and π_k are called a priori estimates of X_k and P_k, respectively.

K_k is usually referred to as the Kalman gain. It can be noted that if the process is highly unpredictable, i.e. Q is large, then JT_k will also become large, and hence the Kalman gain becomes close to 1. In such cases, X is mainly based on the measurement z_k.

On the other hand, if the measurement uncertainty is very large (the extreme case is when no measurement is available), R then becomes very large, and the Kalman gain becomes close to zero. In this case, X is decided mainly by the prediction Ξ_k.

The Kalman filter described previously in this text is preferably used, with a parameter setting according to the following (all scalars):

x (the state) is GU_L-

A, the state transition model, is the channel gain variation model, basically consisting of path loss and fading. For simplicity, A could be set to 1.

B, the control-input model, is zero since we model G_UL as the state. If we would model PSD as the state, B would show how power control commands affect PSD. B would then typically be 1.

U, the control vector, is zero as well. If we would model PSD as the state, u would be the power control commands.

Q, the covariance of the process noise w, is the uncertainty in the fading model. Since we approximated the fading model A with 1 , Q has a very strong dependency towards the actual fading. For channels which vary slowly over time, Q is small, and for channels varying quickly over time, Q will be large.

Z, the observation of x, is the measurement of G_UL- The measurement of G_UL is typically implemented as the difference between the measured received PSD and the transmitted PSD. The transmitted PSD can be estimated from reports from the UE. H, the measurement model, is here set to 1 , which means that we use GUL and not PSD as z. The translation from PSD to GUL is then included in z instead of in H.

R, the covariance of the measurement noise v, depends strongly on the measurement bandwidth as well as the time duration of the reception which differs between SRS on one hand and PUSCH and PUCCH on the other hand. R also depends on the SINR of the signal. R is calculated for each measurement based on the mentioned parameters. The larger values of bandwidth, time duration and SINR, the smaller becomes R.

The Kalman filter then becomes:

Time update ("knowledge based prediction"): Ξk = GUL, k-1

FIK = Pk-i ⁺ Q(fading characteristics)

Measurement update ("correction"):

K_k = π_k(π_k + Rφandwidth, time duration, SINR))^'1

GUL, k - _"k ⁺ Kk(GuL measured, k - ^k) p_k = (i - κ_k)π_k

P could typically be initialized with infinity, meaning that the first measurement is used for initialization, and that initialization of G_UL before the first measurement becomes available does not matter.

It should be noted that k is stepped "forward" (incremented by one step) each subframe, regardless of the presence or absence of measurements. If there is no measurement at a time k, it can be modelled as a measurement with R set to infinity. From the formulae it can be noted that GU_L is not updated; however P is increased.

This means that the longer time since the last measurement, the larger weight a new measurement gets.

From an implementation point of view, the Kalman filter does not need to be run each subframe. It would be enough to run it when a measurement result becomes available, and then it is run N times if it were N subframes since the last measurement.

Kalman Filter for TO

The prediction of TO will suitably use PUSCH, PUCCH and SRS (if available).

The general Kalman filter described first above could suitably be used, with the following parameter settings (all scalars):

x (the state) is TO.

A (the state transition model) is the UE movement model. For simplicity A could be set to 1.

B (the control-input model) is set to 1.

u (the control vector) consists of time alignment commands. It should be noted that U_k is the time alignment command executed by the UE at time k, not the command transmitted by eNodeB at time k. Q (the covariance of the process noise w) is the uncertainty in the UE movement model. It should not be as sensitive for UE differences as the Q for GUL prediction.

z (the observation of x) is the measurement of TO.

H (the measurement model) is set to 1.

R (the covariance of the measurement noise v) strongly depends on the measurement bandwidth as well as the time duration of the reception which differs between SRS and PUSCH/PUCCH. R also depends on the SINR of the signal. It could be noted that SINR characteristics differ between the frequency-multiplexed PUSCH and the code-multiplexed PUCCH. R is calculated for each measurement based on the mentioned parameters. The larger values of bandwidth, time duration and SINR, the smaller becomes R.

The Kalman filter then becomes:

Time update ("predict"): Ξ_k = TO_k-₁ + TA_command_k

EEk = Pk- 1 + Q(UE movement characteristics)

Measurement update ("correct"):

K_k = π_k(π_k + R(bandwidth, time duration, S/Λ/f?))^"1 TOk ⁼ Ξk + Kk(TO_measured, k — Ξ_k) p_k = (1 - κ_k)π_k

As for G_UL, P could typically be initialized with infinity, and k is stepped up each subframe, regardless of presence of measurement. A difference compared to GU_L is however that TO can be updated due to a TA command and not only by new measurements. From an implementation point of view, the Kalman filter is typically run either when a measurement becomes available or when it is time for the UE to effectuate a TA command.

Fig 4 shows a schematic block diagram of a transceiver 400 for use as a controlling node in a system in which the invention is applied. As indicated in fig 4, the transceiver 400 will comprise an antenna, shown as block 410, and will also comprise a receive part 420 and a transmit part 430. In addition, the transceiver 400 also comprises a control means 440 such as a micro processor, as well as a memory 450. Furthermore, the transceiver 400 also comprises an interface 460 towards other components in the system apart from the UEs.

Since these means have now been introduced with respect to their function, they may in the following be referred to by reference number alone, e.g. "the means 410", by which will be meant the antenna 410.

As has emerged from the description above, the transceiver 400 is intended for use as a controlling node of a cell in a cellular communications system such as the system 100 of fig 1 , and uses the means 410, 420, 430, 440, and 450 for controlling traffic to and from users in the cell.

The transceiver 400 also uses the means 410 and 420 for receiving transmissions from users in the cell on a plurality of uplink channels, and uses the means 410 and 430 for transmitting control commands to the users on downlink control channels, said control commands being for parameters in the user's uplink transmissions.

According to the invention, the transceiver uses the means 440, 450 for measuring one and the same parameter in uplink transmissions from a user in both at least a first and a second uplink channel, and also uses the means 440 and 450 for combining these measurements in order to determine one or more of said control commands for the user in question. In one embodiment, the transceiver 400 additionally uses the 440 and 450 for also measuring a second parameter in uplink transmissions from a user in both at least the first and second uplink channels, and also uses the means 440 and 450 for combining these measurements in order to determine at least a second control command for the user in question.

In one version of an embodiment, the transceiver 400 uses the means 440 and 450 for giving different weights to the measurements from the different channels when said control commands are determined.

In one aspect of the invention, the transceiver 400 gives the measurements different weights according to one or more of the following factors:

» The bandwidth of the measurement, » The duration of the measurement, β The SINR, Signal to Interference Noise Ratio, of the measured channel, • The interval of time since the latest previous measurement in the channel.

In one embodiment, the transceiver filters the measurements in order to arrive at the control commands, and as one option, this filtering is carried out in a Kalman filter.

Suitably, one of the measured parameters is the uplink gain, G_UL, of the transmissions from the user. Alternatively, or as a complement, one of the measured parameters is a time offset, TO, of transmissions from the user, i.e. the discrepancy between the expected and actual arrival time of the transmissions from the user. In one embodiment, the transceiver 400 is an eNodeB for an LTE system, Long Term Evolution. In one version of this embodiment, the uplink channels on which measurements are made include two of the following:

• PUSCH, Physical Uplink Shared Channel, • PUCCH, Physical Uplink Control Channel,

• SRS, Sounding Reference Signals.

In another embodiment, the transceiver 400 is a controlling node for a WiMAX system, Worldwide Interoperability for Microwave Access.

The invention is not limited to the examples of embodiments described above and shown in the drawings, but may be freely varied within the scope of the appended claims.

Claims

1. A method (200) for use in a cellular communications system (100) in which a cell (110) in the system is associated with a controlling node (130) which serves to control traffic to and from users (120) in the cell, according to which method users in the cell are enabled to make transmissions (205) to the controlling node on a plurality of uplink channels, and according to which method the controlling node transmits control commands (210) to the users on downlink control channels, said control commands being for parameters in the user's uplink transmissions, the method being characterized in that the controlling node measures (215) one and the same parameter in uplink transmissions from a user in both at least a first and a second uplink channel, and combines (255) these measurements in order to determine (260) one or more of said control commands for the user in question.

2. The method (200) of claim 1 , according to which the controlling node also measures a second parameter (220) in uplink transmissions from said user (120) in both at least the first and second uplink channels and combines these measurements in order to determine at least a second control command for the user in question.

3. The method (200) of claim 1 or 2, according to which the measurements from the different channels are weighted (235) differently when said control commands are determined.

4. The method (200) of claim 3, according to which the measurements are given different weights (240) according to one or more of the following factors:

» The bandwidth of the measurement, β The duration of the measurement, • The SINR, Signal to Interference Noise Ratio, of the measured channel,

• The interval of time since the latest previous measurement in the channel.

5. The method (200) of any of the previous claims, according to which the measurements are filtered (245) in order to arrive at said control commands.

6. The method (200) of claim 5, according to which said filtering is carried out in a Kalman filter (250).

7. The method (200) of any of the previous claims, according to which one of the measured parameters (230) is the uplink gain, G_UL, of the transmissions from the user (120).

8. The method (200) of any of the previous claims, according to which one of the measured parameters (225) is a time offset, TO, of transmissions from the user, i.e. the discrepancy between the expected and actual arrival time of the transmissions from the user (120).

9. The method (200) of any of the previous claims, as applied (265) to an LTE system, Long Term Evolution.

10. The method of claim 9, according to which the uplink channels on which measurements are made include two of the following:

• PUSCH, Physical Uplink Shared Channel,

• PUCCH, Physical Uplink Control Channel, β SRS, Sounding Reference Signals.

11. The method (200) of any of claims 1-8, as applied (270) to a WiMAX system, Worldwide Interoperability for Microwave Access.

12. A transceiver (400) for use as a controlling node (130) of a cell (110) in a cellular communications system (100), being equipped with means (410, 420, 430, 440, 450) for controlling traffic to and from users (120) in the cell (100) and with means (410, 420) for receiving transmissions from users in the cell on a plurality of uplink channels, and with means (410, 430) for transmitting control commands to the users on downlink control channels, said control commands being for parameters in the user's uplink transmissions, the transceiver (400) being characterized in that it is equipped with means (440, 450) for measuring one and the same parameter in uplink transmissions from a user in both at least a first and a second uplink channel, and means (440, 450) for combining these measurements in order to determine one or more of said control commands for the user in question.

13. The transceiver (400) of claim 12, additionally being equipped with means (440, 450) for also measuring a second parameter in uplink transmissions from said user (120) in both at least the first and second uplink channels and means (440, 450) for combining these measurements in order to determine at least a second control command for the user in question.

14. The transceiver (400) of claim 12 or 13, being equipped with means (440 450) for giving different weights to the measurements from the different channels when said control commands are determined.

15. The transceiver (400) of claim 14, in which the measurements are given different weights according to one or more of the following factors:

« The bandwidth of the measurement,

• The duration of the measurement, β The SINR, Signal to Interference Noise Ratio, of the measured channel, • The interval of time since the latest previous measurement in the channel.

16. The transceiver (400) of any of claims 12-15, in which the measurements are filtered in order to arrive at said control commands.

17. The transceiver (400) of claim 16, in which said filtering is carried out in a Kalman filter.

18. The transceiver (400) of any of claims 12-17, in which one of the measured parameters is the uplink gain, G_UL, of the transmissions from the user (120).

19. The transceiver (400) of any of claims 12-18, in which one of the measured parameters is a time offset, TO, of transmissions from the user, i.e. the discrepancy between the expected and actual arrival time of the transmissions from the user (120).

20. The transceiver (400) of any of claims 12-19, being an eNodeB for an LTE system, Long Term Evolution.

21. The transceiver (400) of claim 20, in which the uplink channels on which measurements are made include two of the following:

» PUSCH, Physical Uplink Shared Channel, * PUCCH, Physical Uplink Control Channel,

• SRS, Sounding Reference Signals.

22. The transceiver (400) of any of claims 12-19, being a controlling node for a WiMAX system, Worldwide Interoperability for Microwave Access.