GB2508579A

GB2508579A - Inter-cell interference cancellation

Info

Publication number: GB2508579A
Application number: GB201216346A
Authority: GB
Inventors: Yuan Zhang; li-ke Huang
Original assignee: Aeroflex Ltd
Current assignee: Aeroflex Ltd
Priority date: 2012-09-13
Filing date: 2012-09-13
Publication date: 2014-06-11
Anticipated expiration: 2032-09-13
Also published as: GB2508579B; GB201216346D0

Abstract

A receiver (60) in a cellular communication system receives a signal which includes interference from a neighbouring cell. The receiver generates an estimate for a propagation channel between the receiver and a transmitter of an interfering cell using a moving average (84) of de-patterned pilot symbols (82) in the signal. The receiver then performs a cell search to obtain the cell ID of the interfering cell, and uses the power of the channel estimate (86) to obtain estimates for the bandwidth and number of antennas used by the interfering cell. Interfering pilot symbol estimates are then generated from e.g. a look up table based on cells ID, bandwidth and number of antennas used (88,90). An interference estimation is then generated (92) by multiplying the channel estimate by the pilot estimates, and the resulting estimate is subtracted from the received signal.

Description

INTERFERENCE CANCELLATION

Technical Field

The present application relates to a method for cancelling interference in a signal received by a receiver, and to a receiver which implements such a method. In particular, the present application relates to a method for cancelling inter-ccfl interference in a received orthogonal frequency division multiplexed OFDM) signal.

Background to the Invention

In order to support higher data rates in mobile telecommunications networks, the third generation partnership project (3GPP) introduced a new air interface based on orthogonal frequency division multiple access (OFDMA) techniques as the long term evolution (LTE) of the IJIIVITS network.

One problem that arises in LTE cellular networks is inter-cell interference. This problem is particularly pronounced when a receiver (or "User Equipment" (UE)) is located at or near a cell boundary. An example of such as situation is shown generally at 10 in the schematic illustration of Figure 1, which shows part of a cellular communications system. In thc illustrated cxamplc, a UE 12 rcccivcs signals from a transmitter or "Node B" 14 (identified as eNB 1 in Figure 1) serving the cell in which the TiE 12 is located. However, the TiE 12 may also receive signals intended for other liEs from Node Bs 16, 18 (identified as eNB2 and eNB3 in Figure 1) serving nearby cells, which may interfere with the desired signal from the Node B 14 serving the cell in which the UE 12 is located, leading to reduced reception quality at the TiE 12.

Efforts have been made to address this issue. In one scheme, adjacent Node Bs communicate with each other to schedule data transmissions, such that when one is transmitting data to a target TiE its adjacent Node Bs, which could interfere with the reception of the data by the target IJE, cease transmitting data. Whilst this scheme can bc effcctive, it docs not addrcss a problem of inter-cell intcrfcrcncc arising from transmission of a pilot signal by Node Bs adjacent to the Node B serving the cell in which the receiving UE is located, since transmission of the pilot signal cannot be interrupted.

Summary of Invention

The present application relates to a method of cancelling inter-cell interference in cellular networks.

According to a first aspect of the present invention, there is provided a receiver for a cellular communication system, the receiver being configured to: generate a channel cstimatc for a propagation channel bctween a transmitter of an interfering cell and the rccciver; reccivc information relating to operating parameters of the transmitter of the interfering cell; generate, using the information so received, a pilot symbol estimate for a pilot symbol transmitted by the transmitter of the interfering cell; calculate an estimated interference symbol by multiplying the pilot symbol estimate by the channel estimate; and subtract the estimated interference symbol from a signal received at the receiver.

The receiver may comprise art interference canceflation unit which comprises: a de-patteming unit configured to generate dc-patterned pilot symbols from a signal received by the receiver; and a moving average filter configured to perform a moving average filtering operation on the dc-patterned pilot symbols.

The receiver may be configured to perform a cell search to detect an identifier of the transmitter of the interfering cell.

The receiver may be further configured to estimate a bandwidth and a number of antennas of the transmitter.

The receiver may further comprise a power calculation unit, the power calculation unit bcing configured to estimate thc bandwidth of thc transmitter by calculating a power value of the channel estimate.

The receiver may further comprise a processor configured to generate log likelihood ratios (LLRs) for data symbols contained in thc received signal and to scaling the value of the LLRs so generated.

The processor may bc configured to reduce thc LLR values by a constant scaling factor.

According to a second aspect of the invention there is provided a method for cancellation of adjacent cell interference in a signal received by a receiver of a cellular communication system, the method comprising: generating a channel estimate for a propagation channel between a transmitter of an interfering cell and the receiver; receiving information relating to operating parameters of the transmitter of the interfering cell; generating, using the information so received, a pilot symbol estimate for a pilot symbol transmitted by the transmitter of the interfering cell; calculating an estimated interference symbol by multiplying the pilot symbol estimate by the channel estimate; and subtracting the estimated interference symbol from a signal received at the receiver.

Generating the channel estimate for the propagation channel may comprise generating dc-patterned pilot symbols from the received signals and performing a moving average filtering operation on the dc-patterned pilot symbols.

Receiving information relation to operating parameters of the transmitter may comprise performing a cell search to detect an identifier of the transmitter of the interfering cell.

Receiving information relating to operating parameters of the transmitter may further comprise estimating a bandwidth and a number of antennas of the transmitter.

Estimating the bandwidth of the transmitter may comprise calculating a power value of thc channcl cstimatc.

The method may ftirther comprise generating log likelihood ratios (LLRs for data symbols contained in the received signal and scaling the value of the LLRs so generated.

The LLR values may be reduced by a constant scaling factor.

Brief Description of the Drawings

Embodiments of the invention will now be described, strictly by way of example only, with reference to the accompanying drawings, of which: Figurc I is a schcmatic reprcsentation of part of a cellular nctwork in which a FE is locatcd in a scrving cdl of thc ccllular nctwork and is subjcct to intcr-ccll intcrfcrcncc from adjacent cells of the cellular network; Figure 2 shows an exemplary OFDM resource grid for a receiver in a cellular communications network in which there is one serving cell and one interfering cell; Figurc 3 is a schcmatic rcprescntation of an architecturc for a rccciver; and Figure 4 is a schematic representation of an architecture for a symbol rate interference cancellation unit of the receiver architecture illustrated in Figure 3.

Description of the Embodiments

Referring first to Figure 2, an OFDM resource grid for a receiver in a cellular communications network in which there is one serving cell and one interfering cell is shown gcncrally at 20. In this case both the serving cell and the adjacent cell are served by transmitters of the single input single output (SISO) type. In Figure 2 it is assumed that there is no frequency offset between the serving cell and the interfering ccli, and that any timc offsct between thc scrving ccli and thc interfcring ccli is within the cyclic prefix (CP) length of a transmitted pilot signal. Thus, a pilot symbol Ro contained within a pilot signal transmitted by interfering cell appears in a resource element (RE) of the receiver resource grid intended for data transmitted by the serving ccli, causing intcrfcrcncc (shown with a dottcd background in Figure 2).

A signal model of a resource element R0 at a pilot position in the resource grid of the serving ccli can be written as k@) = (1). pl)(/) + flA-(T)' (1) where H' (i) and p (1) denote, respectively, the channel coefficients and pilot symbol of the serving cell (referred to as Cell 1) at the kth OFDM subcarrier, and flAW denotes noise on the kth OFDM subcarrier. This equation apphes for singie input multiple output (SIMO) and MIMO cases, since the pilot REs are orthogonal among the antennae of a SIMO transmitter. As this equation reflects oniy the pilot position of the serving cell, there is no interference term, and thus channel estimates of the serving cell based on the pilot signal may not reflect the actual level of interference at the data positions of a received signal.

A signal model of a resource element at a data position in the resource grid of the serving cell which is affected by inter-cell interference from a pilot signal of an adjacent cell can be written, for the SISO case, as (I) . (U. (fl. 1') xk(1) = hk @) (1) + h; () + where h' (i) denotes the channel coefficients of the serving cell (Cell 1) experienced by the kth subcarrier of a data signal transmitted by the transmitter of the serving cell, c/' (1) denotes the data symbol on the kth subcarrier of the data signal, h2 (i) denotes the channcl coefficients of the interfcring ceil (Cell 2) expericnccd by the kth siThearrier of an interfering pilot signal transmitted by the tratismitter of the interfering cell, and 2) (1) denotes the pilot symbol on the kth subcarrier of the interfering pilot signal.

For a SIMO case, a signal model of a resource element at a data position in thc resource grid of the sewing cell which is affected by inter-cell interference can be written as rxi (i) = h' 1. d1) (0 + phc) (1) ,(2) (0 + r (2) Lx2 (1)_i h' (0J Lh2 (0 L2 (0J or x(i) = H'0ht'(i) + H20)p2kJ) + whilst for a 2x2 MIMO case the signal of the resource element at a data position of the serving cell which is affected by inter-cell interference can be written as - (0 k (01. H' (0 (01. d' (01 + rh2 (01. 20) + HtO)1 (3) Lx20)1 -h? 0 hiY @H LwY 0 wi (01 d' (01 Lh2) (01 H2 (01' or x(i) = It'(i)sV(i) J1(i -F I12kj)p2Ø) + n(i).

Comparing equation (2) with equation (3), it can be seen that for the SIMO ease and the MIMO case the interference term H(2)(/)p(2)) is the same. This is because the pilot signal is allocated orthogonally among antennae. For one RE of the data signal received by a receiver such as a UE in the serving cell, only the pilot symbol from a single antenna of a transmitter such as a Node B in an interfering cell will interfere.

However, the number of interfering REs will increase, since the number of pilot symbols transmitted in the 2x2 MIMO case is twice that of the SIMO case, resuhing in more severe inter-cell interference.

Accordingly, there is a need to reduce the effect of inter-cell interference arising from pilot signals transmitted by transmitters in adjacent cells.

Figure 3 is a schematic representation of a receiver architecture which can be used in a receiver of a cellular communications system to reduce the effects of this inter-cell interference. It will be appreciated that the schematic representation of Figure 6 presents the architecture as a series of functional blocks, but that the functional blocks do not necessarily represent physical components of a "real world" implementation of the architecture, but are instead intended to represent processing operations undergone by a received signal.

The receiver architecture is shown generally at 60 in Figure 6, and includes parallel first and second fast Fourier transform (FFT) units 62, 64, which are operative to receive encoded data from upstream components of the receiver and to reverse an inverse fast Fourier transform (lEFT) process which takes place in a transmitter transmitting pilot or data symbols to the receiver. Outputs of the FFT units 62, 64 are input to a symbol rate processor 66, which performs symbol-rate processing on symbols received by the architecture 60. Outputs of the symbol rate processor 66 are connected to inputs of a bit rate processor 68, which is configured to perform bit-rate processing on data bits output by the symbol rate processor 66.

The receiver architecture 60 also includes a symbol rate interference cancellation unit and a bit rate interference cancellation unit 72. The operation of the symbol rate interference cancellation unit 70 and the bit rate interference cancellation unit is controlled by an interference cancellation control unit 74.

The symbol rate interference cancellation unit 70 is configured to estimate the interference experienced by the receiver in the sewing cell as a result of the transmission of the pilot signal by a transmitter of a neighbouring cell, and to subtract the resulting estimate of the interference from the signal received at the receiver in the sewing cell, so as to at least partially cancel the interference.

From equations (2) and (3) above, it can be seen that the interference term in the signal model of a resource element affected by inter-cell interference is fi(2)(f)p(2)(j), where (2) denotes the pilot symbol of the interfering pilot signal. Thus, it is this interference term that is estimated by the symbol rate interference cancellation unit 70 and subtractcd form the received signal.

The symbol rate interference cancellation unit 70 estimates the propagation channel H2experienced by the interfering pilot symbols, and, using the estimated propagation channel, generates an estimate of the interference, which is subsequently subtracted from the received signal to at least partially cancel the interference in the received signal.

Applying zero-forcing dc-patterning to equation (2) gives IT2(/)= ? =H2o+H°o. dm0) + (4) r 0) r 0) p -(1) whcre 11(2)0) is a vector of de-patterncd pilot symbols. In equation (4) the data RE of the serving cell represented by the term IJW)0). (2)' becomes interference.

p -(/) Usually the power of a data RE is relatively high, so the estimation in equation (4) cannot be very accurate. The expectation of equation (4) can be expressed as EH(2)(1))= E11(2)(i) +H'0). do) + p -0) p 0)) d =EH(2)(t))= The terms d(i) 9 and approximate to zero, as it can be assumed that p(i) p(i)) thc expectation of a data symbol in the serving cell is zero. This is because di?,) is a complex symbol with QPSK, I6QAM or 64QAM modulation, and it can be assumed that the different modulation points have an equal probability, since the whole constellation is systematic. The expectation of a data symbol is therefore zero, ,and so E(H (2) (1)) E(H (1) (0) (5) Equation (5) gives a good estimate of the mean value of the inference channel 11(2) experienced by the interfering pilot symbols.

Equation (5) is equivalent to using a moving average filter in the time domain on the dc-patterned pilot symbols to generate an estimate of the mean value of the inference channel 11(2). If the length of the moving average filter is within the coherence time of the channel, Ht2 (/) can be approximated by EH(2) (0). Thus, the procedure undertaken by the symbol rate interference cancellation unit 70 to estimate the interference channel H2 experienced by the interfering pilot symbols can be expressed as illustrated in Figure 7.

As can be seen, following the FFT operation performed by FFT unit 62 or 64, the received signal is passed to a dc-patterning unit 82 of the symbol rate interference cancellation unit 70 to generate the dc-patterned pilot symbols H2 (/). The de-patterned pilot symbols are passed to a moving average filter 84, which outputs E(HW). which is used by the symbol rate interference cancellation unit 70 as the estimate of the interference channel H2 experienced by the interfering pilot symbols.

In order to estimate the interference caused by the transmitted pilot symbols of the interfering cell, those pilot symbols must also be estimated by the symbol rate interference cancellation unit 70.

The pilot symbols transmitted by the interfering transmitter are predefined, based upon operating parameters of the transmitter, including a cell identifier (ID), its bandwidth and number of antennas. The symbol rate interference cancellation unit 70 estimates the pilot symbols pt2) in a manner that will be described below.

The receiver performs a cell search for nearby cells, and from this the cell ID of the interfering transmitter is known. This cell ID is passed to the symbol rate interference cancellation unit 70. The symbol rate interference cancellation unit 70 estimates the bandwidth and number of antennas of the interfering transmitter, as will be described below, and uses the resulting estimates of the bandwidth and number of antennas to estimate the pilot symbols pO) of the interfering pilot signal.

In one embodiment, the symbol rate interference cancellation unit 70 sets the bandwidth and number of antennas of thc interfering transmitter to a maximum possible value. For example, the bandwidth may be set to a value of 20 MHz, while the number of antennas may be set to 4. The channel estimates described above are performed using these parameters. If the actual bandwidth of the interfering transmitter is smaller than the maximum (i.e. 20 MHz in this example), the power of pilot symbols outside the signal band will be zero.

This can be expressed mathematically by re-writing equation (5), bearing in mind that H'2(i) = 0 for pilot symbols outside the actual bandwidth. Thus EH(2)(i))= EH°(i) +HO) + As before, the tens 4 cI'(i) 9 and 9 approximate to zero, as it can be 7) (/) p1 (1) assumed that the expectation of a data symbol in the serving cell is zero, for the reason set out above. Thus 0 (6) This suggests that the power of the channel estimates, i.e. E(H12) (i) can be used to detect the signal bandwidth. Thus, as can be seen from Figure 7, a power calculation unit 86 of the symbol rate interference cancellation unit 70 calculates the power of the channel estimates previously calculated.

A similar approach can be applied to antenna number estimation in an antenna number estimation unit 88 of the symbol rate interference cancellation unit 70.

When the antenna number and signal bandwidth have been estimated, the symbol rate interference cancellation unit 70 uses the antenna number and bandwidth estimates to generate pilot symbol estimates. This may be achieved, for example, by inputting the antenna number and bandwidth estimates into a pilot symbol determination unit 90, which may be, for example, a look-up table containing predcfined pilot sequences for particular bandwidth and antenna number values.

Once the symbol rate interference cancellation unit 70 has estimated the interference channel H2 and the pilot symbols p2a,), the interference term can be estimated by an interference estimation unit 92 of the symbol rate interference cancellation unit 70, which multiplies the interference channel H2 by the pilot symbol estimates (2)(/) to simulate the effect of the interference channel on the transmitted pilot symbols. The resulting interference symbol estimates are subtracted from the received signal on a per antenna element basis, prior to data demodulation, to at least partially cancel interference in the data signal received from the transmitter of the cell serving the receiver that is caused by the pilot signal transmitted by a transmitter of a neighbouring cell.

In addition to the symbol rate interference cancellation unit 70, the receiver architecture 60 also includes a bit rate interference cancellation unit 72. The symbol rate interference cancellation process described above will likely not remove all interference from the received signal. Thus, the bit rate interference cancellation unit is operative to apply a scaling factor to log likelihood ratios (LLRs) generated by the symbol rate processor 66 that are affected by interference caused by the pilot signal of the transmitter of an adjacent cell. The individual LLRs that are affected by the interference will be known, as the REs in a given resource block that are affected by interference can be determined once the cell ID, bandwidth and number of antennae of the interfering transmitter arc known, and the bit rate interference cancellation unit 72 is configured to scale the value of these affected LLRs by a constant scaling factor that is derived from the channel estimate used by the symbol rate interference calculation unit.

It will be appreciated that in some circumstances symbol rate interference cancellation or bit ratc interfcrence cancellation will not bc requircd. Thus, thc operation of thc symbol rate interference cancellation unit 70 and the bit rate interference cancellation unit 72 is controlled by the interference cancellation control unit 74. However, the operation of the interference cancellation control unit 74 is beyond the scope of this document, and thus will not be described in detail here.

In the description above, it has been assumed that the serving cell and the interfering cell are time aligned, or that any small timing offset that exists between the serving cell and the interfering cell is smaller than the cyclic prefix length of a transmitted pilot signal, such that the interference from the pilot signal of the interfering cell is aligned with a data RE in the serving cell, as shown in Figure 2. This is a valid assumption for many deployment scenarios.

It will be appreciated that the method and receiver architecture described above provide an efficient and highly effective way of cancelling, at least partially, inter-cell interference in a signal received by a receiver in a serving cell arising from the transmission of pilot signals by transmitters in neighbouring cells.

Claims

CLAIMS1. A receiver for a cellular communication system, the receiver being configured to: generate a channel estimate for a propagation channel between a transmitter of an interfering cell and the receiver; receive information relating to operating parameters of the transmitter of the interfering cell; generate, using the information so received, a pilot symbol estimate for a pilot symbol transmitted by the transmitter of the interfering cell; calculate an estimated interference symbol by multiplying the pilot symbol estimate by the channel estimate; and subtract the estimated interference symbol from a signal received at the receiver.
2. A receiver according to claim I comprising an interference cancellation unit which comprises: a de-patteming unit configured to generate dc-patterned pilot symbols from a signal received by the receiver; and a moving average filter configured to perform a moving average filtering operation on the dc-patterned pilot symbols.
3. A receiver according to claim 1 or claim 2 wherein the receiver is configured to perform a cell search to detect an identifier of the transmitter of the interfering cell.
4. A receiver according to claim 3 wherein the receiver is further configured to estimate a bandwidth and a number of antennas of the transmitter.
5. A receiver according to claim 4 further comprising a power calculation unit, the power calculation unit being configured to estimate the bandwidth of the transmitter by calculating a power value of the channel estimate.
6. A receiver according to any one of the preceding claims wherein the receiver further comprises a processor configured to generate log likelihood ratios (LLR5) for data symbols contained in the received signal and to scaling the value of the LLRs so generated.
7. A receiver according to claim 6 wherein the processor is configured to reduce the LLR values by a constant scaling factor.
8. A method for cancellation of adjacent cell interference in a signal received by a receiver of a cellular communication system, the method comprising: generating a channel estimate for a propagation channel between a transmitter of an interfering cell and the receiver; receiving information relating to operating parameters of the transmitter of the interfering cell; generating, using the information so received, a pilot symbol estimate for a pilot symbol transmitted by the transmitter of the interfering cell; calculating an estimated interference symbol by multiplying the pilot symbol estimate by the channel estimate; and subtracting the estimated interference symbol from a signal received at the receiver.
9. A method according to claim 8 wherein generating the channel estimate for the propagation channel comprises generating dc-patterned pilot symbols from the received signals and performing a moving average filtering operation on the de-patterned pilot symbols.
10. A method according to claim 8 or claim 9 wherein receiving information relation to operating parameters of the transmitter comprises performing a cell search to detect an identifier of the transmitter of the interfering cell.
11. A method according to claim 10 wherein receiving information relating to operating parameters of the transmitter further comprises estimating a bandwidth and a number of antennas of the transmitter.
12. A method according to claim 11 wherein estimating the bandwidth of the transmitter comprises calculating a power value of the chaimel estimate.
13. A method according to any one of the preceding claims further comprising generating log likelihood ratios (LLRs) for data symbols contained in the received signal and scaling the value of the LLRs so generated.
14. A method according to claim 13 wherein the LLR values are reduced by a constant scaling factor.