GB2454513A

GB2454513A - Interference avoidance

Info

Publication number: GB2454513A
Application number: GB0722066A
Authority: GB
Inventors: Yue Wang
Original assignee: Toshiba Research Europe Ltd
Current assignee: Toshiba Europe Ltd
Priority date: 2007-11-09
Filing date: 2007-11-09
Publication date: 2009-05-13
Anticipated expiration: 2027-11-09
Also published as: GB2454513B; GB0722066D0

Abstract

The disclosure concerns shaping a signal intended for transmission in a spectral band in which are defined a plurality of discrete transmission subcarriers. The shaping is intended to define a transmission suppressing notch within a particular portion of the spectral band. For upper and lower edge subcarriers of the intended notch, at least one adjacent subcarrier is identified, the adjacent subcarriers being immediately adjacent each edge subcarrier of the notch, and the intended signal for transmission at the identified adjacent subcarriers is diminished. Notch subcarriers are identified, which are contained within the notch and the signals intended for transmission on those contained subcarriers are suppressed. Interference reducing signals are determined and applied at the edge subcarriers and the subcarriers contained in the notch. The invention is applicable to ultra wideband (UWB) communications interfering with licensed narrowband devices.

Description

Interference Avoidance The present invention relates to a method of, and apparatus for interference avoidance particularly, but not exclusively in the context of wideband communication. The invention has particular application in the field of ultra wideband (UWB) communications.

Interference avoidance may be necessary in circumstances where both wideband and narrowband devices operate in the same frequency range. Figure 1 is an illustration of the overlap between narrowband and wideband signals in such circumstances.

The use of wideband communication could have an impact on operation of narrowband devices within the spectral band used by a wideband device. In essence, the wideband device will, unless avoiding action is taken, cause interference to such narrowband signals. It is known to avoid a wideband device causing interference to a narrowband device by transmitting little or no energy from the wideband device at the frequency bands of the licensed narrowband device. This is achieved by generating a spectral notch at those frequency bands in the wideband signal. Here, the term "wideband" is used in its broadest sense and is not intended to imply conformance to any particular technical standard.

Interference avoidance in such circumstances is important as the user of the narrowband channel is doing so, in many circumstances and jurisdictions, on the basis of an official licence issued by a regulator and so should reasonably expect to be given access to the channel in accordance with that licence and without interference from a wideband transmission. In contrast, many implementations of wideband transmission do so on the basis of taking advantage of the possibility offered by wideband to establish a channel across a wide spectral band, not on the basis of concentrated "spikes" of power in the spectrum but rather on the basis of transmission power being at a low level but spread across a wide spectrum. Such wideband activity can be unlicensed, and often is.

Figure 2 is an illustration of narrowband interference avoidance, in which the power levels of the wideband signal have been reduced to avoid interference in the overlap region with the narrowband signal.

It is noted that the spectrum overlay such as that shown in Figure 1 causes mutual interference to both the wideband (and possibly unlicensed) device and the narrowband licensed device. In the following, the operating frequency band or bands of such narrowband devices are referred to as interference bands. The wideband signal in the operating band of one or more narrowband devices may be suppressed so that the wideband system does not cause significant interference to those devices, and that may be referred to as interference avoidance.

The issue of avoidance of interference with licensed narrowband devices is particularly important for UWB systems. UWB systems occupy an unlicensed bandwidth of over 7.5GHz, which overlays with the spectrum of existing licensed narrowband devices.

Transmission of UWB signals over the licensed frequency band causes interference to the existing licensed narrowband systems. To avoid interference from UWB to such licensed devices, the maximum power of transmitted UWB signals has been limited by the Federal Communications Commission (FCC) in the United States to an extremely low value of -41.3 dBm/MHz. The consequence of this is to impose a rule that the power of the transmitted UWB signal cannot exceed this regulated value at any frequency.

In addition to power regulation requirements, it has also been proposed that when a UWB device detects, or is otherwise aware, that a narrowband licensed device is active in its operating frequency band, priority should be given to the licensed device, and the power of the transmitted UWB signals in such frequency bands should back off to an extent such that little or no energy is transmitted on those frequencies where the narrowband user's signal resides. The power back off for the purpose of interference avoidance can be achieved by generating spectral notches at those frequencies.

Spectral notching using active interference cancellation (AIC) has been proposed for orthogonal frequency division multiplexing (OFDM) transmissions in wideband systems, such as UWB systems, with single input and single output (SISO) antennas in order to avoid causing interference to narrowband signals (H. Yamaguchi, 34th European Microwave Conference, vol. 2, 1105-1108, 2004, and US Patent Application US2006/0080 16). In that context, the term "interference" may refer to or include the additional signal energy (residual interference) from data subcarriers of the wideband signal when the wideband signal is nulled within the narrowband interference band.

The term interference cancellation' may refer to the attempted suppression of any additional signal energy (the "interference") that resides in the desired notch position in the up-sampled spectrum.

In such systems, an AIC algorithm may be used to generate a spectral notch in the desired narrowband interference band by providing AIC tones including additional subcarriers, also referred to as AIC edge tones, at or beyond each edge of the interference band and substituting the data subcarriers at the corresponding frequency subcarriers by the designed AIC tones, thus avoiding causing interference to the 3rd party devices. Generally the AIC tones within, and at or beyond the edge of, the interference band are designed for interference cancellation. Such design or conditioning of the tones in the region of the intended region for interference cancellation can take the form of firstly nulling of tones to achieve residual interference, and then imposition of AIC tones designed to cancel the residual interference. A schematic illustration of AIC tones for producing a frequency notch in an OFDM scheme is provided in Figure 3. The AIC tones shown in Figure 3 comprise nulled AIC tones 2, 4, 6 within the interference band and AIC edge tones 8, 10 at or beyond the edge of the interference band. It is known to obtain the values of the AIC tones used to cancel out the residual interference from data subcarriers in the desired frequency positions in such systems either by performing a least square (LS) calculation (H.

Yamaguchi, 34th European Microwave Conference, vol. 2, 1105-1108, 2004) or a regularized least square calculation (US 2006/008016).

As was mentioned in US 2006/008016, there are instances of average excess power of up to 4.0 dB above the data subcarriers in the up-sampled spectrum when the LS solution of the AIC tones is applied. This causes a problem because the transmit power of the UWB signals must not exceed the power limit of -41.3 dBm/MHz. In order to avoid violation of the UWB power regulation, the transmit power of the transmitted data tones must be lowered by the amount of the excess power, causing excessive performance degradation.

In order to avoid the excess power problem caused by the LS solution, a method of suppressing the power on the AIC tones has been proposed in US2006/0080 16, which will be referred to as the RLS solution in the following description. By assigning a small regularisation factor which tunes the weights of the effort between maximising the notch depth (notch depth is defined as the average/peak power of the resulting spectra within the interference band relative to the data subcarriers) and minimising the power on the AIC tones, the RLS method suppresses the excess power of the AIC tones at the expense of an elevated spectrum notch. Therefore, the resulting notch is generally shallower than that generated by using the LS method, especially when a relatively large regularisation factor is used. Another problem of the RLS solution is that the regularisation factor is determined in an ad-hoc manner and an arbitrarily chosen regularisation factor might result in an inferior performance in terms of notch depth or the excess power. As is shown in Figure 7, a small change in the regularisation factor can cause more than 5dB enhancement of the peak power of the notch. This is undesirable in that it inhibits the effectiveness of the process to properly deliver interference cancellation.

It is also possible to limit the power of the AIC tones after they are computed using the LS or RLS solution. However, the resulting notch is then elevated to higher than -30 dB, as illustrated in Figure 4, as compared to the original notch of deeper than -40 dB when the LS solution is used and deeper than -30 dB when the RLS solution is used.

According to one aspect of the invention, there is provided a method of shaping a signal intended for transmission in a spectral band in which are defined a plurality of discrete transmission subcarriers, said shaping being intended to define a transmission suppressing notch within a particular portion of said spectral band, the method comprising identifying notch subcarriers contained within the portion of said intended notch, edge subcarriers positioned on either side of said notch subcarriers and, for each edge subcarrier, at least one adjacent subcarrier immediately adjacent said edge subcarrier and away from said intended notch position, diminishing the intended signal for transmission at said identified adjacent subcarriers, suppressing intended signals for transmission on said notch subcarriers, and determining and applying interference-reducing signals to be applied at said edge subcarriers and notch subcarriers.

Another aspect of the invention provides apparatus for processing a signal intended for transmission in a spectral band in which are defined a plurality of discrete transmission subcarriers, said apparatus being operable to shape said signal to define a transmission suppressing notch within a particular portion of said spectral band, the apparatus comprising means for diminishing a signal element intended for transmission at subcarriers adjacent edge subcarriers bounding said notch, means for identifying such notch subcarriers as are contained within said notch and for suppressing intended signals for transmission on said contained subcarriers, and means for determining and applying an interference-reducing signal to be applied at each of the contained notch subcarriers and the edge subcarriers.

The invention may have particular application in any wireless communication device that uses OFDM (one antenna or multiple antennas), where interference avoidance is desired. Examples include UWB-equipped PDAs, cameras and laptops.

Any feature in one aspect of the invention may be applied to other aspects of the invention, in any appropriate combination. In particular, apparatus features may be applied to method features and vice versa. In addition, computer program means could be provided to configure a general purpose computer in accordance with the invention.

Computer program means could be introduced by storage medium or by signal, such as download.

Embodiments of the invention will now be described, by way of example only, with reference to the accompanying drawings in which: Figure 1 is an illustration of narrowband and wideband signals occupying overlapping bandwidths; Figure 2 is an illustration of interference avoidance, in which the power levels of a wideband signal are reduced in the operating frequency of a narrowband signal; Figure 3 is an illustration of a frequency notch and associated AIC subcarriers in an OFDM structure; Figure 4 is a graph of notch depth as a function of frequency for notches generated using known AIC algorithms in which the power on the edge AIC tones is limited to the power of data subcarriers; Figure 5 is a schematic diagram of a transmitter and a receiver in combination, in accordance with a specific embodiment of the invention; Figure 6 is a flow diagram of a process in accordance with a specific embodiment of the invention; Figure 7 is a graph of notch depth as a function of frequency for notches generated in accordance with the specific embodiment of the invention, compared with notches generated using known AIC algorithms; Figure 8 is a graph illustrating a portion of the graph of figure 7, as indicated by arrow A in figure 7; Figure 9 is a graph comparing complementary cumulative distribution function (CCDF) of the average notch depth with the prior art examples; and Figure 10 is a graph illustrating bit error rate (BER) performance for the specific

embodiment in comparison with prior art examples.

An OFDM system is shown in Figure 5. This figure shows a transmitter 12 and a receiver 14, each with an antenna and defming an OFDM channel therebetween. The transmitter 12 comprises a data source 16, which provides data (comprising information bits or symbols) to a channel encoder 18. The channel encoder 18 in this example comprises a convolutional coder such as a recursive systematic convolutional (RSC) encoder.

The channel encoder presents the encoded bits to a channel interleaver 20. The channel interleaver 20 interleaves the bits into symbols in a manner that ensures that errors do not arise due to repeated transmission of a bit in a certain position in a data frame from the antenna, or that adjacent bits are separated so that errors due to breaks in transmission are possibly capable of being recovered.

The channel interleaver 20 passes the interleaved symbols to a mapper 22, that maps incoming symbols to a plurality of output tones, each for transmission from the antenna 25.

The mapper 22 passes the output tones to the transmitter antenna 25 via encoding and amplification circuitry 26 which performs an IFFT to convert the frequency tones to time varying signals, and up-samples and amplifies the signals in preparation for transmission by the transmitter antenna 25.

The encoded transmitted signals propagate through a channel 28 defined between the transmit antenna 25 and a corresponding receive antenna 31 of the receiver 14. The receive antenna 31 provides an input to a down converter 32 of the receiver 14.

The receiver 14 of the specific embodiment is configured with the transmitter 12 in mind. The receive antenna 31 is connected to a down converter 32 that is operable to down-convert the received signal and convert the time varying signal into likelihood data representing encoded symbols. The output of the down converter 32 comprises a signal stream which is provided to a channel de-interleaver 34 which reverses the effect of the channel interleaver 20 and outputs convolutional code on the basis of likelihood data provided by the down converter 32.

The convolutional code output by the channel de-interleaver 34 is then presented to a channel decoder 36. In this example, the channel decoder 36 is a Viterbi decoder, which is operable to decode the convolutional code. The output of channel decoder 36 is provided to a data sink 38, for further processing of the data in any desired manner.

The mapper 22 is operable to generate interference-reducing signals, and to modif' each signal, in the form of a symbol, that is passed to it using a respective interference-reducing signal. The mapper 22 includes a processor 40 that is operable to select the interference-reducing signals that create a desired spectral notch without causing excessive power in the up-sampled spectrum. Operation of the mapper 22 and processor 40 is illustrated in overview in the flow chart of Figure 6.

In the preferred mode of operation, the mapper 22 modifies the signals at the tones immediately adjacent to the AIC edge tones and to design AIC tones so as to provide a frequency notch or notches in the output signals at the interference band or bands. The processor 40 applies the procedure to generate the appropriate AIC tones to provide frequency notches and desired AIC tone characteristics, as described in more detail below. The AIC tones are inserted by the mapper 22 into the signals to be transmitted via the antennas, in place of portions of those signals at and around the interference band. Typically, each AIC tone replaces a corresponding tone in a signal. Selection of the AIC tones comprises selecting the amplitude and/or phase of the tone. So, for instance a complex valued tone, having a value of say 1+1 may be replaced by a selected AIC edge tone having a value of, say, 0.776+0.54i. In variants of the preferred embodiment the processor 40 is separate from but linked to the mapper 22.

The interference band or bands may be identified by operation of the transmitter or receiver antennas, which may be used to detect the presence of narrowband signals in the range of operation of the apparatus. The interference band or bands may, alternatively or additionally, be obtained by reference to a radio licensing authority, or other source, which maintains records of licensed narrowband transmissions. The location of the interference band or bands may be pre-programmed or otherwise stored in the processor 40.

To generate a notch or notches in a desired frequency band or bands, two subcarriers, one at each side of the interference band, are selected as AIC edge tones to suppress the residual interference from the data subcarriers. More subcarriers can be used but only two are considered in the following example.

An example of operation of the apparatus of Figure 5 is now considered in more detail.

An OFDM symbol with N=128 subcarriers is to be transmitted. A spectral notch is to be generated at subcarriers 85 to 87, which span a narrowband interference region, using AIC tones at subcarriers 84 to 88. It should be noted that this example uses specific values for the number of subcarriers, and interference tone indices, but it will be understood that these are not exclusive, and the preferred embodiment may be used with a wide range of transmission characteristics. It should also be noted that the formulae used below relate to the single antenna example and to a k-times upsampling ratio, but that in other embodiments any number of antennas and any suitable upsampling rate may be used and equations 1 to 4 may be generalised to the case of any arbitrary number, n, of antennas.

The following steps, set out in the flow chart illustrated in Figure 6, are performed by the mapper 22 and associated processor 40:-F, which is an N x N matrix, is defined as the normalised Fourier transform matrix and Fc (a kW x N matrix) is defmed as the k times up-sampled Fourier matrix. It will be appreciated by the skilled person that, in other embodiments of the invention, other upsampling factors can also be used).

A kernel matrix, P is defined, with its rows being the (k(f +1) -(k -1)) to (k(f -1) -(k -l))th rows of Fk FH and its columns being the f to f,. th columns of the matrix Fk F, where F" is the Hermitian transpose of F. Furthermore, t is used to denote the AIC tones to substitute subcarriers from -to -1r which are length-M (M f,. -f, + I) vectors used to cancel out the residual interference in the interference band. The process of implementing AIC with excess power suppression is summarised as follows.

In step S6, for each N OFDM subcarriers to be transmitted, the power of subcarriers -1 and f,. +1 are backed off by x dB, simply by multiplying the value of these data subcarriers by that is: s(f, -1)=s(f, -1)xV10'° S(f,. +l)=S(f,. +1)xVlO'° For example, X = 3dB is used to get the results in Figures 7-9 and Table I; In step S8, subcarriers f to J',. are nulled among the N =128 subcarriers to be transmitted. The resulting vector is denoted q.

In step SlO, the k times up-sampled spectrum is obtained, of one OFDM symbol generated from q by performing an IFFT (equivalent to left multiplying the vector q by the matrix F", i.e., F" q), and then k times up-sampled FFT over q (i.e., FkFq). In step S12, the (k(f +1)-(k -1)) to (k(fr -I)-(k -l))th frequency positions from the resulting k times up-sampled spectrum are collectively denoted as vector d Instep S14, Yamaguchi's LS method is employed for t, i.e., t = _(PHP)Pd Finally, in step SI 6, subcarriers to fr are substituted by t and the resulting OFDM symbols are transmitted.

By applying a power back off at two data subcarriers adjacent to the edge AIC tones, the excess power in the spectrum can effectively be suppressed compared to that using the LS solution and the resulting notch depth can achieve that of the LS solution.

The embodiment described above offers the potential for improvements over Yamaguchi and US2006/008016 as will be understood from experiments performed whose results are illustrated in figures 7 to 9. These figures show various trials carried out using QPSK constellation. In particular, the trial results illustrated in figures 7 and 8 refer to trials involving LS and RLS with S = 0.01 and 5 0.03. Spectra are up-sampled 4 times and are averaged over 10000 OFDM symbols.

It overcomes the drawbacks of the traditional LS method where large excess power in the up-sampled spectrum can be observed. Providing a notch depth comparable to that using the LS method (Figures 7 and 8), it also effectively suppresses the maximum average power of the up-sampled spectrum by 1.6 dB compared to the LS solution (Table I below), indicating a 1.6 dB gain of the average transmission power when the method of this embodiment is used to perform AIC.

Method Maximum power Peak of the power Average of the (relative to data at the notch power at the notch subcarriers) of the position (dB) position (dB) spectra averaged over 10000 OFDM _________________ symbols (dB) _________________ _________________ LS 1.8193 -41.3791 -48.6454 RLS, 5 =0.01 0.9813 -36.5021 -38.3347 RLS, S = 0.03 0.4841 -29.4642 -33.5799 Power back off on 0.2046 -4 1.7083 -49.0861 tones adjacent to the edge AIC tones ___________________ __________________ Further, it overcomes the drawbacks of the traditional RLS method where power suppression on the edge AIC tones is achieved at the expense of an elevated notch. The method described above can create, on average, a deeper notch than can the RLS method (Figures 7 and 8) and a 10-15 dB gain in the complementary cumulative distribution function (CCDF) of the mean of the notch depth (Figure 9), while achieving a lower maximum average power level compared to the RLS solutions (Table 1).

Finally, the CCDF plot illustrated in figure 10 indicates the probability of the average notch depth exceeds a given value (NotchDeptho). Backing off the power of only two data subcarriers does not cause performance degradation when error correcting codes are used, which are generally essential in OFDM transmission, as is shown in Figure 10 that when punctured convolutional codes with code rates 2/3 and 7/8 are used, the BER performance with and without the power backoff at two subcarriers are almost the same.

The above embodiment describes a simple method of suppressing the excess power of the up-sampled spectrum when performing AIC. In addition to the widely accepted LS AIC method, it simply needs a slight change in the software processing (multiply two subcarriers by 10' ) at the transmitter, and the receiver does not need to have knowledge of whether power back off has been applied at the transmitter or not if phase-modulated signals, such as QPSK, are used for transmission, which is the most widely accepted signal constellation for UWB systems such as in ECMA 368 and WiMedia 1.0.

The method can be used for devices where interference avoidance is desired subject to given transmission power constraints, such as UWB devices.

Other possible embodiments of the invention could employ multiple transmit and receive antennas. In such circumstances, it would simply be required to apply the AIC approach described herein to each antenna.

It will be understood that the present invention has been described above purely by way of example, and modifications of detail can be made within the scope of the invention.

Each feature disclosed in the description, and (where appropriate) the claims and drawings may be provided independently or in any appropriate combination.

Claims

CLAIMS: 1. A method of shaping a signal intended for transmission in a spectral band in which are defined a plurality of discrete transmission subcarriers, said shaping being intended to define a transmission suppressing notch within a particular portion of said spectral band, the method comprising identifying notch subcarriers contained within the portion of said intended notch, edge subcarriers positioned on either side of said notch subcarriers and, for each edge subcarrier, at least one adjacent subcan-ier immediately adjacent said edge subcarrier and away from said intended notch position; diminishing the intended signal for transmission at said identified adjacent subcarriers, suppressing intended signals for transmission on said notch subcarriers, and determining and applying interference-reducing signals to be applied at said edge subcarriers and notch subcarriers.
2. A method in accordance with claim 1 wherein a single pair of adjacent subcarriers is identified.
3. A method in accordance with claim 1 wherein said identifying of adjacent subcarriers comprises identifying a plurality of subcarriers at each edge subcarrier.
4. A method in accordance with any one of the preceding claims wherein said diminishing of said identified adjacent subcarriers comprises diminishing each subcarrier by a corresponding diminution factor.
5. A method in accordance with claim 4 wherein a single diminution factor is applied to each identified adjacent subcarrier.
6. A method in accordance with claim 5 wherein said diminishing comprises reducing each said identified adjacent subcarrier by 3dB.
7. A method in accordance with any one of the preceding claims wherein said determining of an interference reducing signal comprises calculating a plurality of signal elements, one for each notch subcarrier and each edge subcarrier, on the basis of applying an inverse fast Fourier transform (IFFT) over the intended signal including notch subcarriers, edge subcarriers and diminished adjacent subcarriers, applying an up-sampled fast Fourier transform a predetermined number of times, and using the resultant subcarrier information to determine an interference reducing signal element for each of the nulled notch subearriers and edge subcarriers.
8. A method of transmitting a signal in a wideband channel in which a narrowband interference source is present, comprising applying the method in accordance with any preceding claim in respect of the narrowband, and imposing said interference avoiding signal on said channel.
9. Apparatus for processing a signal intended for transmission in a spectral band in which are defmed a plurality of discrete transmission subcarriers, said apparatus being operable to shape said signal to define a transmission suppressing notch within a particular portion of said spectral band, the apparatus comprising means for diminishing a signal element intended for transmission at subcarriers adjacent edge subcarriers bounding said notch, means for identifying such notch subcarriers as are contained within said notch and for suppressing intended signals for transmission on said contained subcarriers, and means for determining and applying an interference-reducing signal to be applied at each of the contained notch subcarriers and the edge subcarriers.
10. Apparatus in accordance with claim 9 wherein said diminishing means is operable on a single adjacent subcarrier identified at each edge subcarrier of the notch.
11. Apparatus in accordance with claim 9 wherein said diminishing means is operable on a plurality of adjacent subcarriers at each edge subcarrier.
12. Apparatus in accordance with any one of claims 9 to 11 wherein said diminishing means is operable to diminish each subcarrier by a corresponding diminution factor.
13. Apparatus in accordance with claim 12 wherein a single diminution factor is applied to each identified adjacent subcarrier.
14. Apparatus in accordance with claim 13 wherein said diminishing means is operable to diminish each said identified adjacent subcarrier by 3dB.
15. Apparatus in accordance with any one of claims 9 to 14 wherein said determining and applying means is operable to calculate a plurality of signal elements, one for each subcarrier contained in said notch and said edge subcarriers, on the basis of applying an inverse fast Fourier transform (IFFT) over the intended signal including said nulled contained subcarriers, edge subcarriers and diminished adjacent subcarriers, applying an up-sampled fast Fourier transform a predetermined number of times, and using the resultant subcarrier information to determine an interference reducing signal element for each of the nulled contained subcarriers and the edge subcarriers.
16. A wireless communications apparatus operable to emit a signal in a wideband channel in which a narrowband interference source is present, comprising apparatus in accordance with any one of claims 9 to 15 and means for imposing said interference avoiding signal on said channel.
17. A computer program product comprising computer executable instructions operable to configure a general purpose computer to perform the method of any one of claims 1 to8.
18. A computer program product in accordance with claim 17 comprising a computer readable storage medium.
19. A computer program product in accordance with claim 17 comprising a computer readable signal.