Title
A Method of and a Communication Device for Adjacent Channel Interference Cancellation in a Radio Telecommunication System.
Field of the Invention
The present. invention relates generally to telecommunication systems having a radio link connection between two or more telecommunication units and, more specifically, to Code Division Multiple Access (CDMA) and Wideband CDMA (WCDMA) radio telecommunication systems having a plurality of geographically spread fixed and/or mobile radio telecommunication units operating on a plurality of radio transmission channels, in a plurality of limited geographical areas, called cells.
Background of the Invention
In CDMA-based telecommunication systems the so-called Near- Far effect poses some problems on the design of the system and the telecommunication units used, i.e. the User Equipment (UE) and Radio base Stations (RBS) of the telecommunication system. If all users in a cell use the same radio channel or frequency, there is a considerable risk that users far away from an RBS do not have enough Up-Link (UL) transmission power to be received by the RBS compared to close-by users, that is users which are more closely spaced to the RBS. Although modern CDMA telecommunication systems are equipped for reducing the UL power of a UE in the vicinity of an RBS, near-by UE's cannot reduce their output level beyond a certain minimum value, due to dynamic range limitations. This is a well-known problem in CDMA systems, which can be
tackled by fast UL power control with sufficient dynamic range or by intra-cell interference cancellation. This technology normally makes use of detailled knowledge of UL transmission characteristics, to remove most of the near-by user signal from the composite RF (Radio Frequency) signal received at an RBS, before attempting to detect the far-away users. A more advanced' solution detects all user simultaneously, using knowledge about codes, code phases, transport formats, timing, etc. assigned to UE. These technologies cannot be applied, however, when there is only limited knowledge about interference present on the desired carrier frequency. In WCDMA, a system or network operator has little or no room to escape interference from adjacent channels using frequency planning, as many of these systems have only three or just two carriers available. Nevertheless, such interference will be present, especially when network deployment is uncoordinated, i.e. the RBS of a first system are relatively far away from the RBS of a second system. For example, assume a far-away UE of the first system. Such UE will transmit at relatively high power in order to be received by an RBS of the first system. However, if these UE are close to an RBS of the second system, they will cause a lot of UL interference at this RBS. At the same time, the UE of the first system will receive a lot of Down-Link (DL) interference from the RBS of the second system. Ultimately, this will force the far-away UE of the first system to disconnect and stop transmission. This, however, solves the interference problems caused in the second system only partly (and in a brute-force manner), because severe problems in the second system can occur already before shutting down of the UE of the first system. It is important to note, that the above described interference scenario cannot be solved by more stringent requirements on the RBS, as it is dominated by UE performance, both in UL and DL. The parameters describing this performance are the Adjacent Channel Selectivity (ACS) and the Adjacent Channel Leakage Ratio (ACLR) .
ACS quantifies how much interference a radio receiver can take on an adjacent radio transmission channel, that is a radio transmission channel other than the desired radio channel, without degrading performance due to imperfect components. ACLR quantifies how much energy a radio transmitter sends out on an adjacent transmission channel, that is a radio transmission channel other than the desired radio transmission channel at which the transmitter is tuned. It will be appreciated that even a very high ACS will not help a receiver in the presence of an adjacent channel interferer with a poor or bad ACLR. Nor will a v ry low ACLR transmitter help a receiver on an adjacent channel with a poor ACS. For example, Universal Mobile Telecommunication System (UMTS) only mandates 33 dB for both ACS and ACLR in the UE, where the RBS has a much better performance. In the above scenario, an RBS faces a lot of co-channel interference from uncoordinated UE, which may persist for a long time. Severe impact is expected in so-called traffic highway sites, because they are often UL limited, because they tend to have low coupling loss to UE, due to Line Of Sight (LOS) propagation. Further, they are supposed to be reused from the Global System for Mobile communications (GSM) to UMTS, and deployment in GSM has not been co-ordinated between operators. The existence of a highway will lead UE close to the RBS in a predetermined manner. Another relevant scenario to be mentioned in this context is interference from UMTS-Time Division Duplex-(TDD) UE and possibly also UMTS-TDD RBS. In practice, several solutions to solve or remedy the co- channel interference problem, as disclosed above, have been discussed. Among which: - co-locate all RBSs; however, this is not feasible, due to site reuse requirements; - position RBSs away from the highway; this is likewise not
feasible, due to site reuse requirements; - direct receive antennas away from the likely origin of interference, e.g. the highway; however, this has an impact on cell planning. - improve UE performance; which is, however, very costly and provides only relief for the newest equipment. - add more UL margin in the cell planning; this solution is feasible only as much as coverage is considered; impact on system stability persists; this solution is also relatively expensive, as more sites are needed.
Summary of the Invention
It is an object of the present invention to cancel or reduce co-channel interference in a radio communication signal received at a radio communication device of a CDMA, in particular a WCDMA, radio telecommunication system, which radio signal is transmitted on a desired radio channel of the system. It is in particular an object of the present invention to provide a method of and a device equipped for cancellation of co-channel interference, not necessitating changes in or putting constraints on cell or site planning of the radio telecommunication system and which method can be principally applied in or with existing UE and RBS, if required. In a first aspect of the present invention, there is provided a method of adjacent channel interference cancellation in a radio communication signal received at a radio communication device and transmitted on a desired radio transmission channel of a CDMA radio transmission system, in particular a WCDMA radio transmission system, the system having a plurality of radio transmission channels, the method comprising the steps of: - obtaining a correction signal representative of the
adjacent channel interference caused by radio signals transmitted on a radio transmission channel other than the desired radio transmission channel and received at the radio communication device, and - obtaining a desired radio communication signal by correcting the received radio communication signal using the correction signal , and wherein - the correction signal is obtained using a non-linear transmission frequency spectrum model of radio signals transmitted by radio communication devices on a radio transmission channel other than the desired radio transmission channel. The method according to the invention is based on the insight that the RF Power Amplifier (PA) frequency spectrum of a radio communication device can be adequately modelled by a non-linear transmission frequency spectrum model. By employing such a non-linear model, in accordance with the present invention, the amount of co-channel interference caused on a desired radio channel by a radio communication device transmitting on a radio channel other than the desired channel can be accurately calculated, providing a correction signal. The radio communication signal received at the desired radio communication channel is than corrected for the co-channel interference using the correction signal as calculated. It will be appreciated that the amount of co-channel interference for a plurality of interfering devices can be calculated using the same or a plurality of different non-linear frequency spectrum models, for example, optimally adapted to a particular radio communication device, i.e. the PA thereof. In a preferred embodiment of the present invention, the receiving radio communication device is a radio base station of the radio transmission system and the correction signal is obtained by the radio base station using a non-linear transmission frequency spectrum model of radio signals transmitted by user equipment devices of the radio
communication system. Correction of the co-channel interference at the RBSs provides the most effective remedy. This, because the PA of the UE-, due to UE transmitter imperfections, provide a relatively heavily distorted WCDMA interference, primarily due to non-linear effects on the modulation. On the other hand, the invention is not limited to its implementation in RBS equipment. In a further embodiment of the present invention, the receiving radio communication device is a user equipment device of the radio transmission system and the correction signal is obtained by the user equipment device using a non-linear transmission frequency spectrum model of radio signals transmitted by radio base stations of the radio communication system. Because of the relatively "clean" signal transmitted by an RBS, implementation of the invention in a UE will be less effective compared to the RBS mplementation. In a UE, interference is dominated by leakage of the clean radio base station signal through the UE filters which, essentially, is a linear process. In a preferred embodiment of the invention, the correction signal is obtained using a non-linear transmission frequency model comprising a polynomial expansion having weighed odd order terms. In a simplified embodiment of the invention, the correction signal is obtained from a weighed third order term of the polynomial expansion. The third order term is most dominant. A simplification of the model results in a simplification of the calculations which have to be performed and saves calculation power and resources, while speeding up the correction process. Weighing of the terms, in yet another embodiment of the invention, may be based on minimizing error on the desired communication signal by adapting the individual weighing factors. Selection of a suitable non-linear transmission frequency
spectrum model of a radio signal, as disclosed above, can be based likewise on minimizing error on the desired communication signal by adapting the individual weighing factors, for example. However, in yet another embodiment of the invention, the non-linear transmission frequency spectrum model is selected based on information identifying a particular radio communication device generating a radio signal at a radio transmission channel other than the desired radio transmission channel. In a preferred embodiment of the invention, the correction signal is obtained following the steps of: providing a baseband representation of the radio communication signal received at the desired radio transmission channel; - providing a baseband representation of a radio signal received at a radio transmission channel other than the desired radio transmission channel; - modelling the baseband representation of the radio signal received at the radio transmission channel other than the desired radio transmission channel using a non-linear transmission frequency spectrum model ; - frequency converting the modelled baseband representation for providing a modeled radio signal having a frequency relation with respect to the baseband representation of the radio communication signal received at the desired radio transmission channel corresponding to the other radio transmission channel in respect of the desired radio transmission channel; - amplitude scaling of the frequency converted modeled baseband representation in respect of the baseband representation of the radio communication signal received at the desired radio transmission channel, providing a reference signal, and - subtracting, from the baseband representation of the radio communication signal received at the desired radio transmission
channel, part of the reference signal comprised by the baseband representation of the radio communication signal received at the desired radio transmission channel, providing the desired radio communication signal at baseband. Processing at baseband has the advantage of being able to apply suitable filter and processing techniques and components, already available in a radio communication unit. Amplitude scaling is based, in yet another embodiment of the invention, on minimizing error on the desired radio communication signal . In a preferred embodiment of the invention, scaling is based on a comparison of pilot signals associated with the radio communication signal received at the desired transmission channel and a radio communication signal received at a radio transmission channel other than the desired radio transmission channel. This embodiment advantageously makes use of the pilot signals available in CDMA and WCDMA. The correction signal may be obtained, in accordance with a preferred embodiment of the invention, using digital signal processing techniques. In a second aspect, the invention provides a communication device arranged for adjacent channel interference cancellation in a radio communication signal received by the radio communication device and transmitted on a desired radio transmission channel of a Code Division Multiple Access radio transmission system, in particular a Wideband Code Division Multiple Access radio transmission system, the system having a plurality of radio transmission channels, the communication device comprising: - means for obtaining a correction signal representative of the adjacent channel interference caused by radio signals transmitted on a radio transmission channel other than the desired radio transmission channel and received at the radio communication device, and
- means for obtaining a desired radio communication signal by correcting the received radio communication signal using the correction signal, wherein the means for obtaining the correction signal comprise a non-linear transmission frequency spectrum model of radio signals transmitted by radio communication devices on a radio transmission channel other than the desired radio transmission channel. The radio communication device is preferably a radio base station of the radio transmission system and the means for obtaining the correction signal comprise a non-linear transmission frequency spectrum model of radio signals transmitted by user equipment devices of the radio communication system. However, the radio communication device may also be a user equipment device of the radio transmission system and the means for obtaining the correction signal comprise a non-linear transmission frequency spectrum model of radio signals transmitted by radio base stations of the radio communication system. In further embodiments of the invention, the radio communication device is suitably adapted to perform any of the steps of the invention, disclosed and discussed above, including but not limited to frequency conversion means and demodulation means, modelling means, amplitude scaling means, correction means, analog-to-digital conversion means, and digital signal processing means. The invention further provides a Code Division Multiple Access radio transmission system, in particular a Wideband Code Division Multiple Access radio transmission system, comprising at least one communication device arranged and operated in accordance with any of the embodiments disclosed above. The invention will now be described in more detail with reference to the appended drawings.
Brief Description of the Drawing
Fig. 1 shows, in a very schematic manner, a first and second uncoordinated telecommunication system deployment, according to the prior art, in a so-called highway scenario. Figure 2 shows a graphic representation of a typical transmission frequency spectrum of a power amplifier of a radio communication device, in particular a UE. Figure 3 shows, in a schematic manner, a block diagram of a communication device according to the present invention, including frequency spectra of the processed radio communication signal.
Detailed Description of the Embodiments
Without the intention of a limitation, the invention will now be described and illustrated with reference to an exemplary embodiment of a telecommunication device. In figure 1, reference numerals 1 and 2 designate an RBS of a first radio telecommunication system and reference numeral 3 designates an RBS of a second radio telecommunication system. The first and second radio telecommunication systems are uncoordinated and are deployed in a so-called highway scenario, that is the RBS 1, 2, 3 are arranged alongside a traffic highway, for example. Reference numeral 4 designates a mobile UE of the first radio system. In the position shown, the UE 4 is in communication with the RBS 2 of the first radio telecommunication system, indicated by an arrow 7. As indicated by arrows 5, 6 the UE 4 is at a relatively short distance to the RBS 3 of the second radio telecommunication system and at a relatively large distance from the RBS 2 of the first radio telecommunication system., indicated by an arrow 7. In this particular
case, the UE 4 transmits at a relatively high output power, providing an RF output signal 7 to be recei ed by the RBS 2. At the same time, because the first and second radio telecommunication systems are uncoordinated, the transmission signal of the UE 4 will be received as an interference signal 8 at the RBS 3 of the second radio telecommunication system. This situation, which not necessarily needs to be a temporal one, because the UE 4 may stay for a longer period of time in the position shown in figure 1, will cause a lot of interference at the RBS 3 of the second radio telecommunication system. That is, the UE 4 causes a lot of UL interference at the RBS 3 and, at the same time, the UE 4 receives a lot of DL interference from the RBS 3 of the second radio telecommunication system. Eventually, the result of this interference may be that the UE 4 disconnects itself and stops transmission, because the connection with the RBS 2 can not be maintained. Although this solves the interference problems caused i the RBS 3 of the second telecommunication system, this situation is not favourable from a point of view of a system operator, providing reliable communication links. The above can be hardly improved by a proper frequency cell planning in, for example, WCDMA because the number of radio transmission channels available for the system as a whole is limited to two or three, at the most. Even if the UE 4 transmits on a radio transmission channel different from the radio transmission channel used by the second radio telecommunication signal, the RBS 3 of the second radio telecommunication system will receive a considerable amount of co-channel interference from the uncoordinated UE 4. The UE 4, because of its imperfect PA of the radio transmitter, produces interferences at the radio transmission channel at which the RBS 3 of the second radio telecommunication system operates. Not only the RBS 3 receives co-channel interference from the UE 4, likewise, the UE 4 receives co-channel interference from the RBS 3. In the latter case, due to the generally more perfect radio transmitters of
a RBS, that is the RBS provides a relatively "clean" radio communication signal, the co-channel interference from the RBS at the UE will be not that severe compared to the co-channel interference caused in the RBS by the UE. The term co-channel interference is to be construed, in the light of the present invention, as a radio signal received on a desired radio communication channel, however not transmitted by a desired radio communication device. Figure 2 shows a typical frequency spectrum 13 of a PA of a UE. Reference numeral 11 indicates the ideal transmission spectrum of an ideal PA, shown in broken lines. The filter response of a radio communication receiver is indicated by reference numeral 12 and shown is in dashed-dotted lines. In this particular case, the receiver is tuned at around base band, i.e. 0 MHz, while the frequency spectrum 13 of the PA of the UE is positioned at a centre frequency of 5 MHz. That is, the transmission spectrum of the UE has a frequency offset of 5 MHz compared to the receiver filter response 12. As can be seen from figure 2, the transmission frequency spectrum 13, due to the imperfections of the PA of the UE, overlaps with the filter response 12 of the receiver, which means that part of the radio communication signal transmitted by the UE is received at the receiver which is not tuned at the frequency of 5 MHz at which the PA of the UE transmits. This part of the signal transmitted by the UE and received at the receiver causes an unwanted interference. This type of interference is called co-channel interference. That is, the receiver, which is not tuned at the frequency at which a UE transmits, will receive an interference • signal from the transmission of the UE due to an imperfect PA of the UE, primarily due to non-linear effects on the modulation by the PA of the UE. Those skilled in the art will appreciate that the radio communication signal of the UE, in practice, will be modulated on a
carrier frequency of a radio transmission channel of a radio telecommunication system. It has been found that the transmission frequency spectrum 13 of the communication signal transmitted by the PA of a UE can be suitably modelled using a non-linear transmission frequency spectrum model of the PA. In particular, a polynomial expansion with odd order terms in the form of: a0 + aj x+a3 x3 + a5 x5 + wherein x = the input signal to the PA of a UE, and aη = term coefficient, i = 1, 3, 5, ... In the case of a linear PA, the output signal would have simply the form a: x. In figure 2, the several components of the polynomial expansion are shown in the form of "pedestals" in the frequency spectrum. From figure 2, it will be immediately clear that the third order term a3 x3 is the principle component of the ACL leakage, causing the co- channel interference problems discussed above. Removal of this third order component from the received radio communication signal on a desired channel will cancel, for the greater part, the co-channel interference in the received radio communication signal. It will be appreciated that the value of the coefficients a0, a15 a3... of the polynomial expansion may depend on a particular type of PA or UE. Figure 3 shows an embodiment of a communication device 200 in accordance with the present invention, arranged for adjacent channel interference cancellation in a radio communication signal received by the radio communication device and transmitted on a desired radio transmission channel of a radio communication network, such as a CDMA radio telecommunication system, in particular a WCDMA radio
telecommunication system. Reference numeral 10 discloses an RF front end receiver of the communication device, arranged for receiving radio communication signals at a radio transmission channel of the radio communication system and for converting same to baseband, providing an output signal 210, having a frequency spectrum as schematically shown in Figure 3, including the desired radio communication signal and adjacent channel interference. As depicted in the frequency spectrum of the output signal 210, the desired radio communication signal is concentrated around 0 Hz and the adjacent channel signal is shifted in frequency compared to the desired signal. The amount of frequency shift depends on the system properties, i.e. channel bandwidth of the telecommunication system. Reference numeral 20 designates frequency conversion and demodulation means, including a Local Oscillator (LO) input, arranged such that a base band version of the adjacent or co-channel signal is provided, the frequency spectrum of which is designated with reference numeral 220. The adjacent channel signal is now concentrated around 0 Hz. The thus obtained base band signal 220 is filtered by a first band filter 30, to remove part of the desired radio communication signal, such that at the output of the band filter 30 an adjacent channel signal 230 is provided. That is, the portion of the desired radio communication signal is removed from the frequency converted and demodulated signal 230, provided by the frequency conversion and demodulation means 20. The adjacent channel signal 230 at the output of the first band filter 30 is now, in accordance with the present invention, fed to modelling means 40, comprising a non-linear transmission frequency spectrum model for modelling the frequency spectrum of the PA of a particular UE, for example the polynomial expansion with odd order terms described above. In a simplified version of the communication device 200, the non-linear transmission frequency spectrum model can be restricted to
the third order term of the polynomial expansion. By feeding the adjacent channel signal 230 at the output of the first band filter 30 into the modelling means 40, a signal is obtained resembling the signal transmitted by the PA of the particular UE, however at base band, indicated by reference numeral 240 and the frequency spectrum of which is shown. See also Figure 2. In order to obtain the adjacent channel signal in relation to the desired radio communication signal at base band, the signal provided by the modelling means 40 has to be frequency converted by frequency conversion means 50, including a Local Oscillator (LO) input, providing the frequency shifted adjacent channel signal 250. That is, the modelled adjacent channel signal 240 is frequency up-converted to take the same frequency position with respect to the output signal 210 of the RF front end receiver 10. The frequency shifted modelled adjacent signal 250 is filtered by a second band filter 70, in order to obtain part of the modelled adjacent channel signal interfering with the desired radio signal, as indicated by reference numeral 260. In the frequency spectra shown, the part of the adjacent channel interference in the desired radio communication signal is hatched. The signal 260 obtained at the output of the second band filter 70 now is a correction signal for correcting the received radio communication signal 210 for cancellation of adjacent channel interference, in accordance with the present invention. This correction is performed by correction means 140, which may take the form of summing or subtracting means. That is, the received radio communication signal 210 provided by the RF front end receiver 10 is filtered, at base band, by a third band filter 60, such to filter away part of the adjacent channel signal beyond the communication frequency band of the desired signal, providing the signal 160. Now, both the filtered received radio communication signal at base band 160 and the correction signal at base band 260 are fed to
the inputs of the correction means 140, such that the adjacent channel interference part is cancelled out of the received radio communication signal at base band, for example subtracted, providing the desired radio communication signal in accordance with the invention at the receiver output 150. In order to correct the received and filtered radio communication signal 160 by the proper amount of co-channel interference, amplitudes scaling means 80, 90, 100, 120 are provided, connected as shown in figure 3. In the embodiment shown, pilot signal detection means 90 are provided, arranged for detecting and measuring the amplitude of a pilot signal transmitted in the radio communication network. By comparing an amplitude of the pilot signal and an amplitude of the desired radio communication signal provided at the output 150 of the radio communication device, for example using summation means 100, an error signal 110 is obtained at the output of the summation means 100, which error signal 110 is representative for the relative amplitude of the radio communication signal. The error signal 110 provides an input for amplitude scaling means 120, which are arranged for selecting a suitable transmission frequency spectrum model for a PA of a particular UE and for adapting 80 the amplitude of the correction signal 260 of the second band filter 70, such that the amplitude of the correction signal is properly adjusted to the amplitude of the received radio communication signal 210 to be corrected. The adaptation means 80 may take the form of controllable attenuator/amplifier means. The dotted lines 130 indicate control and selection lines from the amplitude scaling and model selection means 120 to the modelling means 40 and the adaptation means 80. As discussed above, the scaling means 120 may comprise a plurality of frequency spectrum models of transmitters operative in the
radio communication system and any other radio system. The proper model is selected, among others, by minimizing the error on the desired radio communication signal 150. In the case of a polynomial expansion, as disclosed above, a proper weighing or scaling of the odd order terms is provided, using, for example, Minimum Mean Squared Error (MMSE) methods or Least Mean Square (LMS) adaptive algorithms, which are known to the skilled person. A proper non-linear frequency spectrum model or adaptation of the frequency spectrum model may also be based on knowledge about a particular UE, for example information identifying the UE, which information may be incorporated in the radio communication signal received by the RF front end receiver 10. To this end, the scaling means 120 may be provided with suitable identification means 170, for identifying a particular UE and selecting the proper frequency spectrum model . In figure 3, for illustrative purposes, the several signals obtained in the communication device are illustratively depicted. The dashed part of the signals is the co-channel interference which has to be removed or cancelled from the received radio communication signal, in order to provide the desired radio communication signal in accordance with the present invention. Because the several frequency spectra 150, 160, 210 - 260, are for illustrative purposes only, no frequency or amplitude values have been depicted. Although, in the above, the method of the present invention and the operation of the communication device according to the present invention are elucidated at base band and assuming analog conversion, modulation and filter means, those skilled in the art will appreciate that the means 20-120 of Figure 3 may be replaced by equivalent digital processing means, provided that the radio communication signal received by the RF front end receiver 10 is converted into a digital representation using analog-to-digital converter means 190, as indicated
by dotted lines. Further, instead of simply scaling a non-linear correction signal before subtraction from the received radio communication signal, it may be necessary to have some form of FIR filtering to account for delays in the different processing branches of the radio communication device. That is, the coefficient a3 may be replaced by a tapped delay line, in the case that a simplified non-linear frequency spectrum model in the form of the third order term of the polynomial expansion is used. The methods according to the present invention can be performed by suitable digital signal processing means, such that the method can be implemented in existing radio transmission devices, such as used in an RBS of a CDMA or WCDMA radio communication system. The present invention is not limited to the embodiments shown in the drawings and disclosed above. Those skilled in the art will appreciate that numerous modifications and adaptations to the embodiments can be made, without departing from the novel and inventive scope of the present invention, defined by the attached claims.