WO2001056183A1

WO2001056183A1 - Method and receiver in communication system

Info

Publication number: WO2001056183A1
Application number: PCT/FI2001/000060
Authority: WO
Inventors: Ülo PARTS
Original assignee: Nokia Networks Oy
Priority date: 2000-01-27
Filing date: 2001-01-24
Publication date: 2001-08-02
Also published as: AU2001230284A1; FI20000158A0

Abstract

A method for carrying out parallel interference cancellation from a spread spectrum signal, the method comprising the steps of receiving a combination signal including user signals spread with several spreading codes, composing the user signals of the combination signal in order to form user signal esti-mates, regenerating the spread user signals using the user signal estimates. This method further comprises the steps of generating a scaling coefficient k from the combination signal in order to scale the interference signal to be re-duced, generating the scaled interference signal using the spread user signals and the scaling coefficient k, subtracting the generated scaled interference signal from the combination signal in order to reduce the multi user interference of the user signal.

Description

METHOD AND RECEIVER IN COMMUNICATION SYSTEM

FIELD OF THE INVENTION

The invention relates to a method for carrying out parallel interference cancellation in a radio system employing code division multiple access and to a receiver implementing the method.

BACKGROUND OF THE INVENTION

A radio system employing Code Division Multiple Access (CDMA) allows several sending and receiving stations to communicate on the same radio spectrum frequency band. A spreading code that the user employs to spread the information in his/her baseband signal is allocated for each user during the connection. The receiver of the signal is in turn capable of identifying the information sent by the user by releasing the information sent utilizing a corresponding release code thereof. An advantage with CDMA is, for example, the efficient utilization of a frequency band and the data security of the system. A drawback is in turn that the users operating on the same frequency band cause interference to each other's transmissions owing to the unorthogonality of the spreading codes and the release codes of the spreading and owing to the lack of synchronization between the transmitters. Interference in turn affects the utilization possibilities of the radio system capacity and the quality of the connections.

Various methods have been created in CDMA systems for carrying out interference cancellation (IC). Efficient methods to be used for interference cancellation of several users are methods based on Multi User Detection (MUD). The idea with interference cancellation methods using MUD is that the information of several other users is utilized for identifying the signal of each user. Serial interference cancellation (SIC) is presented for example in US 5,579,304. The idea is to remove in each interference cancellation stage the signal of one or more users in the receiver from the received combination signal of all signals. The signals to be removed from the combination signal dur- ing each interference cancellation round are the strongest, for example, in view of the transmission power, in which case the interference caused by the signals to other users is most likely the strongest. Thus, the users of the weaker signals can be identified in the later interference cancellation stages from a signal, from which the users sending the strongest signals have been purged. Some of the methods based on multi user detection are parallel interference cancellation (PIC) methods, where the signals of all users are identified entirely during each interference cancellation round, and during the signal identification of an individual signal, the signal of all other users than said individual user is reduced from the combination signal. The drawback in such a solution is that multi-purpose interference is cancelled during each interference cancellation round without being aware of the quality of the detected interference. During the initial steps of interference cancellation the interference estimates may be weak, since the signal-to-interference ratio (SIR) is low. Methods cancelling the multi user interference totally do not utilize efficiently the fact that the user signal estimates improve in the later stages, when the signal-to-interference ratio improves.

Publication US 5,644,592 discloses a solution employing prior art parallel interference cancellation that aims to benefit from the fact that the user signal estimates improve on later interference cancellation rounds. The publication presents a method where a scaling coefficient p_k is used to scale the received combination signal and the interference signal formed for each user. In addition, the complement (1- p_k) of the scaling coefficient p_k is used to scale the residual signal obtained from the desired user signal. A significant drawback is included in the solution of the invention.

Several signal processing operations and calculations are carried out in the solution that complicate the implementation of the method and the apparatus solution according to the method.

BRIEF DESCRIPTION OF THE INVENTION It is an object of the invention to provide an improved method and an apparatus for implementing interference cancellation in a communication system. This is achieved with the following method for carrying out parallel interference cancellation from a spread spectrum signal, the method comprising the steps of receiving a combination signal including user signals spread with several spreading codes, composing the user signals of the combination signal in order to generate user signal estimates, regenerating the spread user signals using the user signal estimates. This method is characterized by generating a scaling coefficient k in order to scale the interference signal to be reduced from the combination signal, generating the scaled interference signal using the spread user signals and the scaling coefficient k, subtracting the generated scaled interference signal from the combination signal in order to reduce the multi user interference of the user signal

The invention also relates to a receiver for receiving a spread spectrum signal comprising one or more antennas for receiving a combination sig- nal including user signals spread with several spreading codes, one or more correlators for composing the user signals from the combinations signal in order to generate user signal estimates, one or more regenerators for regenerating the spread user signals using the user signal estimates. The receiver comprises means for generating a scaling coefficient k in order to scale the inter- ference signal to be reduced from the combination signal, means for generating the scaled interference signal using the spread user signals and the scaling coefficient k, a reductor for subtracting the generated scaled interference signal from the combination signal in order to reduce the multi user interference of the user signal. In a radio system employing code division multiple access, such as a mobile communication system, several terminals transmit on the same frequency band. A spreading code, which is significantly different in comparison with the spreading codes of the other terminals, is allocated to the terminal for a call. Each terminal sends a flow of symbols spread with a specific spreading code, which flow of symbols the receiver tries to identify. The solution of the invention aims to cancel multi purpose interference, i.e. interference caused by other users to a recognizable user signal. The receiver, in which the method of the invention and the apparatus implementing the method is carried out, is for example a base station in a CDMA radio network. The method of the invention can also be implemented in other corresponding CDMA radio network receivers where substantially all available spreading codes are known.

In a preferred embodiment of the invention the base station receives several user signals, each one of which being spread with a specific spreading code thereof. The signals are generally multipath propagated, and the same user signal provides the receiver with components scattered and delayed in various ways. As to the capacity questions of the radio network, the spreading codes of different users are not either exactly orthogonal with one another, in which case the users may momentarily cause interference to one another's signals. The invention relates to cancelling the interference caused by other users to a particular user from a received combination signal, so that the information sent by each user can be interpreted in the receiver to best possible effect. The method of the invention is arranged to be implemented in the receiver, in which multi user interference can be cancelled utilizing a parallel interference cancellation method. In such a case, interference caused by as many users as possible is preferably utilized for signal interference cancella- tion of each user. Interference cancellation is carried out in a preferred embodiment of the invention at least during one stage. The practical applications have shown that the best interference cancellation results are obtained using two or three parallel interference cancellation stages. The invention is obviously not restricted to the number of stages in which interference cancellation is carried out.

During each interference cancellation stage the spread of user signals from the combination signal is composed by means of a release code of the spread corresponding with a spreading code for generating a baseband user signal and symbol estimates. Furthermore, the broadband signals of the users are regenerated in the interference cancellation stage using the spreading code of the user and the generated symbol and channel estimates. When the broadband signals of the users are regenerated, they are combined into an interference signal and reduced from the originally received broadband signal. Thus, for example, if the mobile network has 10 users and the aim is to recog- nize the information of user 1 , the signals of the users 2 to 10, which are reduced from the received broadband signal, are regenerated. Parallel interference cancellation can also be implemented to a part of the users only, for example, so that the strongest user signals only, for instance 8 to 10, are reduced in a single interference cancellation stage. In accordance with a preferred embodiment of the invention the amplitude of the regenerated user signal is scaled using the scaling coefficient k in the above interference cancellation stage. The scaling coefficient k preferably exceeds 0 but remains below 1. In accordance with a second preferred embodiment, interference signals multiplied by said scaling coefficient are at first generated from the spread user signals. According to a preferred embodiment the channel estimate is multiplied by the scaling coefficient k. As the spread user signal, symbol estimate and channel estimate, which are multiplied together, are processed in connection with regeneration, it is not relevant for the invention which ones of said product factors are multiplied. However, what is essential for the invention is that only a single multiplication operation is required in order to carry out scaling. According to a preferred embodiment the scaling coefficient k increases between interference cancellation rounds, in other words the coefficient is higher during the later rounds than during the first rounds. What is achieved with this is that as the interference estimates are not of very good quality on the first or on the directly preceding interference cancellation rounds, the effect thereof is reduced. In a preferred embodiment the scaling coefficient increases linearly, thus being for instance 0.5->0.7->0.9, whereby the effect of the later interference cancellation stages significantly exceeds that of the previous stages. The increase of the scaling coefficient can in a preferred embodiment of the invention be bound to the signal-to-interference plus noise ratio. Thus, if the above ratio of the user is great, then as regards the user, only a part of the interference signals is worth cancelling.

The invention provides the significant advantage that the solution of the invention is easy to implement, when the number of calculation operations required for carrying out scaling is small. Only a single constant is required during an interference cancellation round, the constant being used for multiplying a signal component of the interference cancellation stage, such as the interference signal succeeding regeneration. As to the equipment, the solution of the invention provides significant savings as the number of operations to be carried out is reduced.

BRIEF DESCRIPTION OF THE DRAWINGS

In the following the invention will be described by means of the preferred embodiments with reference to the accompanying drawings, in which:

Figure 1 shows roughly a mobile network, Figure 2 is a method diagram showing a preferred embodiment of the invention,

Figure 3 is a block diagram showing a CDMA receiver,

Figure 4 shows an example of the structure of a receiver using parallel interference cancellation, Figure 5 shows a preferred embodiment of an interference cancellation unit, and

Figure 6 shows simulation results obtained using the method of the invention. DETAILED DESCRIPTION OF THE INVENTION

In the following the invention will be described by means of the preferred embodiments with reference to the attached Figures 1 to 6. Figure 1 shows roughly a mobile communication network comprising base stations 100A to 100D. The coverage area of a base station is referred to as a cell indicated in the Figure by C1 to C4 corresponding to the base stations 100A to 100D. The cells may be overlapping and extend over one another, as the Figure shows cell C2 is for example partly overlapped by cells C1 and C3. In the Figure, each cell C1 to C4 area is provided with one or more receivers 102A to 102F, which are for instance mobile phones but may also be other apparatuses provided with radio receiver and/or transmitter properties.

In a radio network using code division multiple access (CDMA), such as a mobile communication network, all users employ the same frequency band simultaneously. The users are distinguished from one another on the basis of a spreading code, by which the information sent by the user is multiplied. Then the information, such as bit flow including speech, is spread to a broad frequency band. The bit rates of the spreading codes used significantly exceed those of the data flow to be sent, being for example 8, 16 or 32. The spreading codes of the users tend to be orthogonally selected, in which case they do not correlate with one another. In practice the spreading codes are not totally orthogonal among each other, and therefore the users interfere with one another. In Figure 1 , the receivers 102D to 102F in cell C4 interfere with each other and are also subjected to interference from the terminals 102A to 102C located within the area of the other cells C1 to C3. Interference be- tween the terminals 102A to 102F is also created when the signal sent by each terminal propagates to the receiver along various paths. What is referred to as multi-path propagation causes the fact that the user signal arrives at the receiver as a signal component delayed in various ways, thus causing interference to the other users. Several methods can be used for cancelling the interference users cause to each other. These methods are in practice applied to base stations, which are aware of all users and of the spreading codes they employ and are therefore able to cancel the interference caused by other users to a particular user. Roughly speaking the interference cancellation methods can be divided into one-user detection methods and multi-user detection methods (MUD). In the one-user detection methods each user is provided with a specific correla- tor group comprising a correlator for one or more multi-path propagated components. A received combination signal including the information of all users is multiplied in each correlator by the spreading code of the user, and in an optimal situation the multiplication brings about the signal of the user. Then the signals of other users are not utilized in user signal detection. As to the multiuser detection methods, the user signal detection utilizes the information obtained from the other user's signals in order to improve the identification of said user signal.

Figure 2 describes a preferred embodiment of the method of the in- vention. In an initial step 200 of the method a radio network using code division multiple access (CDMA), such as a mobile communication network, comprises several transmitters such as mobile phones that transmit information simultaneously on the same frequency band. The information sent by the mobile stations consists of symbols including a speech or data transmission. The symbols are spread onto a broad band by multiplying the user symbols by a spreading code allocated to a user connection. The information sent by the mobile stations is received in a CDMA receiver as a combination signal in step 202, the combination signal comprising one or more multipath propagated signal components of all the users employing the network at that particular mo- ment. In a preferred embodiment the receiver is a RAKE-type receiver located in the base station, in which possibly several multipath propagated components are received from each user. Said components are combined in the receiver in order to allow the best possible signal reception.

A broadband analogue signal is at first converted in the receiver into digital mode and a first detection stage of the user signal is carried out in accordance with step 204. In the detection stage the broadband signal is multiplied by the spreading code of the user, and then in an optimal situation the first symbol estimates can be generated from the symbols sent in the user signal. In a preferred embodiment, a detection stage does not precede the inter- ference cancellation stage, instead zeroes are used as the first user signal estimates. The detection stage is followed by one or more interference cancellation stages, where the interference caused by other users is cancelled from the desired user signal. The user signals are regenerated in each interference cancellation stage in accordance with step 206. Regeneration is carried out using the symbol estimates, channel estimates and the spreading codes of the users generated in the detection stage or in the preceding interference cancel- lation stage. The broadband user signals thus generated are reduced from the original broadband signal using parallel interference cancellation, which is described in more detail in Figure 4.

In accordance with a preferred embodiment of the invention user signals are scaled during the regeneration stage in order to carry out partial interference cancellation in accordance with step 208. In the regeneration stage the scaling coefficient k is used to scale the channel estimate to be fed into the regeneration stage in a preferred embodiment. This refers to, for example, a RAKE receiver that searches the multipath propagated compounds from the user signal, each component having a particular delay and certain amplitude. In a real situation the response of the transmission channel is continuous, in which the amplitude can be shown as the function of the delay. In practice, an impulse response figure is formed of the real response, to which figure for example the five strongest real impulse peaks referred to as impulse response taps are selected. For example, the phase shift and amplitude of one tap that the invention scales are fed into the regenerator of the user signals. In this context, scaling refers to multiplication by a real number, which preferably exceeds 0 but remains below 1. As the regeneration stage generates the user signals as a product, where the factors preferably comprise at least the data bit estimate, spreading code and the amplitude of the channel estimate, it is not significant which ones of said factors are thought to be scaled.

The regeneration stage of the user signals provides as an output a spread user signal for each user, and the user signals are combined to generate an interference signal in accordance with step 210. In a preferred em- bodiment of the invention each user signal obtained during the regeneration is separately scaled using coefficient k or alternatively the interference signal generated from the user signals is scaled using said scaling coefficient k. In step 212 the generated broadband interference signal is reduced in the CDMA receiver from the received broadband combination signal including all the user signals.

A preferred embodiment of the invention is described above. Steps 204 to 212 form an interference cancellation stage. One or more interference cancellation stages can be carried out for a user signal, in which case the scaling coefficient can also be changed between the interference cancellation stages. In the following, the invention will be described with reference to Figure 3, where a block diagram shows a CDMA transmitter and receiver by means of a preferred embodiment. The transmitter is described using blocks 300 to 310 and the receiver using blocks 330 to 340. Data block 300 de- scribes the parts of the receiver required to form the user's speech and data in the receiver. These parts include for example the keypad or microphone of a mobile phone. When the information composed of symbols is obtained from the user, channel coding or interleaving is applied thereto in block 302. The channel coding and interleaving ensure that the information sent can be re- trieved in the receiver, even if all information bits could not be received. Block 304 describes how the multiplication by the spreading code and the spread into broadband form are carried out for the information to be sent. The conversion from digital into analogue mode is performed in block 306, and thereafter the information is transferred to be sent by antenna parts 310 after the radio frequency parts 308.

The user information is sent to the receiver using radio channel 320, which refers to a logical radio channel to be sent on a physical channel. The logical channels of the digital radio systems can in principle be divided into two groups: control and traffic channels. Some of the radio channels are placed in the uplink direction from the terminal towards the cellular radio system, whereas others are placed in the downlink direction from the mobile phone system towards the terminal. The control channel does not allocate radio resources for use, but attends to matters associated with the use of the mobile communication system, such as paging terminals on a paging channel (PCH) that is common to all terminals. In the uplink direction one control channel is for example a random access channel (RACH) by which the terminal requests the network to establish a connection. Radio resources are allocated to the terminal for the actual traffic channels depending on the need for data transmission. A logical traffic channel is DCH (Dedicted Channel) that is used to transfer information from the radio system to the terminal. The UMTS radio system comprises a number of other channels, but the explanation thereof is not relevant in this context.

The frame and burst structures to be used on the physical channels deviate from one another depending on which physical channel is used for transmission. A frame refers here to an entity included in several bursts, in which case one user can be provided with for example the second time slot of the frame for transmitting the burst. An example of such a frame is the PDPCH physical channel frame of the TDD mode in the UMTS, the length of said frame being 10 milliseconds. The frame is divided into sixteen time slots, and the length of each time slot is 0.625 milliseconds. The data packet to be sent in a time slot is referred to as a burst. One such burst may include 2560 chips of information, from which 0 to 1103 include data, and chips 1104 to 1359 comprise the midamble, and chips 1360 to 2463 include data and a 96 chips long guard period is placed at the end of the burst. The data in the midamble is often referred to as the training sequence or the pilot. The training sequence is a set of symbols known by the receiver and said sequence is transmitted to the terminal from the network for example on a forward access channel (FACH) before the actual connection establishment. The training sequence received by the terminal can be used during the connection both in the uplink and downlink direction, but different training se- quences can also be used in the different transmission directions. The receiver, which may be a terminal or a base station, receives bursts on the channel, it generates a channel estimate on the basis of the burst's training sequence. The channel estimate allows the receiver to evaluate how the radio path has distorted the data contents of the burst. Based on the information obtained the receiver may try to correct the data contents of the burst according to the channel estimate utilizing known methods. The training sequence and the impulse response formed thereof may be used to evaluate the channel quality utilizing known methods, such as the C/l ratio (Carrier/Interference), SIR (Signal Interference ratio), bit error rate, or by studying the ratio between the chip energy and the interference power density EJ\₀.

Figure 3 shows a CDMA receiver comprising, in turn, antenna parts 330 for receiving a broadband signal. From the antenna 330 the signal is transferred to radio frequency parts 332, from where the signal is transferred to an A/D converter 334 for carrying out a conversion from analogue mode into digital mode. In a reception block 336 the user signal is separated from the received CDMA signal. The separation occurs for example by composing symbol estimates from the user signal, which symbol estimates can be improved by applying one or more interference cancellation stages to the information. In a preferred embodiment the reception block 336 of the RAKE re- ceiver comprises a delay estimator used to estimate the delays of the mulipath propagated components and to allocate the strongest thereof to the RAKE branches. The user signals are regenerated in the reception block 336 and combined into an interference signal, which can be reduced from the received combination signal. Interference cancellation is preferably implemented at least once. The practical simulations show that the second and even the third interference cancellation round improve the signal quality, but thereafter the improvement in signal quality is not significant. When the final symbol estimates are generated from the signal, said signal is directed to a block 338 eliminating the channel coding and the interleaving of the symbols. Thereafter the user data is directed in the receiver to data processing routines 340, which in the case of a base station refers to transferring data to the base station controller and from there further to a fixed telephone network.

Figure 4 describes the principles of a parallel interference cancellation unit of the CDMA receiver. The interference cancellation unit is placed into the reception unit 336 of the receiver shown in Figure 3. The broadband digital signal is received in a first detection stage 400. A delay unit 404 delays the broadband signal or the samples taken from there so that the delay caused by the interference cancellation stages such as regeneration and composing are taken into account. A correlator group 406A correlates the multipath propagated components of user 1 , group 406B the multipath propagated compo- nents of user 2 and so on. When the first symbol estimates are generated in the detection stage, the user signals are processed in one or more interference cancellation stages 402. A regeneration block 408A regenerates the signal components of user 1 in a broadband mode and directs them to summers 410B to 41 ON. As Figure 4 shows, the spread signals of all other users except user 2 are directed to the summer 410B for generating an interference signal. The interference signal generated in the summer 410B is reduced from the originally received broadband combination signal in a reductor 412B, and a purged signal of user 2 is obtained as the output of said reductor. The purged signal is directed to new correlators in the next detection stage 414A, where new symbol estimates 416A are generated for user 2. New symbol estimates 416A to 416N are accepted as the final symbol estimates or they are used in the following interference cancellation stage in connection with regeneration.

The apparatus according to the receiver unit 336 of the receiver of the invention is explained in the following by means of a preferred embodiment with reference to Figure 5. Samples received in an input buffer 500 are formed of the broadband digital combination signal 320. The buffer 500 receives the combination signal and feeds signal to the interference cancellation as rapidly as the signal can be processed. The signal samples are stored in a sample memory 506. In the sample memory 506 the oldest set of samples is always replaced after an interference cancellation stage with a set of samples to be read from the input buffer 500. A controller 504 coordinates the operation of the sample memory 506, i.e. the controller 504 comprises for example storing means for storing samples into the memory, reading means for reading the samples from the memory and means for removing samples from the sample memory. The receiver unit 336 further comprises a control unit 508, which monitors in accordance with what is shown in Figure 4 the identification of symbol limits in the correlators 414A to 414N, and in the regenerators 408A to 408N. The control unit 508 of the symbol limits is also controlled by the controller 504, which also controls the operation of the input buffer 500. Figure 5 further shows a correlation memory 528, wherein the latest symbol estimates of the users are stored. A symbol processor 524 carries out several tasks, such as combining the multipath propagated components, making symbol decisions, channel estimation and controlling spreading code generators 512. The symbol processor 524 preferably also coordinates interference cancellation and comprises for example means for repeating the interference cancella- tion steps, whereof each one comprises the despreading, regeneration and scaling of user signals, the generation of an interference signal and the reduction thereof from the spread spectrum signal. Symbol estimates generated in the detection stage that precedes the interference cancellation stages or which estimates are zeroes are initially obtained as inputs in the symbol processor 524. The final symbol estimates 416A to 416N are obtained as outputs from the symbol processor 524 from the interference cancellation stages. The residual signal samples are directed to an output buffer 502.

The apparatus according to Figure 5 is used to implement the method of the invention shown in Figure 2, and the scaling of the invention is carried out in a multiplication and adding unit 514. The multiplication and adding unit 514 comprises means for generating a scaling coefficient for an interference cancellation round. The scaling coefficient can be kept constant using means for keeping the scaling coefficient constant on consecutive interference cancellation rounds or the coefficient can be increased using means for in- creasing the scaling coefficient between the interference cancellation rounds. The coefficient can for example be increased linearly so that the coefficient is 0.40 in the first interference cancellation stage, 0.60 in the following and so on, approaching the value 1. The multiplication and adding unit 514 comprises in a preferred embodiment of the invention means for altering the scaling coefficient between the consecutive interference cancellation stages on the basis of the signal-to-interference plus noise ratio. If the signal-to-interference ratio of the detective user is high, the scaling coefficient k is low. In reverse, this means that if the signal-to-interference plus noise ratio is low, it is not preferable to use an interference signal for interference cancellation and then the value of k is low. The multiplication and adding unit 514 also comprises means for generating a scaled interference signal as the output of the unit using the spread user signals and the scaling coefficient k. Scaling is carried out in a preferred embodiment of the invention for the added interference signals. Scaling can also be performed for the regenerated user signals before the addition. Furthermore, in a preferred embodiment of the invention the channel estimate is scaled. The multiplication and adding unit 514 also comprises means for increasing the scaling coefficient between the interference cancellation rounds, whereto the unit obtains the control from the symbol processor 524. The multiplication and adding unit 514 further comprises means for increasing the scaling coefficient between the interference cancellation rounds and means for scaling the amplitude of the user signal generated during regeneration. The unit 514 also comprises means for scaling the channel estimate during regeneration. Thus, the significance of the later interference cancellation rounds increases, what is preferable to the interference cancellation, since the symbol estimates are better on the later interference cancellation rounds than on the earlier interference cancellation rounds. The multiplication and adding unit 514 further comprises means for increasing the scaling coefficient linearly between the interference cancellation rounds. The scaled interference signal is reduced from the combination spread spectrum signal in the reductor 516. The means for carrying out the invention are preferably implemented by software, whereby the receiver placed in the base station 100A to 100D includes a microprocessor, in which the operations of the method described are implemented. The invention can also be implemented for example using an apparatus solution offering the required functionality, such as an ASIC (Application Specific Integrated Circuit), or utilizing separate logic components. It is obvious that the receiver also comprises other parts than those described in Figures 3 to 5, but the description thereof is not relevant in this context.

Figure 6 shows the results obtained in interference cancellation using the method of the invention. In the Figure, an x-axis 600 shows the ratio of the user signal strength Es to the interference noise NO. The scale is a logarithmic decibel scale, where dB = 10log₁₀(Es/N0). Thus, for example, a situation where the user signal ratio to noise is 0.63 corresponds with value -2. A y- axis 602 shows a FER (Frame Error Rate) that describes the number of the received frames including erroneously interpreted information. Simulation is performed 8 for the users. The spreading code used in said simulation is 8 chips long. A curve 604 shows a simulation, where the scaling coefficient is 1 and where the receiver does not include any interference cancellation stages. A curve 606 shows a situation in which the scaling coefficient is 0.7 without any interference cancellation stages. Curves 608 and 610 show the depend- ence of the signal-to-interference ratio and the frame error rate after two interference cancellation rounds. In curve 608 the scaling coefficient k is 1 , i.e. the scaling has not been performed and in curve 610 the scaling coefficient is 0.7 during both interference cancellation rounds. The Figure shows that the difference between curves 608 and 610 is approximately 0.15 dB at both frame er- ror rates. In other words the scaling of the invention provides for example in a situation like simulation a significant improvement to the signal-to-interference- ratio.

Even though the invention has above been explained with reference to the example in the accompanying drawings, it is obvious that the invention is not restricted thereto but can be modified in various ways within the scope of the inventive idea disclosed in the attached claims.

Claims

1. A method for carrying out parallel interference cancellation from a spread spectrum signal, the method comprising the steps of

(202) receiving a combination signal including user signals spread with several spreading codes, (204) composing the user signals of the combination signal in order to generate user signal estimates, (206) regenerating the spread user signals using the user signal estimates, characterized by

(208) generating a scaling coefficient k in order to scale the interference signal to be reduced from the combination signal, (212) generating the scaled interference signal using the spread user signals and the scaling coefficient k,

(214) subtracting the generated scaled interference signal from the combination signal in order to reduce the multi user interference of the user signal.

2. A method as claimed in claim 1, characterized in that said scaling coefficient exceeds 0 but remains below 1.

3. A method as claimed in claim ^characterized in that the multi user interference is cancelled from the combination signal in one or more interference cancellation stages, each one of which comprising the despread- ing, regeneration and scaling of the user signals, the generation of an interference signal and the reduction of the interference signal from the combination signal.

4. A method as claimed in claim 3, characterized in that the scaling coefficient is kept constant on the consecutive interference cancella- tion rounds.

5. A method as claimed in claim 3, characterized in that the scaling coefficient is increased between the consecutive interference cancellation rounds in order to increase the significance of the interference to be cancelled during later interference cancellation rounds.

6. A method as claimed in claim 3, characterized in that the scaling coefficient is increased linearly on the consecutive interference cancellation rounds.

7. A method as claimed in claim 3, characterized in that the scaling coefficient is altered between the consecutive interference cancellation rounds on the basis of the signal-to-interference plus noise ratio.

8. A method as claimed in claim ^characterized in that the amplitude of the user signal generated during regeneration is scaled.

9. A method as claimed in claim 1, characterized in that the channel estimate is scaled during regeneration.

10. A receiver for receiving a spread spectrum signal comprising one or more antennas (330) for receiving a combination signal including user signals spread with several spreading codes, one or more correlators (414A to 414N) for composing the user signals from the combinations signal in order to generate user signal estimates, one or more regenerators (408A to 408N) for regenerating the spread user signals using the user signal estimates, characterized in that the receiver comprises means (514) for generating a scaling coefficient k in order to scale the interference signal to be reduced from the combination signal, means (514) for generating the scaled interference signal using the spread user signals and the scaling coefficient k, a reductor (516) for subtracting the generated scaled interference signal from the combination signal in order to reduce the multi user interference of the user signal.

11. A receiver as claimed in claim 10, characterized in that said scaling coefficient exceeds 0 but remains below 1.

12. A receiver as claimed in claim 10, characterized in that the receiver comprises means for repeating the interference cancellation stages, each one of the interference cancellation stages comprising the de- spreading, regeneration and scaling of the user signals, the generation of an interference signal and the reduction of the interference signal from the combination signal.

13. A receiver as claimed in claim 12, characterized in that the receiver comprises means for increasing the scaling coefficient between the consecutive interference cancellation rounds in order to increase the sig- nificance of the interference to be cancelled during later interference cancellation rounds.

14. A receiver as claimed in claim 12, characterized in that the receiver comprises means to increase the scaling coefficient linearly on the consecutive interference cancellation rounds.

15. A receiver as claimed in claim 12, characterized in that the receiver comprises means to keep the scaling coefficient constant on the consecutive interference cancellation rounds.

16. A receiver as claimed in claim 12, characterized in that the receiver comprises means for altering the scaling coefficient between the consecutive interference cancellation rounds on the basis of the signal-to- interference plus noise ratio.

17. A receiver as claimed in claim 10, characterized in that the receiver comprises means for scaling the amplitude of the user signal generated during regeneration.

18. A receiver as claimed in claim 10, characterized in that the receiver comprises means for scaling the channel estimate during regeneration.