JP2005260991A - Interference removal system and interference removal method of cdma receiver - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To provide an interference removal system and an interference removal method in a CDMA receiver, which can exhibit better interference removal performance. <P>SOLUTION: The system is provided with a plurality of steps 310, 320, 330. Each step is provided with a estimator 312, which provides a estimation signal of the interference that each user (k) receives from other user in the CDMA communication system, a combining means 226 to provide a signal obtained by combining the received input signal and the interference estimating signal, a MAP decoder 240 which is configured to receive the obtained signal after synthesis and generates a decoder output signal. The plurality of steps 310, 320, and 330 are connected so that the decoder output signal from a previous step is applied to the estimator 312 in the next step. The received input signal is applied directly to the each step's combining means 226. <P>COPYRIGHT: (C)2005,JPO&NCIPI

Description

  The present invention relates to a mobile communication system, and in particular to code division multiple access (CDMA), and is directed to the interference cancellation system of a CDMA receiver according to claim 1. The present invention further relates to an interference cancellation method for a CDMA receiver according to claim 5.

  The capability of the CDMA system is limited mainly by the interference experienced by each user in the received signal. In order to increase the capacity in a cell of a CDMA communication system, it is necessary to remove interference as much as possible.

  Also, in a system with high interference, it is necessary to increase the spreading factor in order to maintain a predetermined quality of service (QoS). However, an increase in spreading factor is directly linked to a decrease in the bit rate available for data transmission.

  In the field of CDMA communication systems, the following method is adopted on the receiver side to alleviate the problem of interference.

  As a well-known method, there is a method of collectively detecting received signals of a plurality of users. However, the collective detection method has a drawback in that it requires transposition of a matrix that expands as the number of users increases. For example, if the number of users in a cell is 64, it is necessary to transpose a 64 × 64 matrix, which is very complicated and requires excessive power and time to process.

  Another well-known method is the single user detection method, which is one of the simplest detection methods. This method detects a single user ignoring all other signals. Therefore, even if the result is optimal for one user, the performance in a multi-user environment, that is, the bit error rate is significantly deteriorated.

  Journal of the Institute of Electrical and Electronics Engineers, Vol. 17, no. 12, pp2082-2089, “Repetitive multiuser detection applying antenna train and FEC to multipath channel”, multiple access communication reception equipped with iterative receiver that improves initial estimation by antenna train in multipath propagation environment The machine system is described.

  The third part of this paper details the structure of the iterative receiver. For this purpose, the iterative receiver receives the channel output and performs a combination of soft interference cancellation and a matched filter. The interference canceller performs priority probability estimation (priori probabilities) used to construct a soft estimate of the symbol for each user channel symbol. Next, an interference removal operation in the diffusion region is performed.

  However, there is a need for an interference cancellation system and method that has simpler and sufficient performance in a multi-user environment.

  The present invention has been made in view of this point, and an object thereof is to provide an interference cancellation system and an interference cancellation method in a CDMA receiver that can exhibit better interference cancellation performance.

  This object is achieved by an interference cancellation system having the function according to claim 1 and an interference cancellation method comprising the steps of claim 5.

  That is, the interference cancellation system in the CDMA receiver of the present invention includes a plurality of stages, and each stage (310, 320, 330) receives an estimated signal of interference that each user k receives from other users of the CDMA communication system. An interference estimation means (312) to be provided, a synthesis means (226) to provide a signal obtained by synthesizing the received input signal and the interference estimation signal, and a signal obtained by the synthesis are received. And a decoding means (240) for generating a decoder output signal, wherein the stage (310, 320, 330) includes the decoder output signal of the previous stage to the interference estimation means of the next stage. It is connected as applied, and the received input signal is applied directly to the synthesis means at each stage.

  The interference cancellation method in the CDMA receiver according to the present invention includes a step of providing an interference estimation signal by estimating an interference received by each user k from another user of the CDMA communication system, and a received input signal and the interference estimation signal. A step of providing a signal obtained by synthesizing, a step of decoding the signal obtained by the synthesis to generate a decoder output signal, and a next step based on the decoder output signal of the previous step And a step of directly applying the received input signal to each stage and synthesizing it with the interference estimation signal.

  Embodiments of the invention are constrained by various dependent claims.

  According to the present invention, better interference removal performance can be exhibited.

  Embodiments of the present invention will be described in detail with reference to the following drawings.

  FIG. 1 shows a basic configuration of the transmitter 100, and each user's signal coming from the source 101 is a forward error correction (FEC) code or the like that can be composed of a block code, a convolutional code, or other codes. Encoding is performed using an encoder 102 using an outer code.

  Subsequently, the encoded signal is spread using an internal code that can be constructed from pseudo-noise codes, orthogonal codes, or other codes and combinations thereof, to facilitate multi-channel connections. The internal code generally differs from user to user. The spread modulator 103 is used to modulate the output signal from the encoder 102 and transmits the modulated signal through a communication channel.

  The detailed structure of the transmitter is easily understood by those skilled in the art and need not be described in detail here.

  A basic structure of the receiver 200 is shown in FIG. Here, the received input signal is demodulated by the demodulator 201 and despread by the despreader 202 using the corresponding internal code for each user. After despreading the signal, the signal is input to the interference cancellation and decoding stage 210, details of which are shown in FIGS.

  Next, the output signal from the interference cancellation and decoding stage 210 is obtained using information about the inner code and information contained in the outer code. By removing the interference from the received input signal, the system capacity and data rate within the cell can be increased.

  FIG. 3 is a diagram showing a first embodiment of the interference cancellation system according to the present invention. The interference cancellation and decoding stage 210 consists of a multi-user detector 220 having a first multiple input 221 that receives the receiver input signals of n users of a CDMA communication system. Multi-user detector 220 also includes a second multiple input 222 that receives the output signal of selection switch 260, the purpose of selection switch 260 being described in detail below.

  Multi-user detector 220 includes a plurality of outputs 223 that output detector output signals. Details of the multi-user detector 220 will be described with reference to FIG.

The detector output is connected to a log likelihood calculator 230 that calculates the log likelihood ratio of the applied signal according to the following equation:

  The output of the log likelihood calculator 230 is connected to the input of a MAP decoder 240 that is a soft input / soft output decoder that decodes data information of the received input signal. Further, the decoder 240 outputs the log likelihood ratio estimated for all n users to the input port B of the selection switch 260 using the feedback loop 250.

Next, the structure of the multiuser detector 220 will be described in detail with reference to FIG. The multiple estimated log likelihood ratios ε _k are received at the second multiple detector input 222 and applied to the scaling means to make a tentative decision on the transmitted signal using the tanh function. Therefore, the signal input to the scaling means in the range from −∞ to + ∞ is enlarged / reduced to a signal that can be regarded as an average estimated value of the log likelihood ratio of the input signal in the range from −1 to +1.

The exact mathematical calculation of the scaling means 224 is:
U _k = tanh (0.5 · ε _k )
However, the following relationship has been found to be advantageous by simulation.
−1 for ε _k <−2
U _k = ε _k / 2 for −2 ≦ ε _k ≦ + 2
+1 for ε _k > +2

The output of the scaling unit 224 is connected to a corresponding input of the adding unit 225 that adds the enlarged / reduced signals, and obtains an estimated value of interference that the user k receives from all other users. For this reason, the signal applied to the input of the adding means is first enlarged / reduced by the respective estimated amplitudes before addition, and then expanded by the cross-correlation coefficient ρ _ik between the user i signal and the user k signal. to shrink.

Therefore, the adding means 225 performs the following mathematical operation.
ζ _k = ΣA _i · ρ _ik · U _i
Here, ζ _k is an interference estimated value, A _i is a channel estimated value and includes information on attenuation and phase rotation in the channel, and U _i is an output signal from the scaling means 224.

Finally, the detector includes a combining means 226 for combining the output signal from the input signal r _k and the adding means 225 received. The output of the combining means 226 is provided to a plurality of detector outputs 223.

  Finally, the interference cancellation method in the CDMA receiver will be described in detail.

  The received input signals of n users of the communication system are simultaneously provided to the first multiple detector input 221 and the second multiple detector input 222 via the selection switch 260 during the first iteration. In this case, the selection switch 260 is connected to the input port A and guides the input signal as an output signal of the switch. Thereafter, at each further iteration, the selection switch 260 selects the input port B and outputs the input signal as an output signal.

In multi-user detector 220, the signals provided at the first and second multiple inputs are logically combined to provide an improved signal y _k for subsequent processing. For this reason, the signal at the second multi-detector input 222 is weighted in consideration of the cross-correlation characteristics of the internal code of the adding means 225, including the scaling of the scaling means 224, and an estimate of the received interference is obtained. . Thus, the interference is removed by simply subtracting the weighted signal from the originally received input signal. This is executed by the adding means 226.

  Therefore, the interference removal itself is a subtraction of a weighted signal from the originally received signal. In this way, it is easy to realize interference removal.

  It should be noted that an improvement over conventional systems is obtained even if the iteration is only once, i.e. without using the outer code information provided by the decoder. Since interference cancellation removes more errors than introduced by it, this process can be repeated several times and a subsequent estimate of received interference is obtained from the output of the MAP decoder.

  The output signal of the multiuser detector 220 is applied to the MAP decoder 240 via the log likelihood calculator 230 as described above. The soft output of the decoder provides information bits for use in decoding the information data, while providing code bits that are fed back to the selection switch 260 via the feedback loop 250. Although the need for soft decision output values at the decoder suggests the use of those derived from them, such as convolutional codes or turbo codes, the outer code used by decoder 240 is any forward error correcting code. It should be noted that this method may be used.

  The decoder output signal fed back to the input port B of the selection switch 260 contains the outer code, ie the information contained in the improved decoded signal for each user k. Each time this iteration is performed, the selection switch selects port B, and the signal applied to the second multiple decoder input 222 corresponds to the decoder output signal from the previous iteration.

According to an improved version of the interference cancellation system not shown, instead of calculating the cross-correlation matrix formed by the cross-correlation coefficient ρ _ik that forms the interference estimate ζ _k , the spreading process is used to output the signal from the scaling means 224 It is also applicable to. In this case, the tanh operation is performed on the despread signal, and the provisional determination is repeated before adding to all the users in the adding means 225. According to this refinement, the output signal from the summing means 225 performs the interference cancellation with respect to spread input signal r _k received is inputted to the synthesizing means 226. In other words, the input signal r _k received is not despread shall apply despread the output signal from the combining means 226 by using a spreading code for user k. Also, the feedback signal must be despread before entering the scaling means 224 for initial estimation. In this case, the adding means 225 performs scaling only by the channel estimation A _i before the addition.

  This refinement is particularly important when dealing with asynchronous systems, ie systems in a multirate environment where user data does not arrive at the receiver in time series and / or where different users are operating with different spreading factors.

  Another refinement for making a provisional decision at the receiver is an arbitrary monotonic function that maps input values in the range from -∞ to ∞ to output values in the range from -1 to +1 instead of using the tanh function. Yes, even if it is easier to implement, it may not be optimal in performance.

  In addition to the spreading code, another scrambling code can be applied to the transmission signal. At the receiver, this will either calculate a cross-correlation matrix that combines the spreading and scrambling codes, or include a descrambler after the respreader and before the despreader and re-scrambler.

  In an actual wireless environment, particularly a mobile communication environment, signals are affected by various signal fading and multipath propagation. In order to mitigate such an effect, it is known to introduce a structure that introduces an interleaver in the transmitter, a deinterleaver in the receiver, or a multipath solution in a receiver such as a RAKE receiver.

  Such a new optional function can be installed without any problem in the above-described embodiment of the present invention. In fact, by introducing a different interleaver structure for each user, performance improvement, that is, a large gain can be expected at each iteration.

  For multipath receivers, the removed signal can be used again to act as a RAKE receiver in the improved signal. This results in improved channel estimation, improved signal estimation, and less interference between different RAKE fingers of the same user.

  A second embodiment of the interference cancellation system will be described with reference to FIG. Here, the interference cancellation apparatus is realized as a multistage system including a plurality of stages having the same structure. Further, the constituent elements corresponding to the constituent elements of the first embodiment, that is, the synthesizing means 226 for performing interference cancellation and the MAP decoder 240 are assigned the same reference numerals.

  As shown in detail in the figure, the first stage 310 includes a despreader 311, an estimator 312, and a respreader 313 in order. Further, a despreader 314 and a log likelihood ratio calculator λ (reference number 315) are connected in series between the combining means 226 and the MAP decoder 240.

  The second stage 320, the third stage 330 to the Nth stage have the same structure as the first stage except that the despreader 311 is omitted. Further, the estimator 312 receives the output signal from the MAP decoder 240 as its input instead of the demodulated input signal from the demodulator 201.

  After the last Nth stage, the output of the decoder is regarded as decoded information data.

  It should be noted that FIG. 5 is a simplified diagram and each arrow represents a full vector of the respective signal shown in FIGS.

  Also, in this embodiment, the interference cancellation system has more errors to cancel than incoming errors, and the interference cancellation gain is improved by repeating the interference cancellation process using multiple stages.

  FIG. 6 shows an embodiment of the present invention using the above-described turbo code. The upper half of the figure shows the basic structure of a turbo decoder that includes blocks 401 to 406 and provides a decoder decision output 413. The lower half of the figure shows a process in which the feedback signal 414 is calculated from the results obtained by the MAP decoders 402 and 404. Since turbo decoders are known in the art, those skilled in the art can understand the details of the decoding operation even if they are omitted.

  In short, the signals obtained from the log likelihood calculation stage 401 are input to the first and second MAP decoders 402 and 404, respectively. These signals represent the systematic information that the MAP decoder receives at each first input. Parity information is received at each second input and an information bit metric signal and a sign bit metric signal are provided at the output.

Specifically, for decoder 402, information bit metric 1 is combined with the decoder input signal. The obtained signal is input to the second MAP decoder 404 where it is combined with the systematic information signal from the log likelihood calculator 401. Similarly, at the output of the decoder 404, the external metric signal is fed back to the input of the first MAP decoder 402, where it is combined with the systematic information signal from the log likelihood calculation stage 401. However, the signal of the second MAP decoder 404 is subjected to an interleaving operation (block 403), and the input signal is interleaved according to a predetermined method before being input to the decoder. In addition, an inverse π ^-1 operation (block 405) is performed on the information bit metric signal at the output of the decoder. After the inverse π ⁻¹ operation, the output information bit metric signal is input to the threshold determination block 406, which outputs +1 or −1 depending on the sign of the input signal, and then as the decoder determination signal 413 used.

  Basically, in a rate 1/3 turbo code, of the three bits, the first bit is a systematic bit, the second bit is a parity 1 bit, and the third bit is a parity 2 bit. Systematic information from the MAP decoder 404 is employed for feedback. This is because it is generally an improvement over the MAP decoder 402. However, since this systematic information is interleaved for the MAP decoder 404, the interleaving process is reversed at step 409 for the calculation of systematic feedback. For parity 1 feedback, parity 1 information from the MAP decoder 402 is employed. Parity information from the MAP decoder 404 is adopted for the parity 2 feedback.

  Also, the sign bit metric signals 411, 412 are obtained from the MAP decoders 402, 404 and arranged as two sets of series, with the first bit representing the systematic metric and the second bit representing the parity metric. Accordingly, at block 407, the signal 411 is divided into the systematic and parity metrics of the MAP decoder 402. Similarly, at block 408, signal 412 is divided into systematic and parity metrics for MAP decoder 404.

  These derived metric signals are multiplexed at block 410 so that each of the three feedback information signals 414 follows a systematic / parity 1 / parity 2 order.

  FIG. 7 shows a RAKE receiver to which the interference cancellation system of the present invention is applied. The RAKE receiver is known in the technical field, and its basic structure is well known to those skilled in the art, so a detailed description thereof will be omitted.

  Blocks 501 to 504 represent the first stage and replace demodulator 201 in connection with despreader 311 of FIG. The interference estimation block 505 is the same as the estimator 312 of FIG. Also, the respreader 313 is replaced with blocks 506 and 507 representing the respreading operation and the channel repetition operation, respectively. Further, the synthesizing unit 226 in FIG. 5 corresponds to the synthesizing unit 508. The despreader 314 of FIG. 5 is replaced with blocks 509-512.

  In operation, the interference estimation block 505 receives the signal from the maximum ratio combining means 504 as its input signal, and in block 507 forms a channel replica from the received signal estimates such as radio channel multipath and fading characteristics. Thus, information from finger assignment block 502 is needed to determine the finger delay. Similarly, information from the channel estimation block 503 is necessary to reproduce the fading characteristics of each path.

  The output of the maximum ratio combining means 512 is used as an input to the first stage block 315. The outputs of channel estimation block 511 and finger assignment block 510 are input to a second stage channel duplication block 507. In the next stage, the following changes occur compared to the first stage described above. Blocks 501 to 504 are not executed. In the despreader 507, the signals from the previous blocks 510 and 511 are used in place of the signals of the blocks 502 and 503. Finally, block 505 uses the output of the MAP decoder 240 from the previous stage as an input if present.

  Alternatively, interference estimation block 505 may also include several different blocks to estimate interference. For example, if pilot signals are transmitted, the receiver knows them completely. Thus, instead of making an estimate from the received symbols corresponding to those pilot symbols, the receiver uses the original pilot symbols because it is certain that those symbols have been transmitted. Therefore, the interference estimate for these symbols is most accurate.

The figure which shows the basic structure of the transmitter of a CDMA communication system The figure which shows the basic structure of the applicable receiver of a CDMA communication system. The figure which shows 1st Embodiment of the interference removal system which concerns on this invention. The figure which shows the more detailed structure of the detector used in embodiment shown in FIG. The figure which shows 2nd Embodiment of the interference cancellation system of this invention in a multistage structure. The figure which shows embodiment of this invention using a turbo code | symbol The figure which shows the RAKE receiver which introduced the interference cancellation system of this invention

Claims (7)

It has a plurality of stages, and each stage (310, 320, 330)
Interference estimation means (312) for providing an estimated signal of interference that each user k receives from other users of the CDMA communication system;
Combining means (226) for providing a signal obtained by combining the received input signal and the interference estimation signal;
Decoding means (240) configured to receive a signal obtained by the synthesis and generating a decoder output signal;
The stages (310, 320, 330) are connected such that the decoder output signal of the previous stage is applied to the interference estimation means of the next stage, and the received input signal is Applied directly to the synthesis means,
An interference cancellation system in a CDMA receiver. The first stage (310) is
First despreading means (311) configured to receive an input signal and apply it to said interference estimation means (312);
Respreading means (313) for respreading the interference estimation signal;
The interference cancellation system according to claim 1, further comprising: Each stage (310, 320, 330)
Second despreading means (314) for despreading the signal obtained from the combining means (226);
Log likelihood calculating means (315) for calculating a log likelihood ratio and applying it to the decoding means (240);
The interference cancellation system according to claim 1, further comprising:   The interference cancellation system according to claim 1, wherein the decoding means is a soft input / soft output decoder (240) such as a MAP decoder. Estimating the interference each user k receives from other users of the CDMA communication system and providing an interference estimation signal;
Providing a signal obtained by combining the received input signal and the interference estimation signal;
Decoding a signal obtained by the synthesis to generate a decoder output signal;
Estimating interference of the next stage based on the decoder output signal of the previous stage;
Applying the received input signal directly to each stage and combining it with the interference estimation signal;
A method for canceling interference in a CDMA receiver, comprising: Despreading the input signal before generating the interference estimation signal;
Respreading the interference estimation signal before combining with the input signal;
The method of claim 5, further comprising: Despreading the signal obtained by the synthesis;
Calculating a log likelihood ratio before decoding the despread synthesized signal;
The method of claim 5, further comprising:

Images