A radio receiver in a wireless communication system

Abstract

A method for processing radio signals in a wireless radio receiver for detecting an n-beam channel condition of a propagation channel, the method comprising: Receiving signal samples and estimating multiple channel taps from the samples based on a model of the propagation channel as a delay line with multiple channel taps; Estimating signal power and interference power for each of the estimated channel taps; Filtering the signal power to provide a filtered signal power measurement; Filtering the interference power to provide a filtered interference power measurement; Using the filtered power values ​​to determine the n strongest channel taps; Generating a first comparison parameter by determining the ratio of the filtered signal power quantity and the filtered interference power quantity for the n strongest channel taps, and generating a second comparison parameter by determining the ratio of the filtered signal power quantity and the enhanced interference power quantity for the at least one other channel tap, wherein the ratios are filtered to generate the corresponding first and second comparison parameters; Providing a comparison result based on the first and second comparison parameters and a threshold value, and Identifying an n-ray channel condition of the propagation channel from the comparison result.

Description

[0001] The present invention relates to a radio receiver in a wireless communication system and a method for processing radio signals.

[0002] The transmission of radio signals carrying data in modern wireless communication can be achieved using a variety of different communication systems, often specified by a standard. There is a growing demand for devices capable of operating in a way that supports more than one of these wireless communication systems. Mobile radio receivers contain analog high-frequency (HF) / intermediate-frequency (IF) stages designed to receive and transmit wireless signals via one or more antennas. The output signal of the HF / IF stages is typically converted to the baseband, where an analog-to-digital converter (ADC) converts incoming analog signals into digital signal samples. These samples are then processed for signal detection and decoding of the data into logical values.The ADC can alternatively operate directly at the intermediate frequency (IF), in which case the conversion to the baseband is performed digitally. Several different methods are known for processing the input digital samples to establish signal detection, including rake-receiver processing (also known as computational-receiver processing) and channel equalization processing (also known as channel matching).

[0003] In wireless systems with code-splitting multiple access (CDMA), different physical channels in the code domain are bundled or processed by multiplexing using separate spreading sequences. In the case of orthogonal spreading of codewords, the original data symbols can then be effectively separated in the receiver by unspreading.

[0004] In a wide-area CDMA (WCDMA) radio system, downward link code bundling or multiplexing is performed using orthogonal codes with a variable spread factor (OVSF). However, the OVSF codewords are only orthogonal to each other if ideal time alignment exists. If signal propagation involves multiple paths, the orthogonality of the code is lost, and the unspreading process is affected by multiple access interference (MAI).

[0005] Mobile CDMA radio receivers conventionally use a rake processor, or computational processor, which relies on correlation properties of the spreading sequences. A rake processor is described, for example, in J.G. Proakis, "Digital Communication," New York, McGraw-Hill, 1995. This type of receiver suffers a performance degradation in the presence of code correlation when the MAI (Maximum Interference Index) between a code-bundled transmission is comparable to other sources of noise and interference. Under these conditions, a performance improvement can be achieved by attempting to restore orthogonality between the codes before unspreading.The suboptimal behavior of conventional 3GPP receivers, which rely on computational processing, results in a significant performance loss, particularly for downlink data rates, which increase from 384 kbps for WCDMA specification 99 to high-speed downlink packet access (HDSPA) rates of several Mbps. When code orthogonality is broken due to multiple paths, an effective approach is to use channel equalization instead of computational processing.

[0006] For example, it has been shown that a linear, chip-level equalizer with minimal mean square error (MMSE) offers a significant performance advantage over conventional rake reception, but at the cost of greater implementation complexity. This advantage is particularly important for high-rate data transmission, such as in 3GPP high-speed downlink packet access (HSDPA). However, channel equalization may not provide better performance under all possible conditions. In particular, the use of a channel equalizer provides no advantage under single-beam propagation conditions, i.e., in the absence of multipath signal propagation.

[0007] The above limitations generally depend on the specific equalization algorithm under consideration. In the case of a linear MMSE equalizer, in the presence of a non-frequency-selective response or a fixed channel response, the equalizer relies on estimating the channel impulse response, where a channel estimation error is proportional to the number of channel impulse response samples. In this situation, the response of a computational receiver not only avoids the power loss caused by MAI but actually minimizes the channel estimation error based on the estimation of a single channel sample.

[0008] Similarly, in the case of a least squares (LS) equalizer, receiver performance may be impaired when using the estimation, the channel statistics, performed with a dimension number higher than is required under the special conditions of a non-dispersive channel, i.e., a channel propagation profile with a single tap.

[0009] International publication WO 2009 / 056500 describes a receiver structure capable of selecting the use of a computational receiver or an equalizer. It explains several fundamentals for this selection, including the detection of the receiver operating mode under conditions with spread channels with lower delay. In particular, receiver matching can be based on estimating the mean square (rms) delay spread of the channel or on a measure of the channel energy outside a predefined time window. The inventors recognized that a significant limitation of these schemes lies in the difficulty of distinguishing between a single-beam channel profile and a channel profile with lower delay spread, although the use of channel equalization can still be advantageous.For example, in the case of an HSDPA receiver, the inventors have shown through internal simulation results and performance tests that the use of an equalizer at the chip level can still provide a significant performance advantage in terms of the rake receiver for channel profiles with very low latency spread, such as the Pedestrian A (PA) channel profile of the 3GPP standard.

[0010] US 6,333,953 B1 describes a system and method for selecting a suitable radio sensing technique, in which the extent of time dispersion of a radio channel is estimated based on the number of channel taps required to model the radio channel. US 2004 / 0161029 A1 describes an estimation of a delay time difference between an earliest and a latest received channel path based on the signal-to-noise ratios of the individual signal paths measured at the fingers of a computational receiver. In "A Novel Finger Assignment Algorithm for RAKE Receivers in CDMA Systems," Mohamed Abou-Khousa et al., 2004 IEEE International Conference on Communications, June 20-24, 2004, pp. 2516-2520, a selection of fingers of a rake receiver is described, taking signal-to-noise ratios into account.

[0011] It is an object of this invention to identify n-beam propagation conditions that are able to solve the above problems.

[0012] According to the invention, a method according to claim 1 and a radio receiver according to claim 7 are provided. The dependent claims define further embodiments.

[0013] According to one aspect of the present invention, a method for processing radio signals in a wireless radio receiver is provided to detect an n-beam channel condition of a propagation channel, the method comprising: Receiving signal samples and estimating multiple channel taps from the samples based on a model of the propagation channel as a delay line with multiple channel taps; Estimating signal power and interference power for each of the estimated channel taps; Filtering the signal power to provide a filtered signal power measurement; Filtering the interference power to provide a filtered interference power measurement; Using the filtered power values ​​to determine n strongest channel taps; Generating a first and a second comparison parameter using the n strongest channel taps and at least one further channel tap, in particular generating the first comparison parameter by determining the ratio of the filtered signal power quantity and the filtered noise power quantity for the n strongest channel taps, and generating the second comparison parameter by determining the ratio of the filtered signal power quantity and the filtered noise power quantity for the at least one further channel tap, wherein the ratios are filtered to generate the corresponding first and second comparison parameters; Providing a comparison result based on the first and second comparison parameters and a threshold; and Determining or identifying an n-ray channel condition of the propagation channel from the comparison result.

[0014] Another aspect of the invention provides a radio receiver for a wireless communication system, which includes: a device for receiving signal samples and estimating multiple channel taps from the samples based on a model of a propagation channel as a delay line with multiple channel taps; a device for estimating signal power and interference power for each of the estimated channel taps; a device for filtering the signal power to provide a filtered signal power measurement; a device for filtering the interference power to provide a filter interference power measure; a device for using the filtered power values ​​to determine n strongest channel taps; A device for generating a first and a second comparison parameter using the n strongest channel taps and at least one other channel tap, in particular generating a first comparison parameter by determining the ratio of the filtered signal power and the filtered noise power for the n strongest channel taps, and generating a second comparison parameter by determining the ratio of the filtered signal power and the filtered noise power for the at least one other channel tap, wherein the ratios are filtered to generate the corresponding first and second comparison parameters; a device for providing a comparison result based on the first and second comparison parameters and a threshold; and a device for detecting an n-ray channel condition of the propagation channel from the comparison result.

[0015] Another aspect of the invention provides a computer program product that sets up the method described above when the program product is executed in a processor.

[0016] Another aspect of the invention provides a radio receiver for a wireless communication system, which includes: a wireless interface for receiving digital sample values; a processor that is trained to execute a program that sets up the steps of the procedure defined above; a memory containing a computational processing code sequence and an equalization processing code sequence; wherein the processor is configured to execute the computational processing sequence or the equalization processing sequence depending on the detected n-beam channel condition.

[0017] The following described embodiments relate to the setup of an algorithm in which n = 1 for the detection of single-beam channel conditions in a wireless receiver based on the estimation of the received signal power and the noise plus interference power. The method and the receiver are particularly suitable when applied to the selection of suitable receiver parameters and signal processing algorithms in a 3GPP receiver with Broadband Code Split Multiple Access (WCDMA) / High-Speed ​​Downlink Packet Access (HSDPA). In this case, the invention is particularly advantageous when applied to the selection between computational receiver processing and equalization processing in a receiver.

[0018] The invention is generally applicable in any communication system where there is a need to detect transmission conditions over a non-frequency-selective channel. The invention is specifically, but not exclusively, concerned with the detection of single-beam propagation conditions in a wireless high-speed receiver.

[0019] It should be noted that the interference power includes noise (such as additive Gaussian white noise) and disturbances or interferences (such as from other signals).

[0020] For a better understanding of the present invention and to show how it can be implemented, reference is now made to the accompanying drawings, which: Fig. 1 is a schematic block view of a wireless communication device; Fig. 2 is a block view that shows a selection between rake receiver processing or compute receiver processing and equalizer processing; and Fig. Figure 3 is a schematic block view of a single beam detection system.

[0021] Fig. Figure 1 is a schematic block view of a device or setup, such as that configured in a mobile device to send and receive signals in a wireless communication system. Such a setup can be configured in numerous different ways, but according to Fig. 1 configured as a series of RF / IF stages 32 to receive and transmit wireless signals (TX, RX) via one or more antennas 20. The embodiments of the present invention, as described herein, are primarily concerned with the reception of wireless signals, so that transmitted signals will not be mentioned further. The received signal at the output of the RF / IF stages is typically converted to baseband, with an ADC converting the analog signal into digital samples. The block 32 of Fig. 1 contains components for processing the received radio signals and for providing digital signal samples r(k). This can be achieved in various ways, which are known in the prior art and are not further explained here.

[0022] The sampled values ​​r(k) are fed to a data transfer device 30, which is connected to a processor 22, an instruction memory 24, and a data memory 26. The processor 22 is responsible for processing the sampled values ​​r(k). The processor 22 can execute a number of different functions, which are contained in an instruction memory 24 in the form of code sequences. This provides a so-called "soft" modem, which has a number of advantages, as explained below.

[0023] Fig. Figure 2 is a schematic block view showing a number of different functions executed by processor 22. A first function, designated by block 10, is called a channel parameter estimator. This function estimates a number of different parameters related to the communication channels over which radio signals are transmitted in the wireless communication system. Function 10 provides the output signals or values ​​γ at time k. n (k), n = 1, ..., N C , where N C denotes the number of estimated channel parameters that represent a group of channel parameters derived from the received signal samples r(k). The estimated channel parameters γ n (k) can be used for a number of different purposes. As in Fig. As shown in Figure 2, the samples are fed to a function 12, "Selection of Rake / Equalizer Receiver," which determines whether the received samples are to be processed using a rake receiver or an equalizer receiver. The rake receiver or equalizer receiver is selected by the processor 22, which executes the appropriate code sequence from the instruction memory 24.

[0024] The use of the estimated channel parameters to control the selection of a rake receiver or equalizer receiver (function 12) will now be explained in more detail. Fig. Figure 2 shows a schematic hardware version of the concept. The digital samples r(k) are fed to a switch 4, which has an input 5 that receives the command signal for selecting rake receiver or equalizer processing from function 12. According to this signal, the switch 4 selects a processing path 6 via a rake receiver 7 or a processing path 8 via an equalizer 9. As is known from the prior art, the rake receiver includes a group of computational fingers 7a, 7b for each channel, which is transmitted on a separate channelization code. Each finger belongs to a single drafting / de-spreading device 9 and is further linked to a weighting function 11, and the group of fingers belonging to each channel is linked to an adder 13, which provides a processed output signal on output path 15. Since the way a rake receiver works, or rather...Since the computer receiver is familiar to the expert, this function will not be described further here.

[0025] The equalizer receiver 19 comprises a chip-level equalizer 16 and several design / spreading devices 21a, 21b, for each channel, which are transmitted on a separate channeling code. The output signals of the design / spreading devices are provided on output path 23. An output switch 25 provides the processed output signals on lines 27 for subsequent decoding functions. Switch 25 (like switch 4) is controlled by a control input 5, which receives the command signal for selecting a computational receiver or an equalizer from function 12.

[0026] Although Fig. 2. While a hardware concept for processing the function selection is represented, the same functions can also be based on the software architecture of the Fig. 1 be set up. In this case, it should be noted that in the embodiment of the invention described in Fig. As shown in Figure 1, it is not possible to detect different physical paths (6, 8, 15, 23). Instead, the selection is made with respect to a compute receiver or an equalizer function by downloading different code sequences depending on whether a compute receiver function or an equalizer receiver function is to be executed by processor 22.

[0027] In such a software implementation of the receiver, where computational processing or equalization processing is performed at a given time, the above approach also yields an overall reduction in computational complexity compared to a conventional receiver where channel equalization is implemented as hardware. In this respect, conventional modems based on a hardware implementation must choose between a design dictated by requirements regarding the maximum data rate and one that implements multiple algorithms as separate areas of the silicon chip. These solutions entail higher implementation costs, larger size, and / or higher power consumption, inevitably impacting performance.On the other hand, the proposed solution makes it possible to reduce complexity, size and cost by reusing a common platform to adaptively select the optimal group of signal processing functions capable of maximizing performance and minimizing power consumption.

[0028] It was now on Fig. 3 referred to describe a technique for capturing a single beam, which is set up in the selection of a computational function or an equalization function 12 and the estimation of the channel parameter function 10.

[0029] Let be a wireless communication channel with a total time-variable impulse response h(t,τ)=g T (τ)Ⓧc(t,τ)Ⓧg R (τ) considered, where 'Ⓧ' is the convolution operator, g T (t) and g R (t) the impulse response of the transmit and receive filter and c(t,τ)=∑l=0L−1cl(t)δ(τ−τl) This represents the impulse response at time t of the propagation channel, which is modeled as a delay line with taps at delay iridex τ. With respect to the various channel propagation paths, c(t,τ) is the sum of the complex gains of all paths with delay τ, measured at the current position of the receiver antenna.

[0030] Designated by ĉ ℓ , ℓ = 0,..., L - 1 are the channel taps that are estimated at the receiver. In the case of a computational receiver, the estimated taps can be ĉ ℓ correspond to the monitored (and possibly combined) computing receiver fingers.

[0031] The single-beam detection technique described herein is based on the use of long-term average values ​​of the signal power and the noise-plus interference power corresponding to each estimated channel tap, performed according to the following procedure: ▪ Estimating the signal power P S (ℓ) and the noise plus interference power P I (ℓ) for each of the channel taps ĉ ℓ , ℓ = 0,...,L - 1. ▪ Estimating the quantities P S (ℓ) and P I (ℓ) sufficiently to average over the rapid channel changes due to the decay behavior, in order to thus the filtered sizes P¯S(l) and P¯I(l) for each of the channel taps ĉ ℓ , ℓ = 0,..., L - 1 to obtain. In Fig. 3. The estimation and filtering steps are performed in block 40. In one embodiment, the filter's memory is on the order of a few hundred radio frames (where a single WCDMA / HSPA radio frame has a duration of 10 ms). ▪ Locating the strongest and second strongest channel taps (fingers) based on the ratio between P¯S(l) and P¯I(l) for all channel taps ĉ ℓ, ℓ = 0,...,L - 1 (or based on the value of P¯S(l) for all channel taps ĉ ℓ , ℓ = 0,...,L - 1). PS and PI' These terms are intended to denote the filtered signal power and the noise plus interference power for the strongest tap, and P¯S", and PI" The filtered signal power and the noise-plus interference power for the second strongest tap are to be specified. This step is shown in Block 42. ▪ In block 44, comparison parameters ξ', ξ'' are determined based on the above performance values. P¯S',P¯I',P¯S", and PI" generated. This can be done in many different ways.

[0032] In a first version: ▪ the sizes will be ξ'=P¯S' / P¯I' and ξ"=P¯S" / P¯I" calculated

[0033] In a second version: ▪ the sizes will be P¯S' / P¯I' and P¯S" / P¯I" The filtered power ratios are denoted by ξ' and ξ''.

[0034] In a third variant: ▪ the sizes will be P¯S'⋅P¯I" and P¯S"⋅P¯I' The filtered quantities are denoted by ξ' and ξ''.

[0035] Finally, in comparison block 46, an object is created that controls the selection between computational receiver and equalizer. If ξ' > θ·ξ'' (or equivalently, if ξ' / ξ'' > θ), then it is determined that the receiver operates under single-ray propagation conditions, i.e., that the channel propagation profile contains a single tap; otherwise, it is determined that the channel profile contains more than a single tap.

[0036] For each of the above-listed implementation forms, the threshold θ is optimized from a simulation of the quantities ξ' and ξ'' for the cases of a single-beam and a non-single-beam propagation profile.

[0037] The optimization of θ could also be based on the performance of the receiver processing algorithms to which single-beam detection is applied. In this case, it is better to speak of effective single-beam conditions, which are defined as the conditions under which the receiver processing algorithm designed for a single-beam channel (for example, computational processing in the case of a WCDMA / HSDPA receiver) outperforms the algorithm designed for a non-single-beam channel (for example, equalizer processing in the case of a WCDMA / HSDPA receiver).

[0038] In an alternative implementation, the signal power and the noise-plus interference power of the second strongest tap can be replaced by the combined power obtained from all remaining measured channel taps. ℓ , which are not the strongest channel tap, is derived.

[0039] The proposed approach can be further generalized to detect other channel conditions. For example, the proposed procedure can be readily modified for the detection of an n-beam channel condition, where the quantities PS and PI' as the filtered combined signal power and noise-plus interference power of the n strongest taps are calculated, where P¯S" and PI" which corresponds to the filtered combined signal power and the sum of the noise plus interference power of the remaining channel taps.

[0040] For example, this strategy can be used as an alternative to measuring energy outside a given time window to estimate the length of the channel impulse response.

[0041] The embodiments described above allow detection of single-beam channel conditions based on a comparison between the estimated signal power and the noise-plus interference power of the strongest channel tap and the second strongest channel tap (or based on the strongest channel tap and all other channel taps).

[0042] The embodiments described above allow for differentiation with regard to channels with very short time spreads by relying on long-term averages - by using filters with sufficiently long memory to average over the channel fluctuations due to a rapid decay behavior.

[0043] The selection of the compute / equalizer is based on long-term channel statistics rather than instantaneous channel taps. In particular, if a channel coefficient is present, even if it is temporarily attenuated, performance can increase rapidly enough to make it difficult for the selection process to react in time to switch to the appropriate receiver algorithm. By basing the selection on long-term statistics, a choice is implicitly made in the presence of multiple paths, ensuring that a receiver algorithm capable of handling a multi-path channel is available. Since instantaneous channel realizations do not always reflect long-term statistics, any decision criterion must be based on an average of the channel realizations over a sufficiently long time window.

[0044] In the above embodiment, the filter memory is referenced to the 10 ms WCDMA radio frame. Another important dependency is the channel coherence time (defined as the time interval after which the channel realizations become uncorrelated): Tc=12fD, where f D = f0ν / c is the maximum Doppler frequency.

[0045] For example, at a carrier frequency f0 of 2 GHz and a mobile device speed ν of 3 km / h, several hundred frames correspond to several ten repetitions of the channel coherence time. In preferred embodiments, the filter memory is on the order of greater than ten times the channel coherence time and / or at least an order of magnitude larger than the frame.

[0046] The remaining method for detecting single-beam propagation conditions allows the selection of suitable receiver signal processing for the specific conditions of a non-frequency-selective channel, according to a propagation profile containing a single tap. In particular, it enables switching between a receiver signal processing implementation designed to counteract the effects of multi-path distortion, based, for example, on a chip-level equalizer for WCDMA / HSDPA, and a receiver signal processing implementation optimized for channel conditions and specified decay characteristics, for example, based on a computational receiver for WCDMA / HSDPA.

[0047] For a WCDMA / HSDPA receiver, the choice of compute-receiver processing in the presence of single-beam propagation has the following advantages: • Provides the possibility of reducing receiver complexity and power consumption, especially in the case of a software receiver implementation. • Provides the possibility of achieving the best receiver performance under both frequency-selective conditions and fixed channel conditions in the presence of a limitation of the equalizer performance due to channel estimation errors.

Claims

[1] A method for processing radio signals in a wireless radio receiver for detecting an n-beam channel condition of a propagation channel, the method comprising: Receiving signal samples and estimating multiple channel taps from the samples based on a model of the propagation channel as a delay line with multiple channel taps; Estimating signal power and interference power for each of the estimated channel taps; Filtering the signal power to provide a filtered signal power measurement; Filtering the interference power to provide a filtered interference power measurement; Using the filtered power values ​​to determine the n strongest channel taps; Erzeugen eines ersten Vergleichsparameters durch Bestimmen des Verhältnisses der gefilterten Signalleistungsgröße und der gefilterten Störleistungsgröße für die n stärksten Kanalabgriffe, und Erzeugen eines zweiten Vergleichsparameters durch Bestimmen des Verhältnisses der gefilterten Signalleistungsgröße und der gebieterten Störleistungsgröße für den mindestens einen anderen Kanalabgriff, wobei die Verhältnisse gefiltert werden, um den entsprechenden ersten und zweiten Vergleichsparameter zu erzeugen; Providing a comparison result based on the first and second comparison parameters and a threshold value, and Identifying an n-ray channel condition of the propagation channel from the comparison result. [2] A method according to claim 1, wherein n = 1 and the step of detecting an n-beam channel condition comprises: detecting a single-beam channel condition. [3] A method according to claim 1 or 2, wherein the step of detecting an n-beam channel condition is used to select a rake processing or an equalization processing. [4] A method according to any of the preceding claims, wherein the at least one other channel tap is the second strongest channel tap. [5] A method according to any one of claims 1 to 4, wherein the at least one other channel tap comprises all channel taps except the n strongest taps. [6] A radio receiver for a wireless communication system, comprising: a device for receiving signal samples and estimating multiple channel taps from the samples based on a model of a propagation channel as a delay line with multiple channel taps; a device for estimating signal power and interference power for each of the estimated channel taps; a device for filtering the signal power to provide a filtered signal power level; a device for filtering the interference power to provide a filter interference power quantity; a device for using the filtered power values ​​to determine n strongest channel taps; a device for generating a first comparison parameter by determining the ratio of the filtered signal power quantity and the filtered interference power quantity for the n strongest channel taps, and generating a second comparison parameter by determining the ratio of the filtered signal power quantity and the filtered interference power quantity for the at least one other channel tap, wherein the ratios are filtered to generate the corresponding first and second comparison parameters; a device for providing a comparison result based on the first and second comparison parameters and a threshold; and a device for detecting an n-ray channel condition of the propagation channel from the comparison result. [7] A receiver according to claim 6, wherein the n-beam channel condition is the single-beam channel condition. [8] A receiver according to claim 6 or 7, comprising a computational processing function and a rectification processing function and additionally comprising a device for selecting the computational processing function or the rectification processing function based on the output from the recognition device. [9] A computer program product comprising a program code device which, when executed, implements a method according to any one of claims 1 to 5. [10] A radio receiver for a wireless communication system comprising: a wireless interface for receiving digital sample values; a processor configured to execute a program implementing the steps of any one of claims 1 to 5; a memory containing a rake processing code sequence and an equalization processing code sequence, wherein the processor is configured to execute the rake processing sequence or the equalization processing sequence depending on the detected n-beam channel condition. [11] A radio receiver according to one of claims 6, 7, 8 or 10, installed in a mobile terminal.

Images