GB2344494A

GB2344494A - Digital communications receiver with selectable filtering regime

Info

Publication number: GB2344494A
Application number: GB9826425A
Authority: GB
Inventors: Patrick Mcandrew; Hamid Reza Karimi; Nicholas William Anderson
Original assignee: Motorola Ltd
Current assignee: Motorola Solutions UK Ltd
Priority date: 1998-12-03
Filing date: 1998-12-03
Publication date: 2000-06-07
Also published as: GB9826425D0

Abstract

A digital communications receiver such as a radio frequency (RF) receiver has an input terminal for receiving a signal which includes spectral interference and desired digital data. The input signal is filtered according to a digital filtering regime before processing to recover the digital data. The digital filtering regime is chosen from a plurality of regimes in dependence upon at least one real-time measurement such that the combined influence of the spectral interference and inter-symbol-interference (ISI) introduced by the filter is reduced. The digital filtering regimes may be implemented as FIR or IIR filters with different order (where increasing order reduces adjacent channel interference but increases ISI). The real-time measurement is one which characterises the interference components present so that the optimum filter coefficients can be selected. The measured interference parameters may be mapped to filter coefficients stored in a look-up table.

Description

DIGITAL COMMUNICATIONS RECEIVER AND METHOD Field of the Invention This invention relates to digital communications receivers, and particularly but not exclusively to digital communications receivers in so-called'multimode'software definable radios.

Background of the Invention Within a Software Definable Radio (SDR), the Radio Frequency (RF) receiver apparatus may need to be able to receive and select RF signals in both wide-band and narrow-band regimes. The receiver apparatus typically operates at a fixed sampling rate, and therefore in order to satisfy the requirements of the wide-and narrow-band regimes either a number of RF receivers must be provided, (which is costly) or a single RF receiver must be adapted to receive signals according to both regimes, by performing sample rate conversion between arbitrary sampling rates whilst also providing means to vary the channel selection bandwidth.

These requirements of the single RF receiver give rise to the need for high order decimation and selection filters when the receiver is operating in a narrow-band mode.

This leads to the problem of high implementation complexity and high MIPS (Millions of Instructions Per Second), thereby draining available processing power from the processor of the receiver or system. This is especially true of mobile telephone systems such as GSM, which specify stringent receiver filter characteristics to facilitate operation in the presence of strong interfering signals.

It is therefore desirable to minimise this complexity, especially in SDR systems where processing power is at a premium and could otherwise be used to perform new functions or to improve the performance of existing ones.

This invention seeks to provide a digital communications receiver and method which mitigates the above mentioned disadvantages.

Summary of the Invention According to a first aspect of the present invention there is provided a method of processing digital data signals received via a communications system, comprising the steps of providing an input signal including spectral interference and desired digital data to be recovered; filtering the input signal according to a digital filtering regime in order to provide a filtered signal; and, processing the filtered signal in order to substantially recover the desired data; wherein the digital filtering regime is chosen from a plurality of regimes, in dependence upon at least one real-time measurement, such that the combined influence of the spectral interference and the interference associated with the digital filter upon the desired data is reduced, giving rise to improved recovery of the desired data.

According to a second aspect of the present invention there is provided a digital communications receiver, comprising: an input terminal arranged to receive a signal including spectral interference and desired digital data to be recovered; filtering means arranged to filter the input signal according to a digital filtering regime in order to provide a filtered signal; processing means arranged to process the filtered signal in order to substantially recover the desired data; and measuring means arranged to provide at least one real-time measurement associated with the receiver; wherein the digital filtering regime is chosen from a plurality of regimes, in dependence upon the at least one realtime measurement, such that the combined influence of the spectral interference and the interference associated with the digital filter upon the desired data is reduced, giving rise to improved recovery of the desired data.

Preferably the real-time measurement includes a parameter characterisation of the input signal. The plurality of regimes are preferably mapped in a look-up table of parameter characterisation value sets, such that the digital filtering regime is chosen in dependence upon the closest matching set of parameter characterisation values.

The real-time measurement preferably includes output error metrics of the system, and may also include a feedback error signal.

In this way a digital communications receiver and method are provided in which the contributions of Inter-Symbol-Interference (ISI) introduced by the channel selection filtering process and the spectral interference from other signals are managed in order to achieve optimal recovery of the data. Furthermore a receiver and method are provided with minimised complexity and reduced processing power requirements, leading to improvedperformance.

Brief Description of the Drawings An exemplary embodiment of the invention will now be described with reference to the drawing in which: FIG. 1 shows a block diagram of a preferred embodiment of a digital communications receiver in accordance with the invention; and, FIG. 2 shows an exemplary embodiment of a selection scheme of the receiver of FIG. 1.

Detailed Description of a Preferred Embodiment Referring to FIG. 1, there is shown a digital communications receiver 10 forming part of a Software Definable Radio (SDR) arranged to receive signals over an air interface. The receiver 10 includes a digital filtering and sample rate conversion block 20 and inner receiver block 30. A down-converted received signal m (k) is input to the digital filter block 20. The output of the digital filter block 20, which may be either over-sampled or sampled at the rate of one sample/symbol, is further processed by typical inner receiver functions, namely equalisation, channel estimation, and re-mapping & burst assembly, as shown by equalisation block 32, channel estimation block 34, and re-mapping & burst assembly block 35. These functions include matched filters or equalisation as required by the air interface.

A parameter characterisation block 40 is also coupled to receive the input signal m (k). A filter coefficient mapping block 50 is coupled to the parameter characterisation block 40 and to the digital filter 20. The functions of the filter coefficient mapping block 50 and the parameter characterisation block 40 are further described below.

The digital filter block 20 performing the sample rate conversion, decimation and channel selection may in general be either of Finite Impulse Response (FIR) or Infinite Impulse Response (IIR) form, and the filtering function may be implemented in one or more filter blocks depending upon the specific system design. In any of these cases, the digital filter block 20 has a varying order, and by increasing the order of the channel selection filter, adjacent channel interference and therefore the noise power entering the inner receiver block 30 is reduced. However, increasing the order of the filter also typically involves long time-domain responses, giving rise to the Inter-Symbol Interference (ISI) mentioned above, caused by energy associated with a particular transmitted symbol becoming spread over other symbol periods.

In general there will be an optimum filter order at which the various forms of interference are managed such as to minimise the probability of false detection by the receiver.

Furthermore, there is an optimum filter configuration and hence an optimum set of filter characteristics (including filter order) for each filtering stage which yields the best Bit Error Rate (BER) performance for the receiver 10 under given interference conditions estimated by the parameter characterisation block 40. The optimum set of filter characteristics will vary depending upon the type, level and number of external interference signals present in the channel and represents the best compromise between two competing requirements, the suppression of adjacent channel interference and the minimisation of ISI due to filter order.

The signal m (k) contains both the desired modulated signal and other interference components. The interference components may include co-channel interference, adjacent channel interference and additive noise. These interference components are characterised in real-time using well known spectral and temporal techniques by the parameter characterisation block 40. An optimum set of filter characteristics are then selected using the filter coefficient mapping block 50, and this set of filter characteristics is used by the digital filter block 20 to filter the signal m (k), thereby substantially achieving the lowest possible bit error rate (BER).

The characterisation of the interference components is relatively straightforward, and a typical method for achieving this involves the use of Short Term Fourier Analysis to provide spectral estimates for non-stationary type signals. This analysis may also yield a limited amount of time location information for individual components, but this may be augmented using additional temporal processing methods. From this, a set of Interference Parameters of size L is selected which best characterise the interference components of the signal m (k).

The task of mapping the interference classification data to the filter parameters to yield the best system performance may be a non-linear one. This may be achieved by several methods as follows.

An initial set of possible filter configurations are developed, the size of which is > N.

Since the filters are of low-pass configuration, this set is comprised of variations of the filtering and decimation stages, each stages filter order, type and corner frequency.

Through simulation, the system BER performance under all combinations of interference signals may be obtained. From these results, a subset of the best filter parameters may be selected. This subset will be of size N. This process may also be used for selection of the Interference Parameter set if required. The relationship between the Interference Parameters and the Filter Parameters may be encoded in the form of a lookup table. All of the above are offline tasks, which would comprise part of the typical design process.

Referring now also to FIG. 2, there is shown a lookup table 100 arranged to map interference parameters I 10 to yield an optimum set of filter parameters 130. A mapping block 120 is used to map each of the interference parameters 110 (P1-P6) to an appropriate filter set (S 1-S5). The processing overhead for this operation is very low and the storage requirement is proportional to the degree of granularity desired for the mapping process.

Furthermore, it is possible to develop a set of heuristic rules that replace or augment the offline filter set selection process. Such heuristic rules may take into account other derived real-time values such as output error metrics of the receiver 10 (not shown) or a feedback error signal. In this way real-time design or modification of the required filter characteristics is possible and hence a much larger set of filter characteristics may be utilised, whilst minimising the corresponding storage. The scalar processor in a SDR would typically execute the filter design task.

In this way a technique is provided which allows real-time adaptation of filtering characteristics, and hence implementation complexity, in conjunction with the measured profile of extemal interference signals.

In addition, the process of channel equalisation performed by block 32 of the inner receiver block 30 is typically highly computationally demanding, and often increases exponentially with the time-domain response duration of the channel. For the case in which there is little adjacent channel interference contained in the input signal m (k), the filter orders can be chosen such that the ISI contained within the filter output signal is less than would be the case for a receiver that did not make use of such an adaptive channel selection filtering scheme. Favorably, in this scenario, the complexity of the equalisation block 32 and the MIPS processing demand thereof may be substantially reduced.

It will be appreciated that alternative embodiments to the one described above are possible. For example, mapping techniques other than the look-up table described above may be used.

Furthermore the mapping arrangement as shown in FIG. 2 may include more or less sets of interference parameters 110 and filter parameters 130 than those shown and described.

Also two or more sets of interference parameters 110 may be mapped to the same set of filter parameters 130, or vice-versa.

Claims

Claims 1. A method of processing digital data signals received via a communications system, comprising the steps of : providing an input signal including spectral interference and desired digital data to be recovered; filtering the input signal according to a digital filtering regime in order to provide a filtered signal; and, processing the filtered signal in order to substantially recover the desired data; wherein the digital filtering regime is chosen from a plurality of regimes, in dependence upon at least one real-time measurement, such that the combined influence of the spectral interference and the interference associated with the digital filter upon the desired data is reduced, giving rise to improved recovery of the desired data.
2. A digital communications receiver, comprising: an input terminal arranged to receive a signal including spectral interference and desired digital data to be recovered; filtering means arranged to filter the input signal according to a digital filtering regime in order to provide a filtered signal; processing means arranged to process the filtered signal in order to substantially recover the desired data; and measuring means arranged to provide at least one real-time measurement associated with the receiver; wherein the digital filtering regime is chosen from a plurality of regimes, in dependence upon the at least one real-time measurement, such that the combined influence of the spectral interference and the interference associated with the digital filter upon the desired data is reduced, giving rise to improved recovery of the desired data.
3. The method of claim 1 or receiver of claim 2 wherein the real-time measurement includes a parameter characterisation of the input signal.
4. The method or receiver of any preceding claim wherein the plurality of regimes are mapped in a look-up table of parameter characterisation value sets, such that the digital filtering regime is chosen in dependence upon the closest matching set of parameter characterisation values.
5. The method or receiver of any preceding claim wherein the real-time measurement includes output error metrics of the system.
6. The method or receiver of any preceding claim wherein the real-time measurement includes a feedback error signal.
7. A method substantially as hereinbefore described and with reference to the drawings.
8. A digital communications receiver substantially as hereinbefore described and with reference to the drawings.