Pulse or digital communications – Receivers – Interference or noise reduction
Reexamination Certificate
1998-05-15
2001-12-11
Ghayour, Mohammad H. (Department: 2734)
Pulse or digital communications
Receivers
Interference or noise reduction
C375S229000, C375S230000, C375S231000, C375S232000
Reexamination Certificate
active
06330294
ABSTRACT:
FIELD OF THE INVENTION
The present invention relates to digital communication and, more specifically, to a method of and apparatus for the reception of digital radio signals.
BACKGROUND OF THE INVENTION
The invention has been developed with particular attention to its possible use within mobile radio communication systems such as the systems known as GSM and IS-95. In any case, the invention can be applied in any context wherein the functional reception diagram applied is similar, directly or substantially, to the diagram shown in FIG.
1
.
To illustrate how the invention can be utilized in an existing system, the diagram of a conventional, single antenna, GSM receiver is presented in FIG.
1
. In this diagram, a line
11
receives a baseband digital signal. In addition to an actual useful signal component, the signal comprises a training signal component such as a preamble or a so-called “midamble”, essentially comprising a string of binary characters or bits x
p
(t) which is assumed to be known. The signal received by the antenna (converted in a form suitable for processing in a module
20
X. There the antenna signal is, for instance, converted into baseband and subjected to the conditioning operations currently used in the art for demodulation.
The signal on line
10
is split by a demultiplexing block (not shown in the drawing) between two branches
11
and
12
, intended to convey the training signal or the information signal, respectively. It should be stated that, as is well known to the person skilled in the art, such splitting does not necessarily correspond to an actual routing over two different physical channels, since it can be effected in a virtual manner by means of different processing operations.
The processing performed by the first branch
11
is aimed at obtaining an estimate of the channel impulse response whereon the transmitted signal has propagated. This estimated response can be obtained by analyzing how the training signal is effected by the channel. The aforesaid estimate is usually performed by the correlation (or the matched filtering) effected in a module indicated by
14
. In block
15
, the convolution is calculated between the windowed estimate of the channel impulse response and a set of possible transmitted baseband signals S′
p
(t) (over one bit period) to obtain the signal estimates x(t). These estimates are fed to a processing module
16
where the signal routed over a branch
12
arrives after a possible filtering effected in filter
17
. This filter has a impulse response equal to the windowed ambiguity function (the ambiguity function being, as known, the auto-correlation of the training sequence), i.e. a response given by [X
p
(t)*Xp(T−t)] h
w
(t), where h
w
(t) is a window function. This method has been described by R. Steel, “Mobile Radio Communications”, New York 1992, Chapter 6. In processing module
16
a measurement of the “distance” (incremental metrics for one bit interval) between the generated sequences and the actual received data is performed. The incremental metrics calculated in block
16
are fed through line
18
to a Viterbi processor (known in the art), included in block
19
, where the new metrics for each state are established, as occurs in GSM transmission systems. The Viterbi processor is followed in cascade arrangement by a differential decoder (equally known in the art) which emits the output data stream. Essentially, the received signal on line
10
is subjected to a processing operation that can ideally be seen as a complimentary and opposite action to the one effected by the transmission channel.
In the processing module
16
this signal is subjected to a processing operation that can ideally be seen as a complementary and opposite action to the one effected by the transmission channel. All with the purpose of generating as an output, on the line indicated as
18
, a signal destined to constitute a replica, as faithful as possible, of the transmitted signal, in view of the subsequent decoding. Such decoding can be performed, for instance, by means of a Viterbi decoder
19
, as occurs in GSM transmission systems.
From the analysis of the reference diagram in
FIG. 1
, it is clear that the more articulated is the set of alternations the transmitted signal—and also the training sequence—may have undergone during transmission over the cannel, the more complex and onerous is the set of processing operations performed in elements
14
,
15
,
16
and
19
.
In particular, in mobile radio systems (at least for base stations, but the use of this technique is being extended also to mobile terminals) the use of diversity reception techniques based on the use of a plurality of N receiving antennas has become widespread. The signal received by an array of antennas of this kind in reality comprises N replicas of the same starting signal, which replicas are received by the aforesaid N antennas in a different way (for example due to a different distribution of the echoes, etc.).
The invention exploits this multiplicity of antennas to develop a more robust receiver structure which enhances the communication link quality. The processing of the individual system at the receiving side entails the analysis of a certain number (for example, M) of symbols received successively. A receiver architecture with N antennas should, for the same propagation environment, consider the analysis of N×M symbols. Recovering the transmitted signal in case of the multi-input system essentially entails inverting a system matrix (N×j, where N is the number of antennas and j the number of time instants (i.e. the number of consecutive snapshots) considered necessary to faithfully reproduce the transmitted signal. Direct inversion of this matrix can, if done without due attention, lead to noise amplification and instability. Moreover, it can be rather onerous in terms of time and hardware required, and it hardly appears practical for real-time processing of the received signals, as is required in case, for example, of voice signals.
OBJECT OF THE INVENTION
The object of the present invention therefore is to provide a solution that, though similar to the general diagram shown in
FIG. 1
, does not give rise to the drawbacks described above, further allowing a greater resolution in the performance of the auto-correlation function of the training signal X
p
(t).
SUMMARY OF THE INVENTION
According to the present invention, this object is attained thanks to a method and a system.
For receiving digital signals comprising a training sequence (x
p
(t)) usable to generate an estimate of the transmission channel, wherein the digital signal is received in diversity as a plurality of signal replicas each comprising a respective replica (x
p
(t)*a(t)) of the training sequence. The method comprises the operations:
generating a plurality of versions of the received digital signal separated by a given delay interval (T), each version comprising a respective set of signal replicas;
subjecting each of the versions of the received digital signal to a respective filtering action (
23
.
1
, . . . ,
23
.j) independently of the other versions; the respective filtering action being performed on each of the versions by applying to the respective set of signal replicas a respective first set of filtering coefficients (w*
11
, . . . , w*
N1
; . . . ; w*
1j
, . . . , w*
Nj
) obtained starting from a respective initial set of filtering coefficients,
obtaining the respective initial set of filtering coefficients by subjecting the respective version of the received digital signal to a respective second filtering action performed independently of the respective second filtering actions effected on the other versions of the received digital signal; each of the respective second filtering actions being performed with a respective second set of filtering coefficients identified as a signal (u(t)) able to generate, by convolution (u(t)*x
p
(t)) with the training sequence, a unitary function on a given time slot.
The respective second set of filtering c
Ansbro Andrew Peter
Drakul Spase
Fanigliulo Antonio
Fontana Gianluca
CSELT- Centro Studi E Laboratori Telecomuniicazioni S.p.A.
Dubno Herbert
Ghayour Mohammad H.
LandOfFree
Method of and apparatus for digital radio signal reception does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method of and apparatus for digital radio signal reception, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method of and apparatus for digital radio signal reception will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2583285