Pulse or digital communications – Synchronizers – Synchronizing the sampling time of digital data
Reexamination Certificate
2000-10-04
2003-03-25
Ghayour, Mohammad H. (Department: 2631)
Pulse or digital communications
Synchronizers
Synchronizing the sampling time of digital data
C375S333000, C375S367000
Reexamination Certificate
active
06539071
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The invention relates to a packet transmission system comprising at least a terminal and a head station, the head station comprising means for receiving packets of data transmitted by said terminal with a frequency error &Dgr;f, and error estimation means for computing, on the basis of the received data, a discrete error function Z(&Dgr;{circumflex over (f)}) and for deriving an estimation &Dgr;{circumflex over (f)} of the frequency error corresponding to a maximum value of the error function Z(&Dgr;{circumflex over (f)}) for a given accuracy (Acc).
The invention also relates to a receiver for a packet transmission system intended to receive packets of data transmitted by a terminal with a frequency error &Dgr;f and comprising error estimation means for computing, on the basis of the received data, a discrete error function Z(&Dgr;{circumflex over (f)}) and for deriving an estimation &Dgr;{circumflex over (f)} of the frequency error corresponding to a maximum value of the error function Z(&Dgr;{circumflex over (f)}) for a given accuracy (Acc).
The invention further relates to a method of frequency correction at the receiver end for a packet transmission system, comprising
a step of receiving packets of data from at least one terminal of said system,
an error estimation step for computing, on the basis of the received data, a discrete error function Z(&Dgr;{circumflex over (f)}) and for deriving a frequency error estimation &Dgr;{circumflex over (f)} relative to the data received and corresponding to a maximum value of the error function Z(&Dgr;{circumflex over (f)}) for a given accuracy (Acc).
The invention finds important applications, notably in the field of cable or satellite transmissions with return paths in which a plurality of terminals can transmit data packets to a head station in accordance with a frequency and time division mechanism. Such transmissions from terminals to the head station are referred to as ascending transmissions.
These terminals are generally intended for the consumer. It is thus important to reduce their cost price. To this end, it is advantageous to use low-cost local oscillators which are, however, relatively inaccurate, for generating the carrier frequencies to be used in the ascending transmissions. Typically, the oscillators used have an accuracy varying between 1 ppm (part per million) and 10 ppm. The error resulting in the generated carrier frequency is proportional to the carrier frequency and is larger as the frequencies used are higher. By way of example, in the satellite transmission systems using the frequency band Ka (20 GHz-30 GHz) for the ascending transmissions, the width or frequency error observed for a local oscillator having an accuracy of 1 ppm may reach ±30 kHz (i.e. for a symbol frequency of 100 kHz, the standard frequency width with respect to the symbol frequency is ±30%).
2. Description of Related Art
The article “Frequency estimator with dichotomous search of periodogram peak” by Y. V. Zakharov and T. C. Tozer, published in the magazine Electronics Letters, vol. 35, no. 19, Sep. 16, 1999, describes a frequency estimator which is based on an algorithm named after its authors Rife and Boorstyn for estimating a frequency error in a frequency range which is standardized with respect to the symbol frequency &Dgr;f/B of ±½ (where &Dgr;f is the frequency error and B is the symbol frequency). In the description hereinafter, the estimation of the standard frequency width &Dgr;f/B or &Dgr;f×T
s
(where T
s
is the duration symbol) which is to be determined will be denoted &Dgr;{circumflex over (f)}. The received data are modulated in accordance with a modulation of the PSK (Phase Shift Keying) type for forming symbols.
The estimators are characterized by different parameters, notably by the accuracy of the obtained estimation which is denoted Acc, the magnitude of the acquisition range of the frequency width denoted ±&Dgr;f
max
, the minimum level with respect to the signal-to-noise ratio of the treated signal, and the complexity of the algorithm. Certain parameters have opposite evolutions. Particularly the accuracy of the estimation is less as the acquisition range is larger; and the estimators are more complex as they are capable of functioning at lower signal-to-noise ratios. Rife and Boorstyn have shown that the frequency error &Dgr;f has a maximum probability of being situated at the location of the maximum amplitude of the following function, denoted Z(&Dgr;{circumflex over (f)}):
Z
(
Δ
f
^
)
=
1
L
∑
k
=
0
L
-
1
z
(
k
)
×
ⅇ
-
j2
nk
Δ
f
^
(
1
)
wherein L is the length of observation, i.e. the number of received symbols used for computing the error estimation, and k represents the position of the symbol in the received frame, with z(k)=x(k)×c
k
* if the received symbols are known, where c
k
* is the conjugated complex of the known predetermined symbol c
k
and where x(k) is the symbol received with a frequency error &Dgr;f and is written as x(k)=c
k
×e
2&pgr;j.k&Dgr;f×T
S
+j&phgr;
0
where &phgr;
0
is the inititial phase shift between the local oscillator used at the receiver end and that used at the transmitter end (this phase shift is different for each packet and corresponds to the phase shift for k=0) and with z(k)=e
jM arg|x(k)|
if the received symbols are not known in advance, where M is the number of PSK modulation phases used but in which case the result obtained must be divided by M so as to obtain the estimation &Dgr;{circumflex over (f)}.
The technique proposed in the cited article recommends an error estimation in two steps, a first step in which a first error estimation is obtained by computing a Fast Fourier Transform (FFT) for a certain number of points in each standard interval ±½, and a second step in which successive iterations of error computations are performed for searching the maximum value of the error function in accordance with a dichotomous method by starting from three adjacent points roughly representing of the maximum location. This technique implies that the points are computed in all intervals of acquisition of the standard frequency width ±½ for obtaining the first estimation. In the case of an estimation only using a preamble of predefined data in each received packet for effecting the estimation, currently referred to as DA (Data Aided), the time of acquisition of the error is very limited. It is also preferable to limit the computing power required at the level of the network head by limiting the number of computations. However, this method does not allow adaptation of the number of computations to the real performance of the local oscillators used in the terminals.
SUMMARY OF THE INVENTION
It is an object of the invention to remedy these drawbacks by proposing a system, a receiver and a frequency correction method with which the reception of data packets having a large frequency shift can be improved by limiting the number of computations to a maximum by taking the real performance of the terminals used in the system into account. To this end, a transmission system and a receiver according to the invention as described in the opening paragraph are characterized in that the error correction means comprise iteration means for effecting a first iteration of the computation of the error function Z(&Dgr;{circumflex over (f)}) for 2N+1 error values comprised in a fixed standard range ±&Dgr;f
max
dependent on said terminal for deriving a first frequency error estimation (&Dgr;{circumflex over (f)}(i)), and successive iterations for values comprised between the current estimation previously calculated, and a neighboring value for which the error function Z(&Dgr;{circumflex over (f)}) has the largest amplitude, until a predetermined number of iterations (it) related to the ac
Brajal Americo
Chouly Antoine
Legrand Delphine
Al-Beshrawi Tony
Ghayour Mohammad H.
Gross Russell
Koninklijke Philips Electronics , N.V.
LandOfFree
Frequency correction at the receiver end in a packet... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Frequency correction at the receiver end in a packet..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Frequency correction at the receiver end in a packet... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3047306