Pulse or digital communications – Synchronizers – Synchronizing the sampling time of digital data
Reexamination Certificate
2002-01-29
2004-07-27
Chin, Stephen (Department: 2634)
Pulse or digital communications
Synchronizers
Synchronizing the sampling time of digital data
C375S365000
Reexamination Certificate
active
06768780
ABSTRACT:
BACKGROUND OF THE INVENTION
1. Field of the Invention
The invention relates to communication systems, and more particularly to an apparatus for achieving synchronization in digital receivers used in communication systems.
2. Description of the Related Art
In synchronous digital transmission, information is conveyed by uniformly spaced pulses and the function of any receiver is to isolate these pulses as accurately as possible. However, due to the noisy nature of the transmission channel, the received signal has undergone changes during transmission and a complete estimation of certain reference parameters is necessary prior to data detection. Estimation theory proposes various techniques for estimating these parameters depending on what is known of their characteristics. One such technique is called maximum likelihood (ML). Maximum likelihood estimation assumes the parameters are deterministic or at most slowly varying over the time interval of interest. The term deterministic implies the parameters are unknown but of a constant value and are, therefore not changing over the observation interval. These unknown parameters can cover such factors as the optimum sampling or the phase offset introduced in the channel or induced by the instabilities of the transmitter and receiver oscillators. It is widely recognized that maximum likelihood estimation techniques offer a systematic and conceptually simple guide to the solution of synchronization problems. Maximum likelihood offers two significant advantages: it leads to appropriate circuit configurations and provides near optimum or optimum performance depending on the known channel conditions. See, e.g., J. G. Proakis, “Digital Communications”, Third Edition, McGraw-Hill Publishers, 1995, pp. 333-336.
Maximum likelihood parameter estimation for communication systems facilitates two forms of processing depending on how the data present on the received signal is exploited to assist in parameter estimation. The first is data-aided (DA) estimation wherein known data within the received data stream is exploited to improve the estimation performance. Alternatively, non-data-aided (NDA) estimation is possible wherein the random data is considered a nuisance parameter, which is removed by averaging the received signal over the statistics of the random data. Applying ML as the criterion to derive the NDA ML timing offset estimator, results in a mathematical expression, which is highly non-linear and totally impractical to implement (i.e., the solution requires calculating the natural logarithm of a hyperbolic cosine function of the received signal samples). However, the ML based timing estimators available in the literature are derived by making suitable approximations for the natural logarithm of the hyperbolic cosine non-linearity for extreme values of signal to noise conditions, which generally yield acceptable performance over a wide range of signal to noise conditions. The variance of the timing estimate depends on the applied approximation. In the present invention, an approximation is applied, which offers an excellent compromise between variance performed and implementation cost for the NDA estimation of a timing offset where a phase offset is present on the received signal.
Generally, if the transmitter does not generate a pilot synchronization signal, the receiver must derive symbol timing from the received signal. The term symbol is used in this context to refer to transmitted signals that are phase and or amplitude modulated with discrete phase and or amplitude relationships; each assigned relationship is a symbol that is subject to detection at the receiver. Both the transmitter and receiver employ separate clocks which drift relative to each other, and any symbol synchronization technique must be able to track such drift. Therefore, choosing the proper sampling instants for reliable data detection is critical, and failure to sample at the correct instants leads to inter-symbol interference (ISI), which can be especially severe in sharply bandlimited signals. The term ISI refers to two or more symbols that are superimposed upon each other. Under these circumstances, phase detection of each symbol becomes extremely difficult. Incorrect sampling implies the receiver is inadvertently sampling where the influence of the previous data symbol is still present. See, e.g., J. G. Proakis, “Digital Communications”, Third Edition, McGraw-Hill Publishers, 1995, pp. 536-537.
In a digital receiver, the signal following demodulation is first passed through an anti-aliasing filter which is used to limit the bandwidth of the received signal, and is subsequently sampled asynchronously. Asynchronous sampling implies there is no control over the instant at which the sampling of the continuous time signal occurs.
FIG. 2
illustrates the concept of oversampling a continuous signal at four samples per symbol. The optimum sampling instants correspond to the maximum eye opening and are located approximately at the peaks of the signal pulses. The term “eye opening” refers to the amplitude variations of the signal at the output of the pulse-shaping filter. An “eye” is formed by superimposing the output of the pulse shaping filter for each symbol upon the other until the central portion takes on the shape of an “eye” as illustrated in
FIGS. 1
a
and
1
b
for the case of a BPSK (Binary Phase Shift Keying) modulation scheme. Note that under high signal-to-noise conditions, the “eye” is open whereas at low signal to noise conditions the “eye” is closed.
Among synchronization techniques, a distinction is made between feedforward and feedback systems. A feedback system uses the signal available at the system output to update future parameter estimates. Feedforward systems process the received signal to generate the desired estimate without explicit use of the system output. Whether the design approach is feedforward or feedback, both techniques are related to the maximum likelihood parameter estimation. In an error tracking feedback loop, the timing estimator constantly adjusts the phase of a local clock oscillator to minimize the phase error between the estimated and the optimum sampling instant as illustrated in FIG.
3
. The principle is the same for continuous time or sampled input signals. Feedforward designs however are applied to sampled input signals. A feedforward timing loop as illustrated in
FIG. 4
, initially samples the incoming signal, and then using techniques such as interpolation/decimation estimates the ideal sample and removes the redundant samples. Both feedforward and feedback techniques are used in the current technology. However, it should be noted that there are advantages and disadvantages associated with both approaches, which should be understood prior to deciding the appropriate estimator configuration for a particular design.
Problems with feedback techniques include the acquisition time, the high probability of hangup and cycle slips associated with their phase locked loop (PLL) based structures, especially in the presence of fading. Fading occurs when signal components arriving via different propagation paths add destructively. Hangup occurs when the initial phase error of the estimator is close to an unstable equilibrium point, which can result in an extremely long acquisition time (i.e., a long time for the loop to adjust to the correct phase), in fact, the loop may never recover. Hangup is very serious as it can even occur in perfect channel conditions. See, e.g., H. Meyr, M. Moeneclaey and S. A. Fechtel, “Digital Communication Receivers: Synchronization, Channel Estimation and Signal Processing”, John Wiley Publishers, 1998, pp.94-97. Cycle slips are very harmful to the reliability of the receiver's decisions, because a cycle slip corresponds to the repetition or omission of a channel symbol. H. Meyr, M. Moeneclaey and S. A. Fechtel, “Digital Communication Receivers: Synchronization, Channel Estimation and Signal Processing”, John Wiley Publishers, 1998, pp.385-399. These issues are solely due to the feedback nature
Hatim Baya
Lakkis Ismail
O'Shea Deirdre
Tayebi Masood K.
Chin Stephen
Knobbe Martens Olson & Bear LLP
Ware Cicely
Wireless Facilities, Inc.
LandOfFree
Non-data aided maximum likelihood based feedforward timing... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Non-data aided maximum likelihood based feedforward timing..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Non-data aided maximum likelihood based feedforward timing... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-3190780