Coded data generation or conversion – Digital code to digital code converters – Data rate conversion
Reexamination Certificate
2001-03-27
2002-10-08
JeanPierre, Peguy (Department: 2819)
Coded data generation or conversion
Digital code to digital code converters
Data rate conversion
C708S313000
Reexamination Certificate
active
06462682
ABSTRACT:
TECHNICAL FIELD
The present invention relates to a sample rate conversion method and apparatus.
BACKGROUND
Sample rate conversion, or resampling, is used when digital systems of different sample rates have to be connected. When the sample rates of the systems to be connected are related in a simple way, for example by integer or fractional factors, exact resampling can be used. In many cases, however, there are no simple relations between the sample rates of the systems. In these cases it is necessary to use non-exact resampling, or arbitrary sample rate conversion.
In certain applications there are several systems with differing sample rates that share common hardware. The common hardware can then often be designed so that its sample rate is related in a simple way to one of the systems, but not all. However, it is also often desirable to keep the sample rate of the common hardware somewhat independent of the other sample rates, in order to retain a degree of freedom in the design and optimization of the common hardware (such as optimizing the sample rate of an analog-to-digital converter). An example is digital IF (Intermediate Frequency) processing for multi-standard radio communication.
The two most commonly used solutions for resampling are polynomial interpolation and polyphase FIR (finite impulse response) filters, see [1]-[3].
Polynomial interpolation, or N-th order interpolation, is a way of computing output samples in-between input samples using polynomials. Lagrange interpolation is a well known polynomial interpolation scheme suitable for resampling. Typically zero-order and first-order Lagrange interpolation is used. A drawback of polynomial interpolation is that it often has poor spectral properties [1]. Another drawback of the polynomial method is that it involves a high computational load.
Polyphase FIR filters techniques (see [3], for example) for resampling applications consist of arranging a highly oversampled FIR filter (many coefficients) into several sets, branches or subfilters, and for each output use a different set of filter coefficients for multiplying the input samples. The set of filter coefficients used for a specific output sample represent a fractional delay filter with a delay equal to the delay between the time of an available input sample and the time of the desired output sample. The delay is kept track of by counters or ‘timing NCOs’ (NCO=numerically controlled oscillator). Usually the delay is represented as an integer part, telling which input sample range to use, and a fractional part, telling which subfilter to use. Ideally, for non-rational resampling ratios, infinitely many branches or subfilters are needed for representing all possible delays, but in practice a limited (but large) number of branches are used, approximating all possible delay values with the closest delay available (sample and hold). The error in doing this is the main design tradeoff in this type of resampling. The polyphase method is characterized by having larger memory requirements but somewhat lower computational load than the polynomial method. However, the computational load is still high, and it would be desirable to reduce this load further.
These two methods can also be combined, greatly reducing the number of coefficients needed at the expense of increasing the computational load. When combining polyphase filtering with first-order (linear) interpolation, two ‘raw’ output samples have to be computed for each linearly interpolated output sample, increasing the computational load to twice that of the polyphase filter method itself plus the multiplications needed for the linear interpolation (one per output sample).
SUMMARY
An object of the present invention is an arbitrary sample rate conversion method and apparatus that are less computationally intense than the prior art.
This object is achieved in accordance with the attached claims.
Briefly, the present invention solves this object by providing a technique that is based on polyphase filters, but provides several consecutive output samples from the same subfilter and input sample block by using block convolution. This reduces the computational load. A reduced computational load is especially attractive in a mobile radio communication environment, where it may be exploited for reducing the power consumption, which in turn results in better battery utilization.
REFERENCES:
patent: 5023825 (1991-06-01), Luthra et al.
patent: 5502747 (1996-03-01), McGrath
patent: 5774598 (1998-06-01), Sunshine et al.
patent: 5814750 (1998-09-01), Wang et al.
patent: 6000834 (1999-12-01), Duan
patent: 6115113 (2000-09-01), Flockencier
T.I. Laakso et al., “Splitting the Unit Delay—Tools for Fractional Delay Filter Design”, IEEE Signal Processing Magazine, Jan. 1996, pp. 30-60.
T.A. Ramstad, “Digital Methods for Conversion Between Arbitrary Sampling Frequencies”, IEEE Trans. Acoustics, Speech and Signal Processing, vol. ASSP-32, No. 3, Jun. 1984, pp. 577-591.
A. Oppenheim and R. Schafer, “Discrete-Time Signal Processing”, Prentice Hall, 1989, pp. 548-560.
R. E. Blahut, “Fast Algorithms for Digital Signal Processing”, Addison-Wesley, Inc., 1984, pp. 65-109, 176-186, 292-295.
R.W. Adams and T. Kwan, “A Monolithic Asynchronous Sample-Rate Converter for Digital Audio”, IEEE International Symposium on Circuits and Systems, Jul. 1993, pp. 1963-1966.
I.-S. Lin and S.K. Mitra, “Overlapped Block Digital Filtering”, IEEE Transactions on Circuits and Systems, vol. 43, No. 8, Aug. 1996, pp. 586-596.
Burns Doane Swecker & Mathis L.L.P.
Jean-Pierre Peguy
Telefonaktiebolaget LM Ericsson (publ)
LandOfFree
Sample rate conversion method and apparatus does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Sample rate conversion method and apparatus, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Sample rate conversion method and apparatus will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2991137