Parallel-pipeline implementation of digital signal processing techniques based on modular codes

Kalmykov, Igor A.; Veligosha, A. V.; Kaplun, Dmitry; Klionskiy, Dmitry M.; Gulvanskiy, V. V.

doi:10.1109/scm.2016.7519731

Cited by 6 publications

(6 citation statements)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The block diagram of the proposed converting device is provided in Figure 2 The developed algorithm provides the residue formation for three modular cycles of the interfacing device. Thus, we can conclude that the developed conversion method provides significantly better performance than the existing conversion methods considered in [10].…”

mentioning

confidence: 78%

See 1 more Smart Citation

Improving Calculation Accuracy of Digital Filters Based on Finite Field Algebra

et al. 2019

View full text Add to dashboard Cite

The applications of digital filters based on finite field algebra codes require their conjugation with positional computing structures. Here arises the task of algorithms and structures developed for converting the positional notation codes to finite field algebra codes. The paper proposes a method for codes conversion that possesses several advantages over existing methods. The possibilities and benefits of optimization of the computational channel structure for digital filter functioning based on the codes of finite field algebra are shown. The modified structure of computational channel is introduced. It differs from the traditional structure by the fact that there is no explicit code converter in it. The main principle is that the “reference” values of input samples, which are free from the error of the analog-digital converter, are used as input samples. The proposed approach allows achieving a higher quality of signal processing in advanced digital filters.

show abstract

mentioning

confidence: 78%

“…Such systems are best suited for parallel computing. One of the most fruitful research areas here is the algebra of finite field, which provides an impressive level of internal parallelism [10] to the DSP systems.…”

Section: Introductionmentioning

confidence: 99%

Improving Calculation Accuracy of Digital Filters Based on Finite Field Algebra

et al. 2019

View full text Add to dashboard Cite

show abstract

“…After determining these parameters from Appendices 1 and 2, the most promising RNS moduli sets are these with the lowest possible β. These results can be used for building effective parallel computational systems [15] based on computers with parallel structure like FPGA and GPU [16,17]. The basic idea of a hardware implementation is that an algorithm (division, sign detection, comparison of numbers, reverse conversion) based on a diagonal function requires division by SQ.…”

Section: Balance Metric For Building Effective Computational Systemsmentioning

confidence: 99%

Classification of Moduli Sets for Residue Number System With Special Diagonal Functions

et al. 2020

View full text Add to dashboard Cite

The paper presents algorithms for the generation of Residue Number System (RNS) triples with = 2 − 1 and quadruples with = 2 for some k. Triples and quadruples allow us to design efficient hardware implementations of non-modular operations in RNS such as division, sign detection, comparison of numbers, reverse conversion with using of a diagonal function from requiring division with the remainder by the diagonal module SQ. Division with a remainder in the general case is the most complex arithmetic operation in computer technology. However, the consideration of special cases can significantly simplify this operation and increase the efficiency of hardware implementation. We show that there are 5573 good RNS triples (2301 even and 2372 odd) with elements less than 10 000, as the values of SQ vary from 2 5 − 1 to 2 27 − 1. In contrast, RNS quadruples with = 2 seem to be quite rare. Restricting our search to sums of the elements in a quadruple less than 4000 we find that exactly 31 such quadruples exist. Their values of SQ vary between 2 20 and 2 30 with always even exponent. We suggest the measure of RNS balance and find perfectly balanced RNS among triples according to this measure. We demonstrate the advantages of more balanced quadruples by means of hardware implementation.

show abstract

“…The comparative evaluation was carried out for the algorithm of the non-recursive filter implemented in the algebra of the finite field under the modular (tabular) execution of its basic operation. The basic operation time was selected as 2 ns (the time of reading data from the fast-acting memory device was 0.1-4 ns [16]), while the output sample was calculated in parallel pipeline mode, i.e., in one clock cycle of the processor.…”

Section: Of 14mentioning

confidence: 99%

Optimization of the FIR Filter Structure in Finite Residue Field Algebra

et al. 2018

View full text Add to dashboard Cite

This paper introduces a method for optimizing non-recursive filtering algorithms. A mathematical model of a non-recursive digital filter is proposed and a performance estimation is given. A method for optimizing the structural implementation of the modular digital filter is described. The essence of the optimization is that by using the property of the residue ring and the properties of the symmetric impulse response of the filter, it is possible to obtain a filter having almost a half the length of the impulse response compared to the traditional modular filter. A difference equation is given by calculating the output sample of modules p1 … pn in the modified modular digital filter. The performance of the modular filters was compared with the performance of positional non-recursive filters implemented on a digital signal processor. An example of the estimation of the hardware costs is shown to be required for implementing a modular digital filter with a modified structure. This paper substantiates the expediency of applying the natural redundancy of finite field algebra codes on the example of the possibility to reduce hardware costs by a factor of two. It is demonstrated that the accuracy of data processing in the modular digital filter is higher than the accuracy achieved with the implementation of filters on digital processors. The accuracy advantage of the proposed approach is shown experimentally by the construction of the frequency response of the non-recursive low-pass filters.

show abstract

Parallel-pipeline implementation of digital signal processing techniques based on modular codes

Cited by 6 publications

References 6 publications

Improving Calculation Accuracy of Digital Filters Based on Finite Field Algebra

Improving Calculation Accuracy of Digital Filters Based on Finite Field Algebra

Classification of Moduli Sets for Residue Number System With Special Diagonal Functions

Optimization of the FIR Filter Structure in Finite Residue Field Algebra

Contact Info

Product

Resources

About