 Research Article
 Open Access
SQNR Estimation of FixedPoint DSP Algorithms
 Gabriel Caffarena^{1}Email author,
 Carlos Carreras^{2},
 Juan A. López^{2} and
 Ángel Fernández^{2}
https://doi.org/10.1155/2010/171027
© Gabriel Caffarena et al. 2010
 Received: 1 August 2009
 Accepted: 13 April 2010
 Published: 20 May 2010
Abstract
A fast and accurate quantization noise estimator aiming at fixedpoint implementations of Digital Signal Processing (DSP) algorithms is presented. The estimator enables significant reduction in the computation time required to perform complex wordlength optimizations. The proposed estimator is based on the use of Affine Arithmetic (AA) and it is presented in two versions: (i) a general version suitable for differentiable nonlinear algorithms, and Linear TimeInvariant (LTI) algorithms with and without feedbacks; and (ii) an LTI optimized version. The process relies on the parameterization of the statistical properties of the noise at the output of fixedpoint algorithms. Once the output noise is parameterized (i.e., related to the fixedpoint formats of the algorithm signals), a fast estimation can be applied throughout the wordlength optimization process using as a precision metric the SignaltoQuantization Noise Ratio (SQNR). The estimator is tested using different LTI filters and transforms, as well as a subset of nonlinear operations, such as vector operations, adaptive filters, and a channel equalizer. Fixedpoint optimization times are boosted by three orders of magnitude while keeping the average estimation error down to 4%.
Keywords
 Quantization Noise
 Interval Arithmetic
 Noise Estimation
 Fast Estimation
 Average Estimation Error
1. Introduction
The original infinite precision of an algorithm based on the use of real arithmetic must be reduced to the practical precision bounds imposed by digital computing systems. Wordlength optimization (WLO) aims at the selection of the variables' wordlengths of an algorithm to comply with a certain output noise constraint while optimizing the characteristics of the implementation (e.g., area, speed or power consumption). Normally, the precision loss committed is computed by using a double precision floatingpoint arithmetic description of the algorithm as a reference and, although there are some works on quantization for custom floatingpoint arithmetic [1–3], the common approach is to implement the system using fixedpoint (FxP) arithmetic, since this leads to lower cost implementations in terms of area, speed, and power consumption [4–7].
WLO is a slow process due to the fact that the optimization is very complex (NPhard [8]) and also because of the necessity of a continuous assessment of the algorithm accuracy which may involve a high computational load. This estimation is normally performed adopting a simulationbased approach [7, 9, 10] which leads to exceedingly long design times. However, in the last few years, there have been attempts to provide fast estimation methods based on analytical techniques. These approaches can be applied to Linear TimeInvariant (LTI) systems [6, 11] and to differentiable nonlinear systems [12–15]. As for the noise metric used, they are based on the peak value [15] and on the computation of SQNR [6, 11–14]. Since SQNR is a very popular error metric within DSP systems, our work aims at fast SQNR estimation techniques for LTI and differentiable nonlinear systems.
 (i)
a novel AffineArithmetic (AA) SQNR estimator optimized for LTI algorithms,
 (ii)
a novel AAbased SQNR estimator for LTI and differentiable algorithms. Previous approaches were not able to deal with feedback systems, or produced overestimations.
Our approach enables addressing complex WLO techniques, since the computation times are drastically reduced while providing high levels of accuracy.
The paper is structured as follows. In Section 2, related work is discussed. Section 3 deals with fixedpoint optimization. Section 4 presents the grounds of the novel SQNR estimation proposal. In Section 5, the benchmarks used for validation are described. Performance results are collected in Section 6. And finally, Section 7 draws the conclusions.
2. Related Work
In this section, we focus on those approaches aiming at estimating the quantization noise to avoid the execution of timeconsuming simulations [7, 9, 18] and, therefore, that support fast WLO. We disregard those that are not fully automated [19–22], but consider those that, even though are not implemented within an automatic WLO engine, could be easily integrated within one. Also, we do not consider in this analysis approaches that focus on errorfree implementations [23–25].
The SignaltoQuantization Noise Ratio (SQNR) is a popular quality metric in DSP systems. However, only recently it has been considered in the development of fast quantization noise estimators. Approaches such as [19–25] and also the fully automated [15, 26–28] aim basically at peakvalue estimates. Most of these works are based on the use of (i) interval arithmetic (IA) [29], which produces significant overestimations in general, and intolerable overestimations in the presence of loops; (ii) multiinterval arithmetic (MIA) [30], which improves the results of IA but it still performs poorly in the presence of loops; (iii) affine arithmetic [31], which solves the cancellation problem of IA, and can alleviate overestimation by applying confidence intervals; and (iv) the computation of firstorder derivatives [15, 28], mostly combined with a worstcase analysis, that leads again to overestimation. Due to its interest for DSP applications, only approaches that consider SQNR as a quality metric are fully analyzed in this section.
Fast quantization noise estimation approaches.
Approach  Type  Cyclic complexity  Parameterization complexity  Estimation complexity  Accuracy  Comments 

Constantinides et al. [6]  LTI  YES 
 dot product  High  Steady state 
López et al. [11]  LTI  YES  + dot product  High  Affine arithmetic steady state  
Menard [16]  LTI  YES 
 dot product  High  Graph analysis steady state 
Constantinides [12]  NL  YES 
 dot product  Variance overestimated  Differentiable operations 1st order approx. 
NL  NO 
 dot product + matrixvector mult.  High  Differentiable operations 1st order approx.  
Shi and Brodersen [14]  NL  YES  coeff. curvefitting  dot product + matrixvector mult.  High  Differentiable operations 1st order approx. 
This work (Section 4.3)  LTI  YES 
 dot product  High  Affine arithmetic Steady state 
This work (Section 4.2)  NL  YES  Acycilic:  dot product + matrixvector mult.  High  Affine arithmetic Differentiable op. 1 order approx. 
cycilic: It depends on of loops and stimuli size 
The approaches in the table have been grouped according to the type of algorithm being addressed. The first three rows correspond to approaches aimed at LTI algorithms, the next three rows to those addressing nonlinear algorithms (also valid for LTI systems), and the last two rows describe the features of the two approaches proposed in this paper.
2.1. Linear TimeInvariant Algorithms
Note that and can be computed by means of a graph analysis, and once they are determined, the output noise power can be estimated from and .
In [6] a twostep method is applied where, first, vectors and are computed, and then, expression (2) is used to estimate the output noise variance during WLO. The Parseval Theorem [32] is applied in order to compute expression (5), since it is possible to obtain an equivalent expression that makes use of the impulse response from signal to the output of the systems ( ), instead of using . This highly simplifies the computational cost. If the length of the input vectors is long enough, expression (1) can be estimated with high precision leading to highly accurate quantization noise estimations.
An AAbased approach is presented in [11]. The approach is based again on the computation of for each signal. Due to the characteristics of AA, it is possible to compute all simultaneously. The process has not been divided into parameterization (extraction of vectors and ) and noise estimation. Instead, everything is computed at once. It can be seen in Table 1 that the computational cost is similar to the total cost of [6] (e.g., parameterization plus estimation times). Also, the quality of the estimates is high, since they are based on (1). This approach is further developed in Section 4.3 in order convert it into a twostep method, thus, allowing faster noise estimation (see Table 1, this work—LTI).
The approach in [16] also relies on (1) to present a twostep estimation method. The parameterization is based on the application of graph transforms that allow to obtain the vectors and (5) and (6). As it can be seen in Table 1, the performance in terms of computation time and accuracy is equivalent to the other two approaches.
2.2. NonLinear Systems
The approaches aimed at nonlinear systems are mainly based on perturbation theory, where the effect of the quantization of each algorithm's signal on the quality of the output signal is supposed to be small. This allows to apply firstorder Taylor expansion to each nonlinear operation in order to characterize the effect of the quantization of the inputs of the operations. This constrains the application to algorithms composed of differentiable operations. The existent methods enable us to obtain an expression similar to (2) that relates the wordlengths of signals to the power—also mean and variance—of the quantization noise at the output. This will be further explained in Section 4.2 (19).
In [12] a hybrid method which combines simulations and analytical techniques to estimate the variance of the noise is proposed. The estimator is suitable for nonrecursive and recursive algorithms. The parameterization phase is relatively fast, since it requires simulations for an algorithm with variables. The noise model is based on [33] and second order effects are neglected by applying first order Taylor expansions. However, the paper seems to suggest that the contributions of the signal quantization noises at the output can be added, assuming that the noises are independent. In nonlinear systems, this is a strong assumption that leads to variance underestimation. The accuracy of the method is not supported with any empirical data.
In [14] another method suitable for nonrecursive and recursive algorithms is presented. Here, simulations as well as a curve fitting technique (with variables) are required to parameterize quantization noise. On the one hand, the noise produced by each signal is modeled following the traditional quantization noise model from [34, 35], which is less accurate than [33], and, again, second order statistics are neglected. On the other hand, the expression of the estimated noise power accounts for noise interdependencies, which is a better approach than [12]. The method is tested with an LMS adaptive filter and the accuracy is evaluated graphically. There is no information about computation times.
Finally, in [13] the parameterization is performed by means of simulations and the estimator is suitable only for nonrecursive systems. The accuracy of this approach seems to be the highest since it uses the model from [33] and it accounts for noise interdependencies. Although, the information provided about accuracy is more complete, it is still not sufficient, since the estimator is tested in only a few SQNR scenarios.
2.3. This Work
As aforementioned, we present two approaches: one exclusive for LTI algorithms in steady state, and the other for differentiable algorithms which are a subset of nonlinear algorithms. The LTIoriented approach is based on [11] and it basically enables the division of the estimation process into two steps. One step is devoted to parameterization, while the other is dedicated to perform fast estimations. This method is equivalent to the other methods present in the literature. The advantage that it offers is that now it is possible to analyze the most important finite wordlength effects (SQNR analysis, peak value analysis, dynamic range, limit cycles) using the very same AA simulation engine.
Regarding nonlinear systems, our approach tries to overcome most of the drawbacks of the works presented above. It deals with nonrecursive and recursive systems, using the accurate noise model from [33] and also accounting for noise interdependencies. The parameterization time can be quite long for algorithms that contain loops. However, as we will see in Section 6, the computation times are within standard times, and the benefits of fast estimations make up for the sometimes slow parameterization process.
3. WordLength Optimization
A wraparound scaling strategy is adopted since it requires less hardware than other approaches (i.e., saturation techniques). After scaling, the values of are the minimum possible values that avoid the overflow of signals or, at least, those that reduce the likelihood of overflow to a negligible value. A simulationbased approach is used to carry out scaling [7].
Once scaling is performed, the values of can be fixed during wordlength selection. The right side of Figure 2 shows basic blocks for wordlength selection. The main idea is to iterate trying different wordlength (i.e., ) combinations until the cost is minimized. Each time the wordlength of a signal or a group of signals is changed, the wordlengths must be propagated throughout the graph, task referred to as graph conditioning [6], in order to update the rest of wordlengths. The optimizer control block selects the size of the wordlengths using the values of the previous error and cost estimations and decides when the optimization procedure has finished. The first block in the diagram is the extraction of the quantization noise model (parameterization). The role of this block is to generate a model of the quantization noise at the output due to the FxP format of each signal. This enables to perform a quick error estimation within the optimization loop. The implications of using a fast error estimator are twofold. On the one hand, it is possible to reduce WLO time. On the other hand, more complex optimization techniques can be applied in standard computation times.
4. Quantization Noise Estimation
4.1. Affine Arithmetic
Affine Arithmetic (AA) [31] is an extension of Interval Arithmetic (IA) [29] aimed at the fast and accurate computation of the ranges of signals in a particular mathematical description of an algorithm. Its main feature is that it automatically cancels the linear dependencies of the included uncertainties along the computation path, thus, avoiding the oversizing produced by IA approaches [36]. It has been applied to both, scaling computation [15, 36, 37], and wordlength allocation [1, 15, 36]. Also, a modification called Quantized Affine Arithmetic (QAA) has been applied to the computation of limit cycles [38] and dynamic range analysis of quantized LTI algorithms [37].
4.2. Proposed Estimator: General Expression
Here, we present a method able to estimate the quantization noise power from a single AA simulation. The noise estimation is not based on (1), since this equation only applies to LTI algorithms in steady state and our proposal is more general, since it covers both LTI algorithms and nonlinear algorithms. Also, the parameterization method does not lead to (2)–(6), since these are aimed at LTI algorithms in steady state.
This noise model, which is referred to as the discrete noise model, is an extension of the traditional modeling of quantization error as an additive white noise [34, 35] (continuous noise model). In [33], it is shown that the continuous model can produce an error of up to 200% in comparison to the discrete model.
Thus, it is possible to know at each moment the origin of a particular error term ( ) and the moment when it was generated ( ). The AAbased simulation can be made independent on the particular statistical parameters of each quantization thanks to error term . This is desirable in order to obtain a parameterizable noise model. This error term encapsulates the mean value and the variance of the error term , and now it can be seen as a random variable with variance and mean . This is a reinterpretation of AA, since the error terms are not only intervals, but they also have a probability distribution associated. Once the simulation is finished, it is possible to compute the impact of the quantization noise produced by signal on the output of the algorithms by checking the values of (see (7)). This enables the parameterization of the noise. Once the parameterization is performed, the estimation error produced by any combination of can be easily assessed replacing all by the original expression that accounts for the mean and variance ( ), thus enabling a fast estimation of the quantization error. We will see all the process in the next paragraphs.
where is the value of the output of the algorithm using floatingpoint arithmetic and the summation is the contribution of the quantization noise sources. Note that is a function that depends on the inputs of the algorithm.
 (1)
perform a step AA simulation adding an affine form to each signal ,
 (2)
compute (22)–(24) using previously collected .
The error estimation phase can now be executed very quickly by applying (19)–(21).
 (i)
expressions (17)–(22) can be applied to DSP algorithms including differentiable operations (e.g. multiplications, divisions, etc.) by mean of (9) due to the 1st order approximation,
 (ii)
they are exact for LTI systems in steady state (see the appendix).
4.3. Particularization for LTI Systems
The expressions and the algorithms from the previous subsection can be applied to LTI algorithms, but with a high computational load. In this subsection, we present new expressions to compute the power, mean and variance of the output error for LTI systems in steady state that enable fast estimations.
5. Benchmarks
 (i)
RGB to YCrCb converter (RGB) [6],
 (ii)
8point IDCT ( ) [26],
 (iii)
2ndorder IIR filter ( ) [26],
 (iv)
3rdorder Lattice filter ( ) [39],
 (v)
6thorder transposed direct form II deltaoperator filter ( ) [40],
 (vi)
3 × 3 vector scalar multiplication ( ),
 (vii)
8 × 8 vector scalar multiplication ( ),
 (viii)
MIMO channel equalizer (EQ) [41],
 (ix)
a mean power estimator based on a 1st IIR filter (POW),
 (x)
1storder LMS filter ( ) [12],
 (xi)
2ndorder LMS filter ( ) [12],
 (xii)
5thorder LMS filter ( ) [12],
 (xiii)
3rdorder Volterra adaptive filter ( )[42].
Properties of benchmarks.
Benchmark  LTI  Cyclic  Inputs  Outputs 
 +/−  *  * 

 Input signals 

 YES  NO  3  3  0  4  0  6  0  16  Uniform noise 
 YES  NO  8  8  0  37  0  11  0  48  Uniform noise 
 YES  YES  1  1  2  2  0  2  0  8  Uniform noise 
 YES  YES  1  1  3  9  0  10  0  24  Uniform noise 
 YES  YES  1  1  6  18  0  29  0  62  Uniform noise 
 NO  NO  3  3  0  3  3  0  0  12  Uniform noise 
 NO  NO  8  8  0  8  8  0  0  32  Uniform noise 
 NO  YES*  2  2  64  2  3  2  4  81  MIMO channel Tx [41] 
 NO  YES  1  1  1  1  1  2  0  7  Synthetic tone 
 NO  YES  2  1  3  5  6  3  0  23  Synthetic tone 
 NO  YES  2  1  5  7  8  4  0  30  Synthetic tone 
 NO  YES  2  1  11  13  14  7  0  51  Synthetic tone 
 NO  YES  2  1  2  4  6  4  0  19  Gaussian noise 
All benchmarks are fed with 16bit inputs and 12bit constants and the noise constraint is an SQNR ranging from 40 to 120 dB. The inputs used to perform the noise parameterization as well as the fixedpoint simulation are summarized in the last column of the table.
6. Results
 (1)
compute scaling by means of a floating point simulation,
 (2)
extract noise parameters (22)–(24) performing an AAbased simulation,
 (3)
perform a WPO as in Figure 2 using a gradientdescent approach,
 (4)
perform a single FxP bittrue simulation and use it as reference to compute the performance and accuracy of the estimator.
Performance of the estimation method: Precision.
Estimation error  

Benchmark  [120,100)^{1} dB  [100,80) dB  [80,60) dB  dB  
(dB)^{2} (%)^{3}  (dB) (%)  (dB) (%)  (dB) (%)  
 0.11  0.24  0.09  0.09  0.07  0.17  0.07  0.44 
 0.11  0.11  0.08  0.42  0.21  0.88  0.27  0.68 
 0.04  0.03  0.04  0.04  0.06  0.74  0.04  0.09 
 0.24  0.69  0.18  0.33  0.20  0.15  0.19  0.46 
 0.03  0.01  0.02  0.03  0.03  0.16  0.16  1.16 
 0.07  0.54  0.07  0.11  0.06  0.50  0.09  0.72 
 0.05  0.57  0.04  0.40  0.04  0.57  0.13  1.19 
 0.27  0.98  0.24  0.71  0.29  0.17  0.18  1.52 
 0.39  5.00  0.17  1.55  0.76  5.96  1.12  12.12 
 0.09  0.41  0.14  0.90  0.16  1.74  0.82  6.96 
 0.09  0.46  0.08  0.24  0.15  0.78  0.92  3.73 
 0.09  0.46  0.08  0.07  0.13  1.08  1.09  5.51 
 1.14  3.33  0.49  1.84  0.81  6.70  1.43  16.67 
All  1.14  0.20  0.49  0.09  0.81  1.26  1.43  3.52 
The results yield that the estimator is extremely accurate for LTI algorithms. The mean percentage error is smaller than 1.16%, and the maximum relative error is smaller than 0.24 dB. The quality of the estimates is homogenous within the range dB.
The accuracy for nonlinear algorithms shows some degradation. This is expected, since a 1storder Taylor approximation has been applied (9) in the computation of the quantization noise. Moreover, the presence of loops increases the error in the estimation, since the error due to neglecting Taylor series terms is amplified through the feedback loops. The nonlinear algorithms without loops perform significantly well. The mean percentage error is smaller than 1.52%, and the maximum relative error is smaller than 0.3 dB. This performance is similar to that of LTI algorithms.
The nonlinear algorithms that contain loops have a clearly different behaviour. The mean percentage error is smaller than 16.7%, and the maximum relative error is smaller than 1.43 dB. Now, the accuracy decreases as long as the error constraints get looser. This is due to the aforementioned amplification of the Taylor error terms and also to the fact that the uniformly distributed model for the quantization noise does not remain valid for small SQNRs. The errors due to the quantization noise model introduced by the SQNR ranges used for these experiments are minimum, but, after being propagated through the feedback loops and amplified due to nonlinearities, they become much more noticeable. Anyway, the quality of the estimates is still very high.
The average percentage error is 3.52% which confirms the excellent accuracy obtained by our estimator.
Performance of the estimation method: Computation time.
Bench.  FxP  Param.  Param.  No. of estimates  Estimationbased optim.  Simulationbased optim.  Speedup 

Samples  Samples  time (secs)^{+}  (mean)  (secs)^{+}  (secs)^{+}  
 20000  1  0.00016  141  0.03  76  ×3205 
 20000  1  0.00031  4575  5.77  13774.81  ×2468 
 20000  5000  0.88  19  0.02  4.41  ×270 
 20000  20000  10.80  2276  0.74  2381.51  ×3222 
 20000  5000  6.31  3930  3.47  11206.08  ×3235 
 20000  20000  59  150  0.03  66.86  ×2122 
 20000  20000  330  1739  1.72  2331.79  ×1377 
 16000  16000  61.64  231  0.12  105.78  ×904 
 20000  20000  546.14  97  0.02  21.93  ×1048 
 5000  5000  908.02  712  0.42  163.73  ×394 
 5000  5000  592.11  1032  0.94  310.93  ×331 
 5000  5000  1646.38  2547  7.26  1611.46  ×221 
 5000  5000  212.72  673  0.29  151.13  ×526 
All  —  —  —  —  —  —  ×1486 
The parameterization time goes from 160 secs. to 28 mins. (1646 secs.), and it depends on the size of the input data, the complexity of the algorithm (i.e., number and types of operations), and the presence of loops. The benchmarks clearly show how the parameterization time is increased as long as the number of delays, and therefore loops, increases. These times might seem quite long, but it must be born in mind that the parameterization process is performed only once, and after that the algorithm can be assigned a fixedpoint format as many times as desired using the fast estimator.
The mean number of estimates in the fifth column is shown to give an idea of the complexity of the optimization process. A simulationbased optimization approach would require that very same number of simulations, thus taking a very long time. For instance, the optimization of would approximately require 2500 FxP simulations of 5000 input data. Considering the number of estimations required, the optimization times are extremely fast, ranging from 0.02 secs to 7.26 secs. The speedups obtained in comparison to a simulationbased approach are staggering; boosts from ×221 to ×3235 are obtained. The average boost is ×1486 which proves the advantage of our approach in terms of computation time.
In summary, results show that our approach enables fast and accurate WLO of both LTI and nonlinear DSP algorithms.
7. Conclusions
A novel noise estimation method based on the use of Affine Arithmetic has been presented. This method allows to obtain fast and accurate estimates of the quantization noise at the output of the FxP description of a DSP algorithm. The estimator can be used to perform complex WLO in standard time, leading to significant hardware cost reductions. The method can be applied to differentiable nonlinear DSP algorithms with and without feedbacks.
 (i)
the proposal of a novel AAbased quantization noise estimation for LTI algorithms,
 (ii)
the proposal of a novel AAbased quantization noise estimation for nonlinear algorithms with and without feedbacks,
 (iii)
the average estimation error for LTI systems is smaller than 2%,
 (iv)
the average estimation error for nonlinear systems is smaller than 17%,
 (v)
the computation time of WLO is boosted up to ×3235 (average of ×1486),
The reduction of the computation time of the noise parameterization process, specially in the presence of loops, is to be approached in the near future. Also, the improvement of the quantization model for nonlinear operations is perceived as an interesting research line.
Declarations
Acknowledgments
The authors would like to thank the work of the anonymous reviewers. This work was supported by the Spanish Ministry of Science and Innovation under project TEC200914219C0302.
Authors’ Affiliations
References
 Fang CF, Chen T, Rutenbar RA: Floatingpoint error analysis based on affine arithmetic. IEEE International Conference on Accoustics, Speech, and Signal Processing, April 2003, Hong Kong 561564.Google Scholar
 Gaffar A, Mencer O, Luk W, Cheung P, Shirazi N: Floatingpoint bitwidth analysis via automatic differentiation. International Conference on Field Programmable Technology, 2002 158165.Google Scholar
 Gaffar AA, Mencer O, Luk W, Cheung PYK: Unifying bitwidth optimisation for fixedpoint and floatingpoint designs. Proceedings of the 12th Annual IEEE Symposium on FieldProgrammable Custom Computing Machines (FCCM '04), April 2004 7988.View ArticleGoogle Scholar
 Caffarena G, Constantinides GA, Cheung PYK, Carreras C, NietoTaladriz O: Optimal combined wordlength allocation and architectural synthesis of digital signal processing circuits. IEEE Transactions on Circuits and Systems II 2006, 53(5):339343.View ArticleGoogle Scholar
 Catthoor F, De Man H, Vandewalle J: Simulated annealing based optimization of coefficient and data wordlengths in digital filters. International Journal of Circuit Theory and Applications 1988, 16(4):371390. 10.1002/cta.4490160404View ArticleGoogle Scholar
 Constantinides GA, Cheung PYK, Luk W: Wordlength optimization for linear digital signal processing. IEEE Transactions on ComputerAided Design of Integrated Circuits and Systems 2003, 22(10):14321442. 10.1109/TCAD.2003.818119View ArticleGoogle Scholar
 Sung W, Kum K: Simulationbased wordlength optimization method for fixedpoint digital signal processing systems. IEEE Transactions on Signal Processing 1995, 43(12):30873090. 10.1109/78.476465View ArticleGoogle Scholar
 Constantinides GA, Woeginger GJ: The complexity of multiple wordlength assignment. Applied Mathematics Letters 2002, 15(2):137140. 10.1016/S08939659(01)001070MathSciNetView ArticleMATHGoogle Scholar
 Caffarena G, Fernandez A, Carreras C, NietoTaladriz O: Fixedpoint refinement of OFDMbased adaptive equalizers: a heuristic approach. European Signal Processing Conference, 2004 13531356.Google Scholar
 Cantin MA, Savaria Y, Lavoie P: A comparison of automatic word length optimization procedures. IEEE International Symposium on Circuits and Systems, May 2002 612615.Google Scholar
 López JA, Caffarena G, Carreras C, NietoTaladriz O: Fast and accurate computation of the roundoff noise of linear timeinvariant systems. IET Circuits, Devices and Systems 2008, 2(4):393408. 10.1049/ietcds:20070198View ArticleGoogle Scholar
 Constantinides G: Perturbation analysis for wordlength optimization. IEEE Symposium on FieldProgrammable Custom Computing Machines, 2003 8190.Google Scholar
 Menard D, Rocher R, Scalart P, Sentieys O: SQNR determination in nonlinear and nonrecursive fixedpoint systems. European Signal Processing Conference, 2004 13491352.Google Scholar
 Shi C, Brodersen RW: A perturbation theory on statistical quantization effects in fixedpoint DSP with nonstationary inputs. Proceedings of IEEE International Symposium on Circuits and Systems, May 2004, Vancouver, Canada 3: 373376.Google Scholar
 Lee DU, Gaffar AA, Cheung RCC, Mencer O, Luk W, Constantinides GA: Accuracyguaranteed bitwidth optimization. IEEE Transactions on ComputerAided Design of Integrated Circuits and Systems 2006, 25(10):19901999.View ArticleGoogle Scholar
 Menard D, Sentieys O: A methodology for evaluating the precision of fixedpoint systems. IEEE International Conference on Acoustic, Speech, and Signal Processing, May 2002 31523155.Google Scholar
 Rocher R, Menard D, Herve N, Sentieys O: Fixedpoint configurable hardware components. EURASIP Journal on Embedded Systems 2006, 2006:13.Google Scholar
 Kim S, Kum KII, Sung W: Fixedpoint optimization utility for C and C++ based digital signal processing programs. IEEE Transactions on Circuits and Systems II 1998, 45(11):14551464. 10.1109/82.735357View ArticleGoogle Scholar
 Willems M, Keding H, Grótker T, Meyr H: FRIDGE: an interactive fixedpoint code generation environment for Hw/Swcodesign. IEEE Conference on Acoustics, Speech and Signal Processing, 1997, Munich, Germany 687690.Google Scholar
 Chang M, Hauck S: Precis: a designtime precision aanalysis tool. IEEE Symposium on FieldProgrammable Custom Computing Machines, 2002 229238.View ArticleGoogle Scholar
 Chang ML, Hauck S: Précis: a usercentric wordlength optimization tool. IEEE Design and Test of Computers 2005, 22(4):349361. 10.1109/MDT.2005.92View ArticleGoogle Scholar
 Cmar R, Rijnders L, Schaumont P, Vernalde S, Bolsens I: A methodology and design environment for DSP ASIC fixed point refinement. Proceedings of the Conference on Design, Automation and Test in Europe, 1999 56.Google Scholar
 Benedetti A, Perona P: Bitwidth optimization for configurable DSP's by multiinterval analysis. Proceedings of the 34th Asilomar Conference on Signals, Systems and Computers, November 2000 355359.Google Scholar
 Mahlke S, Ravindran R, Schlansker M, Schreiber R, Sherwood T: Bitwidth cognizant architecture synthesis of custom hardware accelerators. IEEE Transactions on ComputerAided Design of Integrated Circuits and Systems 2001, 20(11):13551371. 10.1109/43.959864View ArticleGoogle Scholar
 Stephenson M, Babb J, Amarasinghe S: Bitwidth analysis with application to silicon compilation. SIGPLAN Conference on Programming Language Design and Implementation (PLDI '00), June 2000 108120.Google Scholar
 Fang CF, Rutenbar RA, Chen T: Fast, accurate static analysis for fixedpoint finiteprecision effects in DSP designs. IEEE/ACM International Conference on Computer Aided Design (ICCAD '03), November 2003 275282.Google Scholar
 López JA, Caffarena G, Carreras C, NietoTaladriz O: Characterization of the quantization properties of similarityrelated DSP structures by means of interval simulations. Proceedings of the 37th Asilomar Conference on Signals, Systems and Computers, November 2003 22082212.Google Scholar
 Wadekar SA, Parker AC: Accuracy sensitive wordlength selection for algorithm optimization. Proceedings of the IEEE International Conference on Computer Design, October 1998 5461.Google Scholar
 Hayes B: A lucid interval. American Scientist 2003, 91(6):484488.View ArticleGoogle Scholar
 Carreras C, López JA, NietoTaladriz O: Bitwidth selection for datapath implementations. International Symposium on System Synthesis, 1999 114.View ArticleGoogle Scholar
 Stolfi J, Figueiredo LH: SelfValidated Numerical Methods and Applications. In Brazilian Mathematics Colloquium Monograph. IMPA, Rio de Janeiro, Brazil; 1997.Google Scholar
 Oppenheim AV, Schafer RW: DiscreteTime Signal Processing. PrenticeHall, Englewood Cliffs, NJ, USA; 1987.MATHGoogle Scholar
 Constantinides GA, Cheung PYK, Luk W: Truncation noise in fixedpoint SFGs. Electronics Letters 1999, 35(23):20132014.View ArticleGoogle Scholar
 Jackson LB: Roundoffnoise analysis for fixedpoint digital filters realized in cascade or parallel form. IEEE Transactions on Audio and Electroacoustics 1970, 18(2):107122. 10.1109/TAU.1970.1162084View ArticleGoogle Scholar
 Oppenheim AV, Weinstein CJ: Effects of finite register length in digital filtering and the fast Fourier transform. Proceedings of the IEEE 1972, 60(8):957976.View ArticleGoogle Scholar
 López J: Evaluación de los Efectos de Cuantificación en las Estructuras de Filtros Digitales Mediante Técnicas de Simulación Basadas en Extensiones de Intervalos, Ph.D. thesis. Universidad Politécnica de Madrid; 2004.Google Scholar
 López JA, Carreras C, NietoTaladriz O: Improved intervalbased characterization of fixedpoint LTI systems with feedback loops. IEEE Transactions on ComputerAided Design of Integrated Circuits and Systems 2007, 26(11):19231933.View ArticleGoogle Scholar
 Lopez J, Caffarena G, Carreras C, NietoTaladriz O: Analysis of limit cycles by means of affine arithmetic computeraided tests. European Signal Processing Conference, 2004 991994.Google Scholar
 Parhi KK: VLSI Digital Signal Processing Systems: Design and Implementation. Wiley, New York, NY, USA; 1999.Google Scholar
 Li G, Zhao Z: On the generalized DFIIt structure and its statespace realization in digital filter implementation. IEEE Transactions on Circuits and Systems I 2004, 51(4):769778. 10.1109/TCSI.2004.823652View ArticleGoogle Scholar
 Fernandez Herrero A, JiménezPacheco A, Caffarena G, Casajus Quiros J: Design and implementation of a hardware module for equalisation in a 4G mimo receiver. International Conference on Field Programmable Logic and Applications (FPL '06), August 2006 765768.Google Scholar
 Ogunfunmi T: Adaptive Nonlinear System Identification: The Volterra and Wiener Approaches. Springer; 2007.View ArticleMATHGoogle Scholar
Copyright
This article is published under license to BioMed Central Ltd. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.