jfriedt / IFCS2018 article

Download as
Browse Code »

Commit 365465b1b586da014bd0c9a31a91139045fdef9b

Authored by jfriedt
1 parent 9db1d56ab9
Exists in master

notations

Showing 1 changed file with 13 additions and 9 deletions Inline Diff

ifcs2018_proceeding.tex
Diff comments   View file @ 365465b
% JMF : revoir l'abstract : on y avait mis le Zynq7010 de la redpitaya en montrant 1 1 % JMF : revoir l'abstract : on y avait mis le Zynq7010 de la redpitaya en montrant
% comment optimiser les perfs a surface finie. Ici aussi on tombait dans le cas ou` 2 2 % comment optimiser les perfs a surface finie. Ici aussi on tombait dans le cas ou`
% la solution a 1 seul FIR n'etait simplement pas synthetisable => fusionner les deux 3 3 % la solution a 1 seul FIR n'etait simplement pas synthetisable => fusionner les deux
% contributions pour le papier TUFFC 4 4 % contributions pour le papier TUFFC
5 5
\documentclass[a4paper,conference]{IEEEtran/IEEEtran} 6 6 \documentclass[a4paper,conference]{IEEEtran/IEEEtran}
\usepackage{graphicx,color,hyperref} 7 7 \usepackage{graphicx,color,hyperref}
\usepackage{amsfonts} 8 8 \usepackage{amsfonts}
\usepackage{amsthm} 9 9 \usepackage{amsthm}
\usepackage{amssymb} 10 10 \usepackage{amssymb}
\usepackage{amsmath} 11 11 \usepackage{amsmath}
\usepackage{algorithm2e} 12 12 \usepackage{algorithm2e}
\usepackage{url,balance} 13 13 \usepackage{url,balance}
\usepackage[normalem]{ulem} 14 14 \usepackage[normalem]{ulem}
% correct bad hyphenation here 15 15 % correct bad hyphenation here
\hyphenation{op-tical net-works semi-conduc-tor} 16 16 \hyphenation{op-tical net-works semi-conduc-tor}
\textheight=26cm 17 17 \textheight=26cm
\setlength{\footskip}{30pt} 18 18 \setlength{\footskip}{30pt}
\pagenumbering{gobble} 19 19 \pagenumbering{gobble}
\begin{document} 20 20 \begin{document}
\title{Filter optimization for real time digital processing of radiofrequency signals: application 21 21 \title{Filter optimization for real time digital processing of radiofrequency signals: application
to oscillator metrology} 22 22 to oscillator metrology}
23 23
\author{\IEEEauthorblockN{A. Hugeat\IEEEauthorrefmark{1}\IEEEauthorrefmark{2}, J. Bernard\IEEEauthorrefmark{2}, 24 24 \author{\IEEEauthorblockN{A. Hugeat\IEEEauthorrefmark{1}\IEEEauthorrefmark{2}, J. Bernard\IEEEauthorrefmark{2},
G. Goavec-M\'erou\IEEEauthorrefmark{1}, 25 25 G. Goavec-M\'erou\IEEEauthorrefmark{1},
P.-Y. Bourgeois\IEEEauthorrefmark{1}, J.-M. Friedt\IEEEauthorrefmark{1}} 26 26 P.-Y. Bourgeois\IEEEauthorrefmark{1}, J.-M. Friedt\IEEEauthorrefmark{1}}
\IEEEauthorblockA{\IEEEauthorrefmark{1}FEMTO-ST, Time \& Frequency department, Besan\c con, France } 27 27 \IEEEauthorblockA{\IEEEauthorrefmark{1}FEMTO-ST, Time \& Frequency department, Besan\c con, France }
\IEEEauthorblockA{\IEEEauthorrefmark{2}FEMTO-ST, Computer Science department DISC, Besan\c con, France \\ 28 28 \IEEEauthorblockA{\IEEEauthorrefmark{2}FEMTO-ST, Computer Science department DISC, Besan\c con, France \\
Email: \{pyb2,jmfriedt\}@femto-st.fr} 29 29 Email: \{pyb2,jmfriedt\}@femto-st.fr}
} 30 30 }
\maketitle 31 31 \maketitle
\thispagestyle{plain} 32 32 \thispagestyle{plain}
\pagestyle{plain} 33 33 \pagestyle{plain}
\newtheorem{definition}{Definition} 34 34 \newtheorem{definition}{Definition}
35 35
\begin{abstract} 36 36 \begin{abstract}
Software Defined Radio (SDR) provides stability, flexibility and reconfigurability to 37 37 Software Defined Radio (SDR) provides stability, flexibility and reconfigurability to
radiofrequency signal processing. Applied to oscillator characterization in the context 38 38 radiofrequency signal processing. Applied to oscillator characterization in the context
of ultrastable clocks, stringent filtering requirements are defined by spurious signal or 39 39 of ultrastable clocks, stringent filtering requirements are defined by spurious signal or
noise rejection needs. Since real time radiofrequency processing must be performed in a 40 40 noise rejection needs. Since real time radiofrequency processing must be performed in a
Field Programmable Array to meet timing constraints, we investigate optimization strategies 41 41 Field Programmable Array to meet timing constraints, we investigate optimization strategies
to design filters meeting rejection characteristics while limiting the hardware resources 42 42 to design filters meeting rejection characteristics while limiting the hardware resources
required and keeping timing constraints within the targeted measurement bandwidths. 43 43 required and keeping timing constraints within the targeted measurement bandwidths.
\end{abstract} 44 44 \end{abstract}
45 45
\begin{IEEEkeywords} 46 46 \begin{IEEEkeywords}
Software Defined Radio, Mixed-Integer Linear Programming, Finite Impulse Response filter 47 47 Software Defined Radio, Mixed-Integer Linear Programming, Finite Impulse Response filter
\end{IEEEkeywords} 48 48 \end{IEEEkeywords}
49 49
\section{Digital signal processing of ultrastable clock signals} 50 50 \section{Digital signal processing of ultrastable clock signals}
51 51
Analog oscillator phase noise characteristics are classically performed by downconverting 52 52 Analog oscillator phase noise characteristics are classically performed by downconverting
the radiofrequency signal using a saturated mixer to bring the radiofrequency signal to baseband, 53 53 the radiofrequency signal using a saturated mixer to bring the radiofrequency signal to baseband,
followed by a Fourier analysis of the beat signal to analyze phase fluctuations close to carrier. In 54 54 followed by a Fourier analysis of the beat signal to analyze phase fluctuations close to carrier. In
a fully digital approach, the radiofrequency signal is digitized and numerically downconverted by 55 55 a fully digital approach, the radiofrequency signal is digitized and numerically downconverted by
multiplying the samples with a local numerically controlled oscillator (Fig. \ref{schema}) \cite{rsi}. 56 56 multiplying the samples with a local numerically controlled oscillator (Fig. \ref{schema}) \cite{rsi}.
57 57
\begin{figure}[h!tb] 58 58 \begin{figure}[h!tb]
\begin{center} 59 59 \begin{center}
\includegraphics[width=.8\linewidth]{images/schema} 60 60 \includegraphics[width=.8\linewidth]{images/schema}
\end{center} 61 61 \end{center}
\caption{Fully digital oscillator phase noise characterization: the Device Under Test 62 62 \caption{Fully digital oscillator phase noise characterization: the Device Under Test
(DUT) signal is sampled by the radiofrequency grade Analog to Digital Converter (ADC) and 63 63 (DUT) signal is sampled by the radiofrequency grade Analog to Digital Converter (ADC) and
downconverted by mixing with a Numerically Controlled Oscillator (NCO). Unwanted signals 64 64 downconverted by mixing with a Numerically Controlled Oscillator (NCO). Unwanted signals
and noise aliases are rejected by a Low Pass Filter (LPF) implemented as a cascade of Finite 65 65 and noise aliases are rejected by a Low Pass Filter (LPF) implemented as a cascade of Finite
Impulse Response (FIR) filters. The signal is then decimated before a Fourier analysis displays 66 66 Impulse Response (FIR) filters. The signal is then decimated before a Fourier analysis displays
the spectral characteristics of the phase fluctuations.} 67 67 the spectral characteristics of the phase fluctuations.}
\label{schema} 68 68 \label{schema}
\end{figure} 69 69 \end{figure}
70 70
As with the analog mixer, 71 71 As with the analog mixer,
the non-linear behavior of the downconverter introduces noise or spurious signal aliasing as 72 72 the non-linear behavior of the downconverter introduces noise or spurious signal aliasing as
well as the generation of the frequency sum signal in addition to the frequency difference. 73 73 well as the generation of the frequency sum signal in addition to the frequency difference.
These unwanted spectral characteristics must be rejected before decimating the data stream 74 74 These unwanted spectral characteristics must be rejected before decimating the data stream
for the phase noise spectral characterization \cite{andrich2018high}. The characteristics introduced between the 75 75 for the phase noise spectral characterization \cite{andrich2018high}. The characteristics introduced between the
downconverter 76 76 downconverter
and the decimation processing blocks are core characteristics of an oscillator characterization 77 77 and the decimation processing blocks are core characteristics of an oscillator characterization
system, and must reject out-of-band signals below the targeted phase noise -- typically in the 78 78 system, and must reject out-of-band signals below the targeted phase noise -- typically in the
sub -170~dBc/Hz for ultrastable oscillator we aim at characterizing. The filter blocks will 79 79 sub -170~dBc/Hz for ultrastable oscillator we aim at characterizing. The filter blocks will
use most resources of the Field Programmable Gate Array (FPGA) used to process the radiofrequency 80 80 use most resources of the Field Programmable Gate Array (FPGA) used to process the radiofrequency
datastream: optimizing the performance of the filter while reducing the needed resources is 81 81 datastream: optimizing the performance of the filter while reducing the needed resources is
hence tackled in a systematic approach using optimization techniques. Most significantly, we 82 82 hence tackled in a systematic approach using optimization techniques. Most significantly, we
tackle the issue by attempting to cascade multiple Finite Impulse Response (FIR) filters with 83 83 tackle the issue by attempting to cascade multiple Finite Impulse Response (FIR) filters with
tunable number of coefficients and tunable number of bits representing the coefficients and the 84 84 tunable number of coefficients and tunable number of bits representing the coefficients and the
data being processed. 85 85 data being processed.
86 86
\section{Finite impulse response filter} 87 87 \section{Finite impulse response filter}
88 88
We select FIR filter for their unconditional stability and ease of design. A FIR filter is defined 89 89 We select FIR filter for their unconditional stability and ease of design. A FIR filter is defined
by a set of weights $b_k$ applied to the inputs $x_k$ through a convolution to generate the 90 90 by a set of weights $b_k$ applied to the inputs $x_k$ through a convolution to generate the
outputs $y_k$ 91 91 outputs $y_k$
$$y_n=\sum_{k=0}^N b_k x_{n-k}$$ 92 92 $$y_n=\sum_{k=0}^N b_k x_{n-k}$$
93 93
As opposed to an implementation on a general purpose processor in which word size is defined by the 94 94 As opposed to an implementation on a general purpose processor in which word size is defined by the
processor architecture, implementing such a filter on an FPGA offer more degrees of freedom since 95 95 processor architecture, implementing such a filter on an FPGA offer more degrees of freedom since
not only the coefficient values and number of taps must be defined, but also the number of bits 96 96 not only the coefficient values and number of taps must be defined, but also the number of bits
defining the coefficients and the sample size. For this reason, and because we consider pipeline 97 97 defining the coefficients and the sample size. For this reason, and because we consider pipeline
processing (as opposed to First-In, First-Out FIFO memory batch processing) of radiofrequency 98 98 processing (as opposed to First-In, First-Out FIFO memory batch processing) of radiofrequency
signals, High Level Synthesis (HLS) languages \cite{kasbah2008multigrid} are not considered but 99 99 signals, High Level Synthesis (HLS) languages \cite{kasbah2008multigrid} are not considered but
the problem is tackled at the Very-high-speed-integrated-circuit Hardware Description Language (VHDL) level. 100 100 the problem is tackled at the Very-high-speed-integrated-circuit Hardware Description Language (VHDL) level.
Since latency is not an issue in a openloop phase noise characterization instrument, the large 101 101 Since latency is not an issue in a openloop phase noise characterization instrument, the large
numbre of taps in the FIR, as opposed to the shorter Infinite Impulse Response (IIR) filter, 102 102 numbre of taps in the FIR, as opposed to the shorter Infinite Impulse Response (IIR) filter,
is not considered as an issue as would be in a closed loop system. 103 103 is not considered as an issue as would be in a closed loop system.
104 104
The coefficients are classically expressed as floating point values. However, this binary 105 105 The coefficients are classically expressed as floating point values. However, this binary
number representation is not efficient for fast arithmetic computation by an FPGA. Instead, 106 106 number representation is not efficient for fast arithmetic computation by an FPGA. Instead,
we select to quantify these floating point values into integer values. This quantization 107 107 we select to quantify these floating point values into integer values. This quantization
will result in some precision loss. 108 108 will result in some precision loss.
109 109
%As illustrated in Fig. \ref{float_vs_int}, we see that we aren't 110 110 %As illustrated in Fig. \ref{float_vs_int}, we see that we aren't
%need too coefficients or too sample size. If we have lot of coefficients but a small sample size, 111 111 %need too coefficients or too sample size. If we have lot of coefficients but a small sample size,
%the first and last are equal to zero. But if we have too sample size for few coefficients that not improve the quality. 112 112 %the first and last are equal to zero. But if we have too sample size for few coefficients that not improve the quality.
113 113
% JMF je ne comprends pas la derniere phrase ci-dessus ni la figure ci dessous 114 114 % JMF je ne comprends pas la derniere phrase ci-dessus ni la figure ci dessous
% AH en gros je voulais dire que prendre trop peu de bit avec trop de coeff, ça induit ta figure (bien mieux faite que moi) 115 115 % AH en gros je voulais dire que prendre trop peu de bit avec trop de coeff, ça induit ta figure (bien mieux faite que moi)
% et que l'inverse trop de bit sur pas assez de coeff on ne gagne rien, je vais essayer de la reformuler 116 116 % et que l'inverse trop de bit sur pas assez de coeff on ne gagne rien, je vais essayer de la reformuler
117 117
%\begin{figure}[h!tb] 118 118 %\begin{figure}[h!tb]
%\includegraphics[width=\linewidth]{images/float-vs-integer.pdf} 119 119 %\includegraphics[width=\linewidth]{images/float-vs-integer.pdf}
%\caption{Impact of the quantization resolution of the coefficients} 120 120 %\caption{Impact of the quantization resolution of the coefficients}
%\label{float_vs_int} 121 121 %\label{float_vs_int}
%\end{figure} 122 122 %\end{figure}
123 123
\begin{figure}[h!tb] 124 124 \begin{figure}[h!tb]
\includegraphics[width=\linewidth]{images/demo_filtre} 125 125 \includegraphics[width=\linewidth]{images/demo_filtre}
\caption{Impact of the quantization resolution of the coefficients: the quantization is 126 126 \caption{Impact of the quantization resolution of the coefficients: the quantization is
set to 6~bits -- with the horizontal black lines indicating $\pm$1 least significant bit -- setting 127 127 set to 6~bits -- with the horizontal black lines indicating $\pm$1 least significant bit -- setting
the 30~first and 30~last coefficients out of the initial 128~band-pass 128 128 the 30~first and 30~last coefficients out of the initial 128~band-pass
filter coefficients to 0 (red dots).} 129 129 filter coefficients to 0 (red dots).}
\label{float_vs_int} 130 130 \label{float_vs_int}
\end{figure} 131 131 \end{figure}
132 132
The tradeoff between quantization resolution and number of coefficients when considering 133 133 The tradeoff between quantization resolution and number of coefficients when considering
integer operations is not trivial. As an illustration of the issue related to the 134 134 integer operations is not trivial. As an illustration of the issue related to the
relation between number of fiter taps and quantization, Fig. \ref{float_vs_int} exhibits 135 135 relation between number of fiter taps and quantization, Fig. \ref{float_vs_int} exhibits
a 128-coefficient FIR bandpass filter designed using floating point numbers (blue). Upon 136 136 a 128-coefficient FIR bandpass filter designed using floating point numbers (blue). Upon
quantization on 6~bit integers, 60 of the 128~coefficients in the beginning and end of the 137 137 quantization on 6~bit integers, 60 of the 128~coefficients in the beginning and end of the
taps become null, making the large number of coefficients irrelevant and allowing to save 138 138 taps become null, making the large number of coefficients irrelevant and allowing to save
processing resource by shrinking the filter length. This tradeoff aimed at minimizing resources 139 139 processing resource by shrinking the filter length. This tradeoff aimed at minimizing resources
to reach a given rejection level, or maximizing out of band rejection for a given computational 140 140 to reach a given rejection level, or maximizing out of band rejection for a given computational
resource, will drive the investigation on cascading filters designed with varying tap resolution 141 141 resource, will drive the investigation on cascading filters designed with varying tap resolution
and tap length, as will be shown in the next section. Indeed, our development strategy closely 142 142 and tap length, as will be shown in the next section. Indeed, our development strategy closely
follows the skeleton approach \cite{crookes1998environment, crookes2000design, benkrid2002towards} 143 143 follows the skeleton approach \cite{crookes1998environment, crookes2000design, benkrid2002towards}
in which basic blocks are defined and characterized before being assembled \cite{hide} 144 144 in which basic blocks are defined and characterized before being assembled \cite{hide}
in a complete processing chain. In our case, assembling the filter blocks is a simpler block 145 145 in a complete processing chain. In our case, assembling the filter blocks is a simpler block
combination process since we assume a single value to be processed and a single value to be 146 146 combination process since we assume a single value to be processed and a single value to be
generated at each clock cycle. The FIR filters will not be considered to decimate in the 147 147 generated at each clock cycle. The FIR filters will not be considered to decimate in the
current implementation: the decimation is assumed to be located after the FIR cascade at the 148 148 current implementation: the decimation is assumed to be located after the FIR cascade at the
moment. 149 149 moment.
150 150
\section{Filter optimization} 151 151 \section{Filter optimization}
152 152
A basic approach for implementing the FIR filter is to compute the transfer function of 153 153 A basic approach for implementing the FIR filter is to compute the transfer function of
a monolithic filter: this single filter defines all coefficients with the same resolution 154 154 a monolithic filter: this single filter defines all coefficients with the same resolution
(number of bits) and processes data represented with their own resolution. Meeting the 155 155 (number of bits) and processes data represented with their own resolution. Meeting the
filter shape requires a large number of coefficients, limited by resources of the FPGA since 156 156 filter shape requires a large number of coefficients, limited by resources of the FPGA since
this filter must process data stream at the radiofrequency sampling rate after the mixer. 157 157 this filter must process data stream at the radiofrequency sampling rate after the mixer.
158 158
An optimization problem \cite{leung2004handbook} aims at improving one or many 159 159 An optimization problem \cite{leung2004handbook} aims at improving one or many
performance criteria within a constrained resource environment. Amongst the tools 160 160 performance criteria within a constrained resource environment. Amongst the tools
developed to meet this aim, Mixed-Integer Linear Programming (MILP) provides the framework to 161 161 developed to meet this aim, Mixed-Integer Linear Programming (MILP) provides the framework to
formally define the stated problem and search for an optimal use of available 162 162 formally define the stated problem and search for an optimal use of available
resources \cite{yu2007design, kodek1980design}. 163 163 resources \cite{yu2007design, kodek1980design}.
164 164
First we need to ensure that our problem is a real optimization problem. When 165 165 First we need to ensure that our problem is a real optimization problem. When
designing a processing function in the FPGA, we aim at meeting some requirement such as 166 166 designing a processing function in the FPGA, we aim at meeting some requirement such as
the throughput, the computation time or the noise rejection noise. However, due to limited 167 167 the throughput, the computation time or the noise rejection noise. However, due to limited
resources to design the process like BRAM (high performance RAM), DSP (Digital Signal Processor) 168 168 resources to design the process like BRAM (high performance RAM), DSP (Digital Signal Processor)
or LUT (Look Up Table), a tradeoff must be generally searched between performance and available 169 169 or LUT (Look Up Table), a tradeoff must be generally searched between performance and available
computational resources: optimizing some criteria within finite, limited 170 170 computational resources: optimizing some criteria within finite, limited
resources indeed matches the definition of a classical optimization problem. 171 171 resources indeed matches the definition of a classical optimization problem.
172 172
Specifically the degrees of freedom when addressing the problem of replacing the single monolithic 173 173 Specifically the degrees of freedom when addressing the problem of replacing the single monolithic
FIR with a cascade of optimized filters are the number of coefficients $N_i$ of each filter $i$ and 174 174 FIR with a cascade of optimized filters are the number of coefficients $N_i$ of each filter $i$ and
the number of bits $C_i$ representing the coefficients. Because each FIR in the chain is fed the output of the previous stage, 175 175 the number of bits $C_i$ representing the coefficients. Because each FIR in the chain is fed the output of the previous stage,
the optimization of the complete processing chain within a constrained resource environment is not 176 176 the optimization of the complete processing chain within a constrained resource environment is not
trivial. The resource occupation of a FIR filter is considered as $C_i \times N_i$ which is 177 177 trivial. The resource occupation of a FIR filter is considered as $C_i \times N_i$ which aims
the number of bits needed in a worst case condition to represent the output of the FIR. Such an 178 178 at approximating the number of bits needed in a worst case condition to represent the output of the
179 FIR. Indeed, the number of bits generated by the FIR is $(C_i+D_i)\times\log_2(N_i)$ with $D_i$
180 the number of bits needed to represent the data $x_k$ generated by the previous stage, but the
181 $\log$ function is avoided for its incompatibility with a linear programming description, and
182 the simple product is approximated as the number of gates needed to perform the calculation. Such an
occupied area estimate assumes that the number of gates scales as the number of bits and the number 179 183 occupied area estimate assumes that the number of gates scales as the number of bits and the number
of coefficients, but does not account for the detailed implementation of the hardware. Indeed, 180 184 of coefficients, but does not account for the detailed implementation of the hardware. Indeed,
various FPGA implementations will provide different hardware functionalities, and we shall consider 181 185 various FPGA implementations will provide different hardware functionalities, and we shall consider
at the end of the design a synthesis step using vendor software to assess the validity of the solution 182 186 at the end of the design a synthesis step using vendor software to assess the validity of the solution
found. As an example of the limitation linked to the lack of detailed hardware consideration, Block Random 183 187 found. As an example of the limitation linked to the lack of detailed hardware consideration, Block Random
Access Memory (BRAM) used to store filter coefficients are not shared amongst filters, and multiplications 184 188 Access Memory (BRAM) used to store filter coefficients are not shared amongst filters, and multiplications
are most efficiently implemented by using DSP blocks whose input word 185 189 are most efficiently implemented by using DSP blocks whose input word
size is finite. DSPs are a scarce resource to be saved in a practical implementation. Keeping a high 186 190 size is finite. DSPs are a scarce resource to be saved in a practical implementation. Keeping a high
abstraction on the resource occupation is nevertheless selected in the following discussion in order 187 191 abstraction on the resource occupation is nevertheless selected in the following discussion in order
to leave enough degrees of freedom in the problem to try and find original solutions: too many 188 192 to leave enough degrees of freedom in the problem to try and find original solutions: too many
constraints in the initial statement of the problem leave little room for finding an optimal solution. 189 193 constraints in the initial statement of the problem leave little room for finding an optimal solution.
190 194
\begin{figure}[h!tb] 191 195 \begin{figure}[h!tb]
\begin{center} 192 196 \begin{center}
\includegraphics[width=.5\linewidth]{schema2} 193 197 \includegraphics[width=.5\linewidth]{schema2}
\caption{Shape of the filter transmitted power $P$ as a function of frequency: 194 198 \caption{Shape of the filter transmitted power $P$ as a function of frequency:
the bandpass BP is considered to occupy the initial 195 199 the bandpass BP is considered to occupy the initial
40\% of the Nyquist frequency range, the stopband the last 40\%, allowing 20\% transition 196 200 40\% of the Nyquist frequency range, the stopband the last 40\%, allowing 20\% transition
width.} 197 201 width.}
\label{rejection-shape} 198 202 \label{rejection-shape}
\end{center} 199 203 \end{center}
\end{figure} 200 204 \end{figure}
201 205
Following these considerations, the model is expressed as: 202 206 Following these considerations, the model is expressed as:
\begin{align} 203 207 \begin{align}
\begin{cases} 204 208 \begin{cases}
\mathcal{R}_i &= \mathcal{F}(N_i, C_i)\\ 205 209 \mathcal{R}_i &= \mathcal{F}(N_i, C_i)\\
\mathcal{A}_i &= N_i * C_i\\ 206 210 \mathcal{A}_i &= N_i \times C_i\\
\Delta_i &= \Delta _{i-1} + \mathcal{P}_i 207 211 \Delta_i &= \Delta _{i-1} + \mathcal{P}_i
\end{cases} 208 212 \end{cases}
\label{model-FIR} 209 213 \label{model-FIR}
\end{align} 210 214 \end{align}
To explain the system \ref{model-FIR}, $\mathcal{R}_i$ represents the stopband rejection dependence with $N_i$ and $C_i$, $\mathcal{A}$ 211 215 To explain the system \ref{model-FIR}, $\mathcal{R}_i$ represents the stopband rejection dependence with $N_i$ and $C_i$, $\mathcal{A}_i$
is a theoretical area occupation of the processing block on the FPGA as discussed earlier, and $\Delta_i$ is the total rejection for the current stage $i$. 212 216 is a theoretical area occupation of the processing block on the FPGA as discussed earlier, and $\Delta_i$ is the total rejection for the current stage $i$.
Since the function $\mathcal{F}$ cannot be explictly expressed, we run simulations to determine the rejection depending 213 217 Since the function $\mathcal{F}$ cannot be explictly expressed, we run simulations to determine the rejection depending
on $N_i$ and $C_i$. However, selecting the right filter requires a clear definition of the rejection criterion. Selecting an 214 218 on $N_i$ and $C_i$. However, selecting the right filter requires a clear definition of the rejection criterion. Selecting an
incorrect criterion will lead the linear program solver to produce a solution which might not meet the user requirements. 215 219 incorrect criterion will lead the linear program solver to produce a solution which might not meet the user requirements.
Hence, amongst various criteria including the mean or median value of the FIR response in the stopband as will 216 220 Hence, amongst various criteria including the mean or median value of the FIR response in the stopband as will
be illustrated lated (section \ref{median}), we have designed 217 221 be illustrated lated (section \ref{median}), we have designed
a criterion aimed at avoiding ripples in the passband and considering the maximum of the FIR spectral response in the stopband 218 222 a criterion aimed at avoiding ripples in the passband and considering the maximum of the FIR spectral response in the stopband
(Fig. \ref{rejection-shape}). The bandpass criterion is defined as the sum of the absolute values of the spectral response 219 223 (Fig. \ref{rejection-shape}). The bandpass criterion is defined as the sum of the absolute values of the spectral response
in the bandpass, reminiscent of a standard deviation of the spectral response: this criterion must be minimized to avoid 220 224 in the bandpass, reminiscent of a standard deviation of the spectral response: this criterion must be minimized to avoid
ripples in the passband. The stopband transfer function maximum must also be minimized in order to improve the filter 221 225 ripples in the passband. The stopband transfer function maximum must also be minimized in order to improve the filter
rejection capability. Weighing these two criteria allows designing the linear program to be solved. 222 226 rejection capability. Weighing these two criteria allows designing the linear program to be solved.
223 227
\begin{figure}[h!tb] 224 228 \begin{figure}[h!tb]
\includegraphics[width=\linewidth]{images/noise-rejection.pdf} 225 229 \includegraphics[width=\linewidth]{images/noise-rejection.pdf}
\caption{Rejection as a function of number of coefficients and number of bits} 226 230 \caption{Rejection as a function of number of coefficients and number of bits}
\label{noise-rejection} 227 231 \label{noise-rejection}
\end{figure} 228 232 \end{figure}
229 233
The objective function maximizes the noise rejection ($\max(\Delta_{i_{\max}})$) while keeping resource occupation below 230 234 The objective function maximizes the noise rejection ($\max(\Delta_{i_{\max}})$) while keeping resource
a user-defined threshold, or aims at minimizing the area needed to reach a given rejection ($\min(S_q)$ in 231 235 occupation below a user-defined threshold, or as will be discussed here, aims at minimizing the area
the forthcoming discussion, Eqs. \ref{cstr_size} and \ref{cstr_rejection}). 232 236 needed to reach a given rejection ($\min(S_q)$ in the forthcoming discussion, Eqs. \ref{cstr_size}
The MILP solver is allowed to choose the number of successive 233 237 and \ref{cstr_rejection}). The MILP solver is allowed to choose the number of successive
filters, within an upper bound. The last problem is to model the noise rejection. Since filter 234 238 filters, within an upper bound. The last problem is to model the noise rejection. Since filter
noise rejection capability is not modeled with linear equations, a look-up-table is generated 235 239 noise rejection capability is not modeled with linear equations, a look-up-table is generated
for multiple filter configurations in which the $C_i$, $D_i$ and $N_i$ parameters are varied: for each 236 240 for multiple filter configurations in which the $C_i$, $D_i$ and $N_i$ parameters are varied: for each
one of these conditions, the low-pass filter rejection is stored as computed by the frequency response 237 241 one of these conditions, the low-pass filter rejection is stored as computed by the frequency response
of the digital filter (Fig. \ref{noise-rejection}). Various rejection criteria have been investigated, 238 242 of the digital filter (Fig. \ref{noise-rejection}). Various rejection criteria have been investigated,
including mean value of the stopband response, median value of the stopband response, or as finally 239 243 including mean value of the stopband response, median value of the stopband response, or as finally
selected, maximum value in the stopband. An intuitive analysis of the chart of Fig. \ref{noise-rejection} 240 244 selected, maximum value in the stopband. An intuitive analysis of the chart of Fig. \ref{noise-rejection}
hints at an optimum 241 245 hints at an optimum
set of tap length and number of bit for representing the coefficients along the line of the pyramidal 242 246 set of tap length and number of bit for representing the coefficients along the line of the pyramidal
shaped rejection capability function. 243 247 shaped rejection capability function.
244 248
Linear program formalism for solving the problem is well documented: an objective function is 245 249 Linear program formalism for solving the problem is well documented: an objective function is
defined which is linearly dependent on the parameters to be optimized. Constraints are expressed 246 250 defined which is linearly dependent on the parameters to be optimized. Constraints are expressed
as linear equation and solved using one of the available solvers, in our case GLPK\cite{glpk}. 247 251 as linear equation and solved using one of the available solvers, in our case GLPK\cite{glpk}.
With the notation explain in system \ref{model-FIR}, we have defined our linear problem like this: 248 252 With the notation explain in system \ref{model-FIR}, we have defined our linear problem like this:
\paragraph{Variables} 249 253 \paragraph{Variables}
\begin{align*} 250 254 \begin{align*}
x_{i,j} \in \lbrace 0,1 \rbrace & \text{ $i$ is a given filter} \\ 251 255 x_{i,j} \in \lbrace 0,1 \rbrace & \text{ $i$ is a given filter} \\
& \text{ $j$ is the stage} \\ 252 256 & \text{ $j$ is the stage} \\
& \text{ If $x_{i,j}$ is equal to 1, the filter is selected} \\ 253 257 & \text{ If $x_{i,j}$ is equal to 1, the filter is selected} \\
\end{align*} 254 258 \end{align*}
\paragraph{Constants} 255 259 \paragraph{Constants}
\begin{align*} 256 260 \begin{align*}
\mathcal{F} = \lbrace F_1 ... F_p \rbrace & \text{ All possible filters}\\ 257 261 \mathcal{F} = \lbrace F_1 ... F_p \rbrace & \text{ All possible filters}\\
& \text{ $p$ is the number of different filters} \\ 258 262 & \text{ $p$ is the number of different filters} \\
% N(i) & \text{ % Constant to let the 259 263 % N(i) & \text{ % Constant to let the
% number of coefficients %} \\ & \text{ 260 264 % number of coefficients %} \\ & \text{
% for filter $i$}\\ 261 265 % for filter $i$}\\
% C(i) & \text{ % Constant to let the 262 266 % C(i) & \text{ % Constant to let the
% number of bits of %}\\ & \text{ 263 267 % number of bits of %}\\ & \text{
% each coefficient for filter $i$}\\ 264 268 % each coefficient for filter $i$}\\
\mathcal{S}_{\max} & \text{ Total space available inside the FPGA} 265 269 \mathcal{S}_{\max} & \text{ Total space available inside the FPGA}
\end{align*} 266 270 \end{align*}
\paragraph{Constraints} 267 271 \paragraph{Constraints}
\begin{align} 268 272 \begin{align}
1 \leq i \leq p & \nonumber\\ 269 273 1 \leq i \leq p & \nonumber\\
1 \leq j \leq q & \text{ $q$ is the max of filter stage} \nonumber \\ 270 274 1 \leq j \leq q & \text{ $q$ is the max of filter stage} \nonumber \\
\forall j, \mathlarger{\sum_{i}} x_{i,j} = 1 & \text{ At most one filter by stage} \nonumber\\ 271 275 \forall j, \mathlarger{\sum_{i}} x_{i,j} = 1 & \text{ At most one filter by stage} \nonumber\\
\mathcal{S}_0 = 0 & \text{ initial occupation} \nonumber\\ 272 276 \mathcal{S}_0 = 0 & \text{ initial occupation} \nonumber\\
\forall j, \mathcal{S}_j = \mathcal{S}_{j-1} + \mathlarger{\sum_i (x_{i,j} \times \mathcal{A}_i)} \label{cstr_size} \\ 273 277 \forall j, \mathcal{S}_j = \mathcal{S}_{j-1} + \mathlarger{\sum_i (x_{i,j} \times \mathcal{A}_i)} \label{cstr_size} \\
\mathcal{S} \leq \mathcal{S}_{\max}\nonumber \\ 274 278 \mathcal{S}_j \leq \mathcal{S}_{\max}\nonumber \\
\mathcal{N}_0 = 0 & \text{ initial rejection}\nonumber\\ 275 279 \mathcal{N}_0 = 0 & \text{ initial rejection}\nonumber\\
\forall j, \mathcal{N}_j = \mathcal{N}_{j-1} + \mathlarger{\sum_i (x_{i,j} \times \mathcal{R}_i)} \label{cstr_rejection} \\ 276 280 \forall j, \mathcal{N}_j = \mathcal{N}_{j-1} + \mathlarger{\sum_i (x_{i,j} \times \mathcal{R}_i)} \label{cstr_rejection} \\
\mathcal{N}_q \geqslant 160 & \text{ an user defined bound}\nonumber\\ 277 281 \mathcal{N}_q \geqslant 160 & \text{ an user defined bound}\nonumber\\
& \text{ (e.g. 160~dB here)}\nonumber\\\nonumber 278 282 & \text{ (e.g. 160~dB here)}\nonumber\\\nonumber
\end{align} 279 283 \end{align}
\paragraph{Goal} 280 284 \paragraph{Goal}
\begin{align*} 281 285 \begin{align*}
\min \mathcal{S}_q 282 286 \min \mathcal{S}_q
\end{align*} 283 287 \end{align*}
284 288
The constraint \ref{cstr_size} means the occupation for the current stage $j$ depends on 285 289 The constraint \ref{cstr_size} means the occupation for the current stage $j$ depends on
the previous occupation and the occupation of current selected filter (it is possible 286 290 the previous occupation and the occupation of current selected filter (it is possible
that no filter is selected for this stage). And the second one \ref{cstr_rejection} 287 291 that no filter is selected for this stage). And the second one \ref{cstr_rejection}
means the same thing but for the rejection, the rejection depends the previous rejection 288 292 means the same thing but for the rejection, the rejection depends the previous rejection
plus the rejection of selected filter. 289 293 plus the rejection of selected filter.
290 294
\subsection{Low bandpass ripple and maximum rejection criteria} 291 295 \subsection{Low bandpass ripple and maximum rejection criteria}
292 296
The MILP solver provides a solution to the problem by selecting a series of small FIR with 293 297 The MILP solver provides a solution to the problem by selecting a series of small FIR with
increasing number of bits representing data and coefficients as well as an increasing number 294 298 increasing number of bits representing data and coefficients as well as an increasing number
of coefficients, instead of a single monolithic filter. 295 299 of coefficients, instead of a single monolithic filter.
296 300
\begin{figure}[h!tb] 297 301 \begin{figure}[h!tb]
% \includegraphics[width=\linewidth]{images/compare-fir.pdf} 298 302 % \includegraphics[width=\linewidth]{images/compare-fir.pdf}
\includegraphics[width=\linewidth]{images/fir-mono-vs-fir-series-noise-fixe-jmf-light.pdf} 299 303 \includegraphics[width=\linewidth]{images/fir-mono-vs-fir-series-noise-fixe-jmf-light.pdf}
\caption{Comparison of the rejection capability between a series of FIR and a monolithic FIR 300 304 \caption{Comparison of the rejection capability between a series of FIR and a monolithic FIR
with a cutoff frequency set at half the Nyquist frequency.} 301 305 with a cutoff frequency set at half the Nyquist frequency.}
\label{compare-fir} 302 306 \label{compare-fir}
\end{figure} 303 307 \end{figure}
304 308
Fig. \ref{compare-fir} exhibits the 305 309 Fig. \ref{compare-fir} exhibits the
performance comparison between one solution and a monolithic FIR when selecting a cutoff 306 310 performance comparison between one solution and a monolithic FIR when selecting a cutoff
frequency of half the Nyquist frequency: a series of 5 FIR and a series of 10 FIR with the 307 311 frequency of half the Nyquist frequency: a series of 5 FIR and a series of 10 FIR with the
same space usage are provided as selected by the MILP solver. The FIR cascade provides improved 308 312 same space usage are provided as selected by the MILP solver. The FIR cascade provides improved
rejection than the monolithic FIR at the expense of a lower cutoff frequency which remains to 309 313 rejection than the monolithic FIR at the expense of a lower cutoff frequency which remains to
be tuned or compensated for. 310 314 be tuned or compensated for.
311 315
312 316
The resource occupation when synthesizing such FIR on a Xilinx FPGA is summarized as Tab. \ref{t1}. 313 317 The resource occupation when synthesizing such FIR on a Xilinx FPGA is summarized as Tab. \ref{t1}.
We have considered a set of resources representative of the hardware platform we work on, 314 318 We have considered a set of resources representative of the hardware platform we work on,
Avnet's Zedboard featuring a Xilinx XC7Z020-CLG484-1 Zynq System on Chip (SoC). The results reported in 315 319 Avnet's Zedboard featuring a Xilinx XC7Z020-CLG484-1 Zynq System on Chip (SoC). The results reported in
Tab. \ref{t1} emphasize that implementing the monolithic single FIR is impossible due to 316 320 Tab. \ref{t1} emphasize that implementing the monolithic single FIR is impossible due to
the insufficient hardware resources (exhausted LUT resources), while the FIR cascading 5 or 10 317 321 the insufficient hardware resources (exhausted LUT resources), while the FIR cascading 5 or 10
filters fit in the available resources. However, in all cases the DSP resources are fully 318 322 filters fit in the available resources. However, in all cases the DSP resources are fully
used: while the design can be synthesized using Xilinx proprietary Vivado 2016.2 software, 319 323 used: while the design can be synthesized using Xilinx proprietary Vivado 2016.2 software,
implementing the design fails due to the excessive resource usage preventing routing the signals 320 324 implementing the design fails due to the excessive resource usage preventing routing the signals
on the FPGA. Such results emphasize on the one hand the improvement prospect of the optimization 321 325 on the FPGA. Such results emphasize on the one hand the improvement prospect of the optimization
procedure by finding non-trivial solutions matching resource constraints, but on the other 322 326 procedure by finding non-trivial solutions matching resource constraints, but on the other
hand also illustrates the limitation of a model with an abstraction layer that does not account 323 327 hand also illustrates the limitation of a model with an abstraction layer that does not account
for the detailed architecture of the hardware. 324 328 for the detailed architecture of the hardware.
325 329
\begin{table}[h!tb] 326 330 \begin{table}[h!tb]
\caption{Resource occupation on a Xilinx Zynq-7000 series FPGA when synthesizing the FIR cascade 327 331 \caption{Resource occupation on a Xilinx Zynq-7000 series FPGA when synthesizing the FIR cascade
identified as optimal by the MILP solver within a finite resource criterion. The last line refers 328 332 identified as optimal by the MILP solver within a finite resource criterion. The last line refers
to available resources on a Zynq-7020 as found on the Zedboard.} 329 333 to available resources on a Zynq-7020 as found on the Zedboard.}
\begin{center} 330 334 \begin{center}
\begin{tabular}{|c|cccc|}\hline 331 335 \begin{tabular}{|c|cccc|}\hline
FIR & BlockRAM & LookUpTables & DSP & rejection (dB)\\\hline\hline 332 336 FIR & BlockRAM & LookUpTables & DSP & rejection (dB)\\\hline\hline
1 (monolithic) & 1 & 76183 & 220 & -162 \\ 333 337 1 (monolithic) & 1 & 76183 & 220 & -162 \\
5 & 5 & 18597 & 220 & -160 \\ 334 338 5 & 5 & 18597 & 220 & -160 \\
10 & 8 & 24729 & 220 & -161 \\\hline\hline 335 339 10 & 8 & 24729 & 220 & -161 \\\hline\hline
\textbf{Zynq 7020} & \textbf{420} & \textbf{53200} & \textbf{220} & \\\hline 336 340 \textbf{Zynq 7020} & \textbf{420} & \textbf{53200} & \textbf{220} & \\\hline
%\begin{tabular}{|c|ccccc|}\hline 337 341 %\begin{tabular}{|c|ccccc|}\hline
%FIR & BRAM36 & BRAM18 & LUT & DSP & rejection (dB)\\\hline\hline 338 342 %FIR & BRAM36 & BRAM18 & LUT & DSP & rejection (dB)\\\hline\hline
%1 (monolithic) & 1 & 0 & {\color{Red}76183} & 220 & -162 \\ 339 343 %1 (monolithic) & 1 & 0 & {\color{Red}76183} & 220 & -162 \\
%5 & 0 & 5 & {\color{Green}18597} & 220 & -160 \\ 340 344 %5 & 0 & 5 & {\color{Green}18597} & 220 & -160 \\
%10 & 0 & 8 & {\color{Green}24729} & 220 & -161 \\\hline\hline 341 345 %10 & 0 & 8 & {\color{Green}24729} & 220 & -161 \\\hline\hline
%\textbf{Zynq 7020} & \textbf{140} & \textbf{280} & \textbf{53200} & \textbf{220} & \\\hline 342 346 %\textbf{Zynq 7020} & \textbf{140} & \textbf{280} & \textbf{53200} & \textbf{220} & \\\hline
\end{tabular} 343 347 \end{tabular}
\end{center} 344 348 \end{center}
%\vspace{-0.7cm} 345 349 %\vspace{-0.7cm}
\label{t1} 346 350 \label{t1}
\end{table} 347 351 \end{table}
348 352
\subsection{Alternate criteria}\label{median} 349 353 \subsection{Alternate criteria}\label{median}
350 354
Fig. \ref{compare-fir} provides FIR solutions matching well the targeted transfer 351 355 Fig. \ref{compare-fir} provides FIR solutions matching well the targeted transfer
function, namely low ripple in the bandpass defined as the first 40\% of the frequency 352 356 function, namely low ripple in the bandpass defined as the first 40\% of the frequency
range and maximum rejection of 160~dB in the last 40\% stopband. We illustrate now, for 353 357 range and maximum rejection of 160~dB in the last 40\% stopband. We illustrate now, for
demonstrating the need to properly select the optimization criterion, two cases of poor 354 358 demonstrating the need to properly select the optimization criterion, two cases of poor
filter shapes obtained by selecting the mean value and median value of the rejection, 355 359 filter shapes obtained by selecting the mean value and median value of the rejection,
with no consideration for the ripples in the bandpass. The results of the optimizations, 356 360 with no consideration for the ripples in the bandpass. The results of the optimizations,
in these cases, are shown in Figs. \ref{compare-mean} and \ref{compare-median}. 357 361 in these cases, are shown in Figs. \ref{compare-mean} and \ref{compare-median}.
358 362
\begin{figure}[h!tb] 359 363 \begin{figure}[h!tb]
\includegraphics[width=\linewidth]{images/fir-mono-vs-fir-series-noise-fixe-mean-light.pdf} 360 364 \includegraphics[width=\linewidth]{images/fir-mono-vs-fir-series-noise-fixe-mean-light.pdf}
\caption{Comparison of the rejection capability between a series of FIR and a monolithic FIR 361 365 \caption{Comparison of the rejection capability between a series of FIR and a monolithic FIR
with a cutoff frequency set at half the Nyquist frequency.} 362 366 with a cutoff frequency set at half the Nyquist frequency.}
\label{compare-mean} 363 367 \label{compare-mean}
\end{figure} 364 368 \end{figure}
365 369
In the case of the mean value criterion (Fig. \ref{compare-mean}), the solution is not 366 370 In the case of the mean value criterion (Fig. \ref{compare-mean}), the solution is not
acceptable since the notch at the end of the transition band compensates for some unacceptable 367 371 acceptable since the notch at the end of the transition band compensates for some unacceptable
rise in the rejection close to the Nyquist frequency. Applying such a filter might yield excessive 368 372 rise in the rejection close to the Nyquist frequency. Applying such a filter might yield excessive
high frequency spurious components to be aliased at low frequency when decimating the signal. 369 373 high frequency spurious components to be aliased at low frequency when decimating the signal.
Similarly, the lack of criterion on the bandpass shape induces a shape with poor flatness and 370 374 Similarly, the lack of criterion on the bandpass shape induces a shape with poor flatness and
and slowly decaying transfer function starting to attenuate spectral components well before the 371 375 and slowly decaying transfer function starting to attenuate spectral components well before the
transition band starts. Such issues are partly aleviated by replacing a mean rejection value with 372 376 transition band starts. Such issues are partly aleviated by replacing a mean rejection value with
a median rejection value (Fig. \ref{compare-median}) but solutions remain unacceptable for 373 377 a median rejection value (Fig. \ref{compare-median}) but solutions remain unacceptable for
the reasons stated previously and much poorer than those found with the maximum rejection criterion 374 378 the reasons stated previously and much poorer than those found with the maximum rejection criterion
selected earlier (Fig. \ref{compare-fir}). 375 379 selected earlier (Fig. \ref{compare-fir}).
376 380
\begin{figure}[h!tb] 377 381 \begin{figure}[h!tb]
\includegraphics[width=\linewidth]{images/fir-mono-vs-fir-series-noise-fixe-median-light.pdf} 378 382 \includegraphics[width=\linewidth]{images/fir-mono-vs-fir-series-noise-fixe-median-light.pdf}
\caption{Comparison of the rejection capability between a series of FIR and a monolithic FIR 379 383 \caption{Comparison of the rejection capability between a series of FIR and a monolithic FIR
with a cutoff frequency set at half the Nyquist frequency.} 380 384 with a cutoff frequency set at half the Nyquist frequency.}
\label{compare-median} 381 385 \label{compare-median}
\end{figure} 382 386 \end{figure}
383 387
\section{Filter coefficient selection} 384 388 \section{Filter coefficient selection}
385 389
The coefficients of a single monolithic filter are computed as the impulse response 386 390 The coefficients of a single monolithic filter are computed as the impulse response
of the filter transfer function, and practically approximated by a multitude of methods 387 391 of the filter transfer function, and practically approximated by a multitude of methods
including least square optimization (Matlab's {\tt firls} function), Hamming or Kaiser windowing 388 392 including least square optimization (Matlab's {\tt firls} function), Hamming or Kaiser windowing
(Matlab's {\tt fir1} function). 389 393 (Matlab's {\tt fir1} function).
390 394
\begin{figure}[h!tb] 391 395 \begin{figure}[h!tb]
\includegraphics[width=\linewidth]{images/fir1-vs-firls} 392 396 \includegraphics[width=\linewidth]{images/fir1-vs-firls}
\caption{Evolution of the rejection capability of least-square optimized filters and Hamming 393 397 \caption{Evolution of the rejection capability of least-square optimized filters and Hamming
FIR filters as a function of the number of coefficients, for floating point numbers and 8-bit 394 398 FIR filters as a function of the number of coefficients, for floating point numbers and 8-bit
encoded integers.} 395 399 encoded integers.}
\label{2} 396 400 \label{2}
\end{figure} 397 401 \end{figure}
398 402
Cascading filters opens a new optimization opportunity by 399 403 Cascading filters opens a new optimization opportunity by
selecting various coefficient sets depending on the number of coefficients. Fig. \ref{2} 400 404 selecting various coefficient sets depending on the number of coefficients. Fig. \ref{2}
illustrates that for a number of coefficients ranging from 8 to 47, {\tt fir1} provides a better 401 405 illustrates that for a number of coefficients ranging from 8 to 47, {\tt fir1} provides a better
rejection than {\tt firls}: since the linear solver increases the number of coefficients along 402 406 rejection than {\tt firls}: since the linear solver increases the number of coefficients along
the processing chain, the type of selected filter also changes depending on the number of coefficients 403 407 the processing chain, the type of selected filter also changes depending on the number of coefficients
and evolves along the processing chain. 404 408 and evolves along the processing chain.
405 409
\section{Conclusion} 406 410 \section{Conclusion}
407 411
We address the optimization problem of designing a low-pass filter chain in a Field Programmable Gate 408 412 We address the optimization problem of designing a low-pass filter chain in a Field Programmable Gate
Array for improved noise rejection within constrained resource occupation, as needed for 409 413 Array for improved noise rejection within constrained resource occupation, as needed for
real time processing of radiofrequency signal when characterizing spectral phase noise 410 414 real time processing of radiofrequency signal when characterizing spectral phase noise
characteristics of stable oscillators. The flexibility of the digital approach makes the result 411 415 characteristics of stable oscillators. The flexibility of the digital approach makes the result
best suited for closing the loop and using the measurement output in a feedback loop for 412 416 best suited for closing the loop and using the measurement output in a feedback loop for
controlling clocks, e.g. in a quartz-stabilized high performance clock whose long term behavior 413 417 controlling clocks, e.g. in a quartz-stabilized high performance clock whose long term behavior
is controlled by non-piezoelectric resonator (sapphire resonator, microwave or optical 414 418 is controlled by non-piezoelectric resonator (sapphire resonator, microwave or optical
atomic transition). 415 419 atomic transition).
416 420
\section*{Acknowledgement} 417 421 \section*{Acknowledgement}
418 422
This work is supported by the ANR Programme d'Investissement d'Avenir in 419 423 This work is supported by the ANR Programme d'Investissement d'Avenir in
progress at the Time and Frequency Departments of the FEMTO-ST Institute 420 424 progress at the Time and Frequency Departments of the FEMTO-ST Institute
(Oscillator IMP, First-TF and Refimeve+), and by R\'egion de Franche-Comt\'e. 421 425 (Oscillator IMP, First-TF and Refimeve+), and by R\'egion de Franche-Comt\'e.
The authors would like to thank E. Rubiola, F. Vernotte, G. Cabodevila for support and 422 426 The authors would like to thank E. Rubiola, F. Vernotte, G. Cabodevila for support and
fruitful discussions. 423 427 fruitful discussions.
424 428
\bibliographystyle{IEEEtran} 425 429 \bibliographystyle{IEEEtran}
\balance 426 430 \balance
\bibliography{references,biblio} 427 431 \bibliography{references,biblio}
\end{document} 428 432 \end{document}
429 433
\section{Contexte d'ordonnancement} 430 434 \section{Contexte d'ordonnancement}
Dans cette partie, nous donnerons des d\'efinitions de termes rattach\'es au domaine de l'ordonnancement 431 435 Dans cette partie, nous donnerons des d\'efinitions de termes rattach\'es au domaine de l'ordonnancement
et nous verrons que le sujet trait\'e se rapproche beaucoup d'un problème d'ordonnancement. De ce fait 432 436 et nous verrons que le sujet trait\'e se rapproche beaucoup d'un problème d'ordonnancement. De ce fait
nous pourrons aller plus loin que les travaux vus pr\'ec\'edemment et nous tenterons des approches d'ordonnancement 433 437 nous pourrons aller plus loin que les travaux vus pr\'ec\'edemment et nous tenterons des approches d'ordonnancement
et d'optimisation. 434 438 et d'optimisation.
435 439
\subsection{D\'efinition du vocabulaire} 436 440 \subsection{D\'efinition du vocabulaire}
Avant tout, il faut d\'efinir ce qu'est un problème d'optimisation. Il y a deux d\'efinitions 437 441 Avant tout, il faut d\'efinir ce qu'est un problème d'optimisation. Il y a deux d\'efinitions
importantes à donner. La première est propos\'ee par Legrand et Robert dans leur livre \cite{def1-ordo} : 438 442 importantes à donner. La première est propos\'ee par Legrand et Robert dans leur livre \cite{def1-ordo} :
\begin{definition} 439 443 \begin{definition}
\label{def-ordo1} 440 444 \label{def-ordo1}
Un ordonnancement d'un système de t\^aches $G\ =\ (V,\ E,\ w)$ est une fonction $\sigma$ : 441 445 Un ordonnancement d'un système de t\^aches $G\ =\ (V,\ E,\ w)$ est une fonction $\sigma$ :
$V \rightarrow \mathbb{N}$ telle que $\sigma(u) + w(u) \leq \sigma(v)$ pour toute arête $(u,\ v) \in E$. 442 446 $V \rightarrow \mathbb{N}$ telle que $\sigma(u) + w(u) \leq \sigma(v)$ pour toute arête $(u,\ v) \in E$.
\end{definition} 443 447 \end{definition}
444 448
Dit plus simplement, l'ensemble $V$ repr\'esente les t\^aches à ex\'ecuter, l'ensemble $E$ repr\'esente les d\'ependances 445 449 Dit plus simplement, l'ensemble $V$ repr\'esente les t\^aches à ex\'ecuter, l'ensemble $E$ repr\'esente les d\'ependances
des t\^aches et $w$ les temps d'ex\'ecution de la t\^ache. La fonction $\sigma$ donne donc l'heure de d\'ebut de 446 450 des t\^aches et $w$ les temps d'ex\'ecution de la t\^ache. La fonction $\sigma$ donne donc l'heure de d\'ebut de
chacune des t\^aches. La d\'efinition dit que si une t\^ache $v$ d\'epend d'une t\^ache $u$ alors 447 451 chacune des t\^aches. La d\'efinition dit que si une t\^ache $v$ d\'epend d'une t\^ache $u$ alors
la date de d\'ebut de $v$ sera plus grande ou \'egale au d\'ebut de l'ex\'ecution de la t\^ache $u$ plus son 448 452 la date de d\'ebut de $v$ sera plus grande ou \'egale au d\'ebut de l'ex\'ecution de la t\^ache $u$ plus son
temps d'ex\'ecution. 449 453 temps d'ex\'ecution.
450 454
Une autre d\'efinition importante qui est propos\'ee par Leung et al. \cite{def2-ordo} est : 451 455 Une autre d\'efinition importante qui est propos\'ee par Leung et al. \cite{def2-ordo} est :
\begin{definition} 452 456 \begin{definition}
\label{def-ordo2} 453 457 \label{def-ordo2}
L'ordonnancement traite de l'allocation de ressources rares à des activit\'es avec 454 458 L'ordonnancement traite de l'allocation de ressources rares à des activit\'es avec
l'objectif d'optimiser un ou plusieurs critères de performance. 455 459 l'objectif d'optimiser un ou plusieurs critères de performance.
\end{definition} 456 460 \end{definition}
457 461
Cette d\'efinition est plus g\'en\'erique mais elle nous int\'eresse d'avantage que la d\'efinition \ref{def-ordo1}. 458 462 Cette d\'efinition est plus g\'en\'erique mais elle nous int\'eresse d'avantage que la d\'efinition \ref{def-ordo1}.
En effet, la partie qui nous int\'eresse dans cette première d\'efinition est le respect de la pr\'ec\'edance des t\^aches. 459 463 En effet, la partie qui nous int\'eresse dans cette première d\'efinition est le respect de la pr\'ec\'edance des t\^aches.
Dans les faits les dates de d\'ebut ne nous int\'eressent pas r\'eellement. 460 464 Dans les faits les dates de d\'ebut ne nous int\'eressent pas r\'eellement.
461 465
En revanche la d\'efinition \ref{def-ordo2} sera au c\oe{}ur du projet. Pour se convaincre de cela, 462 466 En revanche la d\'efinition \ref{def-ordo2} sera au c\oe{}ur du projet. Pour se convaincre de cela,
il nous faut d'abord d\'efinir quel est le type de problème d'ordonnancement qu'on traite et quelles 463 467 il nous faut d'abord d\'efinir quel est le type de problème d'ordonnancement qu'on traite et quelles
sont les m\'ethodes qu'on peut appliquer. 464 468 sont les m\'ethodes qu'on peut appliquer.
465 469
Les problèmes d'ordonnancement peuvent être class\'es en diff\'erentes cat\'egories : 466 470 Les problèmes d'ordonnancement peuvent être class\'es en diff\'erentes cat\'egories :
\begin{itemize} 467 471 \begin{itemize}
\item T\^aches ind\'ependantes : dans cette cat\'egorie de problèmes, les t\^aches sont complètement ind\'ependantes 468 472 \item T\^aches ind\'ependantes : dans cette cat\'egorie de problèmes, les t\^aches sont complètement ind\'ependantes
les unes des autres. Dans notre cas, ce n'est pas le plus adapt\'e. 469 473 les unes des autres. Dans notre cas, ce n'est pas le plus adapt\'e.
\item Graphe de t\^aches : la d\'efinition \ref{def-ordo1} d\'ecrit cette cat\'egorie. La plupart du temps, 470 474 \item Graphe de t\^aches : la d\'efinition \ref{def-ordo1} d\'ecrit cette cat\'egorie. La plupart du temps,
les t\^aches sont repr\'esent\'ees par une DAG. Cette cat\'egorie est très proche de notre cas puisque nous devons \'egalement ex\'ecuter 471 475 les t\^aches sont repr\'esent\'ees par une DAG. Cette cat\'egorie est très proche de notre cas puisque nous devons \'egalement ex\'ecuter
des t\^aches qui ont un certain nombre de d\'ependances. On pourra même dire que dans certain cas, 472 476 des t\^aches qui ont un certain nombre de d\'ependances. On pourra même dire que dans certain cas,
on a des anti-arbres, c'est à dire que nous avons une multitude de t\^aches d'entr\'ees qui convergent vers une 473 477 on a des anti-arbres, c'est à dire que nous avons une multitude de t\^aches d'entr\'ees qui convergent vers une
t\^ache de fin. 474 478 t\^ache de fin.
\item Workflow : cette cat\'egorie est une sous cat\'egorie des graphes de t\^aches dans le sens où 475 479 \item Workflow : cette cat\'egorie est une sous cat\'egorie des graphes de t\^aches dans le sens où
il s'agit d'un graphe de t\^aches r\'ep\'et\'e de nombreuses de fois. C'est exactement ce type de problème 476 480 il s'agit d'un graphe de t\^aches r\'ep\'et\'e de nombreuses de fois. C'est exactement ce type de problème
que nous traitons ici. 477 481 que nous traitons ici.
\end{itemize} 478 482 \end{itemize}
479 483
Bien entendu, cette liste n'est pas exhaustive et il existe de nombreuses autres classifications et sous-classifications 480 484 Bien entendu, cette liste n'est pas exhaustive et il existe de nombreuses autres classifications et sous-classifications
de ces problèmes. Nous n'avons parl\'e ici que des cat\'egories les plus communes. 481 485 de ces problèmes. Nous n'avons parl\'e ici que des cat\'egories les plus communes.
482 486
Un autre point à d\'efinir, est le critère d'optimisation. Il y a là encore un grand nombre de 483 487 Un autre point à d\'efinir, est le critère d'optimisation. Il y a là encore un grand nombre de
critères possibles. Nous allons donc parler des principaux : 484 488 critères possibles. Nous allons donc parler des principaux :
\begin{itemize} 485 489 \begin{itemize}
\item Temps de compl\'etion total (ou Makespan en anglais) : ce critère est l'un des critères d'optimisation 486 490 \item Temps de compl\'etion total (ou Makespan en anglais) : ce critère est l'un des critères d'optimisation
les plus courant. Il s'agit donc de minimiser la date de fin de la dernière t\^ache de l'ensemble des 487 491 les plus courant. Il s'agit donc de minimiser la date de fin de la dernière t\^ache de l'ensemble des
t\^aches à ex\'ecuter. L'enjeu de cette optimisation est donc de trouver l'ordonnancement optimal permettant 488 492 t\^aches à ex\'ecuter. L'enjeu de cette optimisation est donc de trouver l'ordonnancement optimal permettant
la fin d'ex\'ecution au plus tôt. 489 493 la fin d'ex\'ecution au plus tôt.
\item Somme des temps d'ex\'ecution (Flowtime en anglais) : il s'agit de faire la somme des temps d'ex\'ecution de toutes les t\^aches 490 494 \item Somme des temps d'ex\'ecution (Flowtime en anglais) : il s'agit de faire la somme des temps d'ex\'ecution de toutes les t\^aches
et d'optimiser ce r\'esultat. 491 495 et d'optimiser ce r\'esultat.
\item Le d\'ebit : ce critère quant à lui, vise à augmenter au maximum le d\'ebit de traitement des donn\'ees. 492 496 \item Le d\'ebit : ce critère quant à lui, vise à augmenter au maximum le d\'ebit de traitement des donn\'ees.
\end{itemize} 493 497 \end{itemize}
494 498
En plus de cela, on peut avoir besoin de plusieurs critères d'optimisation. Il s'agit dans ce cas d'une optimisation 495 499 En plus de cela, on peut avoir besoin de plusieurs critères d'optimisation. Il s'agit dans ce cas d'une optimisation
multi-critères. Bien entendu, cela complexifie d'autant plus le problème car la solution la plus optimale pour un 496 500 multi-critères. Bien entendu, cela complexifie d'autant plus le problème car la solution la plus optimale pour un
des critères peut être très mauvaise pour un autre critère. De ce cas, il s'agira de trouver une solution qui permet 497 501 des critères peut être très mauvaise pour un autre critère. De ce cas, il s'agira de trouver une solution qui permet
de faire le meilleur compromis entre tous les critères. 498 502 de faire le meilleur compromis entre tous les critères.
499 503
\subsection{Formalisation du problème} 500 504 \subsection{Formalisation du problème}
\label{formalisation} 501 505 \label{formalisation}
Maintenant que nous avons donn\'e le vocabulaire li\'e à l'ordonnancement, nous allons pouvoir essayer caract\'eriser 502 506 Maintenant que nous avons donn\'e le vocabulaire li\'e à l'ordonnancement, nous allons pouvoir essayer caract\'eriser
formellement notre problème. En effet, nous allons reprendre les contraintes \'enonc\'ees dans la sections \ref{def-contraintes} 503 507 formellement notre problème. En effet, nous allons reprendre les contraintes \'enonc\'ees dans la sections \ref{def-contraintes}
et nous essayerons de les formaliser le plus finement possible. 504 508 et nous essayerons de les formaliser le plus finement possible.
505 509
Comme nous l'avons dit, une t\^ache est un bloc de traitement. Chaque t\^ache $i$ dispose d'un ensemble de paramètres 506 510 Comme nous l'avons dit, une t\^ache est un bloc de traitement. Chaque t\^ache $i$ dispose d'un ensemble de paramètres
que nous nommerons $\mathcal{P}_{i}$. Cet ensemble $\mathcal{P}_i$ est propre à chaque t\^ache et il variera d'une 507 511 que nous nommerons $\mathcal{P}_{i}$. Cet ensemble $\mathcal{P}_i$ est propre à chaque t\^ache et il variera d'une
t\^ache à l'autre. Nous reviendrons plus tard sur les paramètres qui peuvent composer cet ensemble. 508 512 t\^ache à l'autre. Nous reviendrons plus tard sur les paramètres qui peuvent composer cet ensemble.
509 513
Outre cet ensemble $\mathcal{P}_i$, chaque t\^ache dispose de paramètres communs : 510 514 Outre cet ensemble $\mathcal{P}_i$, chaque t\^ache dispose de paramètres communs :
\begin{itemize} 511 515 \begin{itemize}
\item Dur\'ee de la t\^ache : Comme nous l'avons dit auparavant, dans le cadre d'un FPGA le temps est compt\'e en nombre de coup d'horloge. 512 516 \item Dur\'ee de la t\^ache : Comme nous l'avons dit auparavant, dans le cadre d'un FPGA le temps est compt\'e en nombre de coup d'horloge.
En outre, les blocs sont toujours sollicit\'es, certains même sont capables de lire et de renvoyer une r\'esultat à chaque coups d'horloge. 513 517 En outre, les blocs sont toujours sollicit\'es, certains même sont capables de lire et de renvoyer une r\'esultat à chaque coups d'horloge.
Donc la dur\'ee d'une t\^ache ne peut être le laps de temps entre l'entr\'ee d'une donn\'ee et la sortie d'une autre. Nous d\'efinirons la 514 518 Donc la dur\'ee d'une t\^ache ne peut être le laps de temps entre l'entr\'ee d'une donn\'ee et la sortie d'une autre. Nous d\'efinirons la
dur\'ee comme le temps de traitement d'une donn\'ee, c'est à dire la diff\'erence de temps entre la date de sortie d'une donn\'ee 515 519 dur\'ee comme le temps de traitement d'une donn\'ee, c'est à dire la diff\'erence de temps entre la date de sortie d'une donn\'ee
et de sa date d'entr\'ee. Nous nommerons cette dur\'ee $\delta_i$. % Je devrais la nomm\'ee w comme dans la def2 516 520 et de sa date d'entr\'ee. Nous nommerons cette dur\'ee $\delta_i$. % Je devrais la nomm\'ee w comme dans la def2
\item La pr\'ecision : La pr\'ecision d'une donn\'ee est le nombre de bits significatifs qu'elle compte. En effet, au fil des traitements 517 521 \item La pr\'ecision : La pr\'ecision d'une donn\'ee est le nombre de bits significatifs qu'elle compte. En effet, au fil des traitements
les pr\'ecisions peuvent varier. On nomme donc la pr\'ecision d'entr\'ee d'une t\^ache $i$ comme $\pi_i^-$ et la pr\'ecision en sortie $\pi_i^+$. 518 522 les pr\'ecisions peuvent varier. On nomme donc la pr\'ecision d'entr\'ee d'une t\^ache $i$ comme $\pi_i^-$ et la pr\'ecision en sortie $\pi_i^+$.
\item La fr\'equence du flux en entr\'ee (ou sortie) : Cette fr\'equence repr\'esente la fr\'equence des donn\'ees qui arrivent (resp. sortent). 519 523 \item La fr\'equence du flux en entr\'ee (ou sortie) : Cette fr\'equence repr\'esente la fr\'equence des donn\'ees qui arrivent (resp. sortent).
Selon les t\^aches, les fr\'equences varieront. En effet, certains blocs ralentissent le flux c'est pourquoi on distingue la fr\'equence du 520 524 Selon les t\^aches, les fr\'equences varieront. En effet, certains blocs ralentissent le flux c'est pourquoi on distingue la fr\'equence du
flux en entr\'ee et la fr\'equence en sortie. Nous nommerons donc la fr\'equence du flux en entr\'ee $f_i^-$ et la fr\'equence en sortie $f_i^+$. 521 525 flux en entr\'ee et la fr\'equence en sortie. Nous nommerons donc la fr\'equence du flux en entr\'ee $f_i^-$ et la fr\'equence en sortie $f_i^+$.
\item La quantit\'e de donn\'ees en entr\'ee (ou en sortie) : Il s'agit de la quantit\'e de donn\'ees que le bloc s'attend à traiter (resp. 522 526 \item La quantit\'e de donn\'ees en entr\'ee (ou en sortie) : Il s'agit de la quantit\'e de donn\'ees que le bloc s'attend à traiter (resp.
est capable de produire). Les t\^aches peuvent avoir à traiter des gros volumes de donn\'ees et n'en ressortir qu'une partie. Cette 523 527 est capable de produire). Les t\^aches peuvent avoir à traiter des gros volumes de donn\'ees et n'en ressortir qu'une partie. Cette
fois encore, il nous faut donc diff\'erencier l'entr\'ee et la sortie. Nous nommerons donc la quantit\'e de donn\'ees entrantes $q_i^-$ 524 528 fois encore, il nous faut donc diff\'erencier l'entr\'ee et la sortie. Nous nommerons donc la quantit\'e de donn\'ees entrantes $q_i^-$
et la quantit\'e de donn\'ees sortantes $q_i^+$ pour une t\^ache $i$. 525 529 et la quantit\'e de donn\'ees sortantes $q_i^+$ pour une t\^ache $i$.
\item Le d\'ebit d'entr\'ee (ou de sortie) : Ce paramètre correspond au d\'ebit de donn\'ees que la t\^ache est capable de traiter ou qu'elle 526 530 \item Le d\'ebit d'entr\'ee (ou de sortie) : Ce paramètre correspond au d\'ebit de donn\'ees que la t\^ache est capable de traiter ou qu'elle
fournit en sortie. Il s'agit simplement de l'expression des deux pr\'ec\'edents paramètres. Nous d\'efinirons donc la d\'ebit entrant de la 527 531 fournit en sortie. Il s'agit simplement de l'expression des deux pr\'ec\'edents paramètres. Nous d\'efinirons donc la d\'ebit entrant de la
t\^ache $i$ comme $d_i^-\ =\ q_i^-\ *\ f_i^-$ et le d\'ebit sortant comme $d_i^+\ =\ q_i^+\ *\ f_i^+$. 528 532 t\^ache $i$ comme $d_i^-\ =\ q_i^-\ *\ f_i^-$ et le d\'ebit sortant comme $d_i^+\ =\ q_i^+\ *\ f_i^+$.
\item La taille de la t\^ache : La taille dans les FPGA \'etant limit\'ee, ce paramètre exprime donc la place qu'occupe la t\^ache au sein du bloc. 529 533 \item La taille de la t\^ache : La taille dans les FPGA \'etant limit\'ee, ce paramètre exprime donc la place qu'occupe la t\^ache au sein du bloc.
Nous nommerons $\mathcal{A}_i$ cette taille. 530 534 Nous nommerons $\mathcal{A}_i$ cette taille.
\item Les pr\'ed\'ecesseurs et successeurs d'une t\^ache : cela nous permet de connaître les t\^aches requises pour pouvoir traiter 531 535 \item Les pr\'ed\'ecesseurs et successeurs d'une t\^ache : cela nous permet de connaître les t\^aches requises pour pouvoir traiter
la t\^ache $i$ ainsi que les t\^aches qui en d\'ependent. Ces ensemble sont not\'es $\Gamma _i ^-$ et $ \Gamma _i ^+$ \\ 532 536 la t\^ache $i$ ainsi que les t\^aches qui en d\'ependent. Ces ensemble sont not\'es $\Gamma _i ^-$ et $ \Gamma _i ^+$ \\
%TODO Est-ce vraiment un paramètre ? 533 537 %TODO Est-ce vraiment un paramètre ?
\end{itemize} 534 538 \end{itemize}
535 539
Ces diff\'erents paramètres communs sont fortement li\'es aux \'el\'ements de $\mathcal{P}_i$. Voici quelques exemples de relations 536 540 Ces diff\'erents paramètres communs sont fortement li\'es aux \'el\'ements de $\mathcal{P}_i$. Voici quelques exemples de relations
que nous avons identifi\'ees : 537 541 que nous avons identifi\'ees :
\begin{itemize} 538 542 \begin{itemize}
\item $ \delta _i ^+ \ = \ \mathcal{F}_{\delta}(\pi_i^-,\ \pi_i^+,\ d_i^-,\ d_i^+,\ \mathcal{P}_i) $ donne le temps d'ex\'ecution 539 543 \item $ \delta _i ^+ \ = \ \mathcal{F}_{\delta}(\pi_i^-,\ \pi_i^+,\ d_i^-,\ d_i^+,\ \mathcal{P}_i) $ donne le temps d'ex\'ecution
de la t\^ache en fonction de la pr\'ecision voulue, du d\'ebit et des paramètres internes. 540 544 de la t\^ache en fonction de la pr\'ecision voulue, du d\'ebit et des paramètres internes.
\item $ \pi _i ^+ \ = \ \mathcal{F}_{p}(\pi_i^-,\ \mathcal{P}_i) $, la fonction $F_p$ donne la pr\'ecision en sortie selon la pr\'ecision de d\'epart 541 545 \item $ \pi _i ^+ \ = \ \mathcal{F}_{p}(\pi_i^-,\ \mathcal{P}_i) $, la fonction $F_p$ donne la pr\'ecision en sortie selon la pr\'ecision de d\'epart
et les paramètres internes de la t\^ache. 542 546 et les paramètres internes de la t\^ache.
\item $d_i^+\ =\ \mathcal{F}_d(d_i^-, \mathcal{P}_i)$, la fonction $F_d$ donne le d\'ebit sortant de la t\^ache en fonction du d\'ebit 543 547 \item $d_i^+\ =\ \mathcal{F}_d(d_i^-, \mathcal{P}_i)$, la fonction $F_d$ donne le d\'ebit sortant de la t\^ache en fonction du d\'ebit
sortant et des variables internes de la t\^ache. 544 548 sortant et des variables internes de la t\^ache.
\item $A_i^+\ =\ \mathcal{F}_A(\pi_i^-,\ \pi_i^+,\ d_i^-,\ d_i^+, \mathcal{P}_i)$ 545 549 \item $A_i^+\ =\ \mathcal{F}_A(\pi_i^-,\ \pi_i^+,\ d_i^-,\ d_i^+, \mathcal{P}_i)$
\end{itemize} 546 550 \end{itemize}
Pour le moment, nous ne sommes pas capables de donner une d\'efinition g\'en\'erale de ces fonctions. Mais en revanche, 547 551 Pour le moment, nous ne sommes pas capables de donner une d\'efinition g\'en\'erale de ces fonctions. Mais en revanche,
sur quelques exemples simples (cf. \ref{def-contraintes}), nous parvenons à donner une \'evaluation de ces fonctions. 548 552 sur quelques exemples simples (cf. \ref{def-contraintes}), nous parvenons à donner une \'evaluation de ces fonctions.
549 553
Maintenant que nous avons donn\'e toutes les notations utiles, nous allons \'enoncer des contraintes relatives à notre problème. Soit 550 554 Maintenant que nous avons donn\'e toutes les notations utiles, nous allons \'enoncer des contraintes relatives à notre problème. Soit
un DGA $G(V,\ E)$, on a pour toutes arêtes $(i, j)\ \in\ E$ les in\'equations suivantes : 551 555 un DGA $G(V,\ E)$, on a pour toutes arêtes $(i, j)\ \in\ E$ les in\'equations suivantes :
552 556
\paragraph{Contrainte de pr\'ecision :} 553 557 \paragraph{Contrainte de pr\'ecision :}
Cette in\'equation traduit la contrainte de pr\'ecision d'une t\^ache à l'autre : 554 558 Cette in\'equation traduit la contrainte de pr\'ecision d'une t\^ache à l'autre :
\begin{align*} 555 559 \begin{align*}
\pi _i ^+ \geq \pi _j ^- 556 560 \pi _i ^+ \geq \pi _j ^-
\end{align*} 557 561 \end{align*}
558 562
\paragraph{Contrainte de d\'ebit :} 559 563 \paragraph{Contrainte de d\'ebit :}
Cette in\'equation traduit la contrainte de d\'ebit d'une t\^ache à l'autre : 560 564 Cette in\'equation traduit la contrainte de d\'ebit d'une t\^ache à l'autre :
\begin{align*} 561 565 \begin{align*}
d _i ^+ = q _j ^- * (f_i + (1 / s_j) ) & \text{ où } s_j \text{ est une valeur positive de temporisation de la t\^ache} 562 566 d _i ^+ = q _j ^- * (f_i + (1 / s_j) ) & \text{ où } s_j \text{ est une valeur positive de temporisation de la t\^ache}
\end{align*} 563 567 \end{align*}
564 568
\paragraph{Contrainte de synchronisation :} 565 569 \paragraph{Contrainte de synchronisation :}
Il s'agit de la contrainte qui impose que si à un moment du traitement, le DAG se s\'epare en plusieurs branches parallèles 566 570 Il s'agit de la contrainte qui impose que si à un moment du traitement, le DAG se s\'epare en plusieurs branches parallèles
et qu'elles se rejoignent plus tard, la somme des latences sur chacune des branches soit la même. 567 571 et qu'elles se rejoignent plus tard, la somme des latences sur chacune des branches soit la même.
Plus formellement, s'il existe plusieurs chemins disjoints, partant de la t\^ache $s$ et allant à la t\^ache de $f$ alors : 568 572 Plus formellement, s'il existe plusieurs chemins disjoints, partant de la t\^ache $s$ et allant à la t\^ache de $f$ alors :
\begin{align*} 569 573 \begin{align*}
\forall \text{ chemin } \mathcal{C}1(s, .., f), 570 574 \forall \text{ chemin } \mathcal{C}1(s, .., f),
\forall \text{ chemin } \mathcal{C}2(s, .., f) 571 575 \forall \text{ chemin } \mathcal{C}2(s, .., f)
\text{ tel que } \mathcal{C}1 \neq \mathcal{C}2 572 576 \text{ tel que } \mathcal{C}1 \neq \mathcal{C}2
\Rightarrow 573 577 \Rightarrow
\sum _{i} ^{i \in \mathcal{C}1} \delta_i = \sum _{i} ^{i \in \mathcal{C}2} \delta_i 574 578 \sum _{i} ^{i \in \mathcal{C}1} \delta_i = \sum _{i} ^{i \in \mathcal{C}2} \delta_i
\end{align*} 575 579 \end{align*}
576 580
\paragraph{Contrainte de place :} 577 581 \paragraph{Contrainte de place :}
Cette in\'equation traduit la contrainte de place dans le FPGA. La taille max de la puce FPGA est nomm\'e $\mathcal{A}_{FPGA}$ : 578 582 Cette in\'equation traduit la contrainte de place dans le FPGA. La taille max de la puce FPGA est nomm\'e $\mathcal{A}_{FPGA}$ :
\begin{align*} 579 583 \begin{align*}
\sum ^{\text{t\^ache } i} \mathcal{A}_i \leq \mathcal{A}_{FPGA} 580 584 \sum ^{\text{t\^ache } i} \mathcal{A}_i \leq \mathcal{A}_{FPGA}
\end{align*} 581 585 \end{align*}
582 586
\subsection{Exemples de mod\'elisation} 583 587 \subsection{Exemples de mod\'elisation}
\label{exemples-modeles} 584 588 \label{exemples-modeles}
Nous allons maintenant prendre quelques blocs de traitement simples afin d'illustrer au mieux notre modèle. 585 589 Nous allons maintenant prendre quelques blocs de traitement simples afin d'illustrer au mieux notre modèle.
Pour tous nos exemple, nous prendrons un d\'ebit en entr\'ee de 200 Mo/s avec une pr\'ecision de 16 bit. 586 590 Pour tous nos exemple, nous prendrons un d\'ebit en entr\'ee de 200 Mo/s avec une pr\'ecision de 16 bit.
587 591
Prenons tout d'abord l'exemple d'un bloc de d\'ecimation. Le but de ce bloc est de ralentir le flux en ne gardant 588 592 Prenons tout d'abord l'exemple d'un bloc de d\'ecimation. Le but de ce bloc est de ralentir le flux en ne gardant