Mourad Fakhfakh Esteban Tlelo-Cuautle Patrick Siarry Editors Computational Intelligence in Analog and Mixed-Signal (AMS) and Radio-Frequency (RF) Circuit Design Computational Intelligence in Analog and Mixed-Signal (AMS) and Radio-Frequency (RF) Circuit Design Mourad Fakhfakh Esteban Tlelo-Cuautle (cid:129) Patrick Siarry Editors Computational Intelligence in Analog and Mixed-Signal (AMS) and Radio-Frequency (RF) Circuit Design 123 Editors MouradFakhfakh Patrick Siarry Department ofElectronics Laboratory LiSSi(EA 3956) ENET’Com, University of Sfax UniversitéParis-Est Créteil Sfax Vitry-sur-Seine Tunisia France Esteban Tlelo-Cuautle Department ofElectronics INAOE Tonantzintla, Puebla Mexico ISBN978-3-319-19871-2 ISBN978-3-319-19872-9 (eBook) DOI 10.1007/978-3-319-19872-9 LibraryofCongressControlNumber:2015942631 SpringerChamHeidelbergNewYorkDordrechtLondon ©SpringerInternationalPublishingSwitzerland2015 Thisworkissubjecttocopyright.AllrightsarereservedbythePublisher,whetherthewholeorpart of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission orinformationstorageandretrieval,electronicadaptation,computersoftware,orbysimilarordissimilar methodologynowknownorhereafterdeveloped. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publicationdoesnotimply,evenintheabsenceofaspecificstatement,thatsuchnamesareexemptfrom therelevantprotectivelawsandregulationsandthereforefreeforgeneraluse. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authorsortheeditorsgiveawarranty,expressorimplied,withrespecttothematerialcontainedhereinor foranyerrorsoromissionsthatmayhavebeenmade. Printedonacid-freepaper SpringerInternationalPublishingAGSwitzerlandispartofSpringerScience+BusinessMedia (www.springer.com) Preface Computational intelligence has been an astounding success in the engineering domain, particularly in electronic design. Over the last two decades, improved techniqueshaveraisedtheproductivityofdesignerstoaremarkabledegree.Indeed, intheareasofdigital,analog,radio-frequency,andmixed-signalengineering,there is a focused effort on trying to automate all levels of the design flow of electronic circuits, a field where it was long assumed that progress demanded a skilled designer’s expertise. Thus, new computational-based modeling, synthesis and design methodologies, and applications of optimization algorithms have been proposed for assisting the designer’s task. This book offers the reader a collection of recent advances in computational intelligence—algorithms,designmethodologies,andsynthesistechniques—applied tothedesignofintegratedcircuitsandsystems.Ithighlightsnewbiasingandsizing approaches and optimization techniques and their application to the design of high-performance digital, VLSI, radio-frequency, and mixed-signal circuits and systems. As editors, we invited experts from related design disciplines to contribute overviews of their particular fields, and we grouped these into the following: (cid:129) Volume1,“ComputationalIntelligenceinAnalogandMixed-Signal(AMS)and Radio-Frequency (RF) Circuit Design,” contains 17 chapters, divided into two parts: “Analog and Mixed-Signal Applications” (Chaps. 1–8) and “Radio- Frequency Design” (Chaps. 9–17). (cid:129) Volume 2, “Computational Intelligence in Digital and Network Designs and Applications,” contains 12 chapters, divided into three parts: “Digital Circuit Design”(Chaps.1–6),“NetworkOptimization”(Chaps.7–8),and“Applications” (Chaps.9–12). Here, we present detailed descriptions of the chapters in both volumes. v vi Preface Volume 1—Computational Intelligence in Analog and Mixed-Signal (AMS) and Radio-Frequency (RF) Circuit Design Part I—Analog and Mixed-Signal Applications Chapter 1, “I-Flows: A Novel Approach to Computational Intelligence for Analog Circuit Design Automation Through Symbolic Data Mining and Knowledge- IntensiveReasoning,”waswrittenbyFanshuJiao,SergioMontano,CristianFerent, and Alex Doboli. It presents an overview of the authors’ ongoing work toward devisinganewapproachtoanalogcircuitsynthesis.Theapproachcomputationally implements some of the facets of knowledge-intensive reasoning that humans perform when tackling new design problems. This is achieved through a synthesis flow that mimics reasoning using a domain-specific knowledge structure with two components: an associative part and a causal reasoning part. The associative part groups known circuit schematics into abstractions based on the similarities and differencesoftheirstructuralfeatures.Thecausalreasoningcomponentincludesthe starting ideas as well as the design sequences that create the existing circuits. Chapter 2, “Automatic Synthesis of Analog Integrated Circuits Including Efficient Yield Optimization,” was written by Lucas C. Severo, Fabio N. Kepler, and Alessandro G. Girardi. Here, the authors show the main aspects and implica- tions of automatic sizing, including yield. Different strategies for accelerating performance estimation and design space search are addressed. The analog sizing problemisconvertedintoanonlinearoptimizationproblem,andthedesignspaceis explored using metaheuristics based on genetic algorithms. Circuit performance is estimated by electrical simulations, and the generated optimal solution includes yield prediction as a design constraint. The method was applied for the automatic design of a 12-free-variables two-stage amplifier. The resulting sized circuit pre- sented 100 % yield within a 99 % confidence interval, while achieving all the performance specifications in a reasonable processing time. The authors imple- mentedanefficientyield-orientedsizingtoolwhichgeneratesrobustsolutions,thus increasing the number offirst-time-right analog integrated circuit designs. Chapter3,“ApplicationofComputationalIntelligenceTechniquestoMaximize Unpredictability in Multiscroll Chaotic Oscillators,” was written by Victor Hugo Carbajal-Gómez,EstebanTlelo-Cuautle,andFranciscoV.Fernández.Itappliesand comparesthreecomputationalintelligencealgorithms—thegeneticalgorithm(GA), differential evolution (DE), and particle swarm optimization (PSO)—to maximize the positive Lyapunov exponent in a multiscroll chaotic oscillator based on a sat- urated nonlinear function series based on the modification of the standard settings ofthecoefficientvaluesofthemathematicaldescription,andtakingintoaccountthe correct distribution of the scrolls drawing the phase-space diagram. The experi- mentalresultsshowthattheDEandPSOalgorithmshelptomaximizethepositive Lyapunov exponent of truncated coefficients over the continuous spaces. Chapter 4, “Optimization and Cosimulation of an Implantable Telemetric System by Linking System Modelsto Nonlinear Circuits,” was written by Yao Li, Hao Zou, Yasser Moursy, Ramy Iskander, Robert Sobot, and Marie-Minerve Louërat.Itpresentsaplatformformodeling,design,optimization,andcosimulation Preface vii of mixed-signal systems using the SystemC-AMS standard. The platform is based on a bottom-up design and top-down simulation methodologies. In the bottom-up design methodology, an optimizer is inserted to perform a knowledge-aware opti- mization loop. During the process, a Peano trajectory is applied for global explo- ration and the Nelder–Mead Simplex optimization method is applied for local refinement. The authors introduce an interface between system-level models and their circuit-level realizations in the proposed platform. Moreover, a transient simulationschemeisproposedtosimulatenonlineardynamicbehaviorofcomplete mixed-signal systems. The platform is used to design and verify a low-power CMOS voltage regulator for an implantable telemetry system. Chapter 5, “Framework for Formally Verifying Analog and Mixed-Signal Designs,” was written by Mohamed H. Zaki, Osman Hasan, Sofiène Tahar, and Ghiath Al-Sammane. It proposes a complementary formal-based solution to the verification of analog and mixed-signal (AMS) designs. The authors use symbolic computation to model and verify AMS designs through the application of induction-based model checking. They also propose the use of higher order logic theorem proving to formally verify continuous models of analog circuits. To test andvalidatetheproposedapproaches,theydevelopedprototypeimplementationsin Mathematica and HOL and target analog and mixed-signal systems such as delta sigma modulators. Chapter 6, “Automatic Layout Optimizations for Integrated MOSFET Power Stages,” was written by David Guilherme, Jorge Guilherme, and Nuno Horta. It presents a design automation approach that generates automatically error-free area and parasitic optimized layout viewsof output power stages consisting ofmultiple power MOSFETs. The tool combines a multitude of constraints associated with DRC, DFM, ESD rules, current density limits, heat distribution, and placement. It uses several optimization steps based on evolutionary computation techniques that precede a bottom-uplayout construction ofeach powerMOSFET, its optimization for area and parasitic minimization, and its optimal placement within the output stage power topology network. Chapter 7, “Optimizing Model Precision in High Temperatures for Efficient Analog and Mixed-Signal Circuit Design Using Modern Behavioral Modeling Techniques: an Industrial Case Study,” was written by Sahbi Baccar, Timothée Levi, Dominique Dallet, and François Barbara. It deals with the description of a modeling methodology dedicated to simulation of AMS circuits in high tempera- tures (HT). A behavioral model of an op-amp is developed using VHDL-AMS in order to remedy the inaccuracy of the SPICE model. The precision of the model simulation in HT was improved thanks to the VHDL-AMS model. Almost all known op-amp parameters were inserted into the model which was developed manually. Future work can automate the generation of such a behavioral model to describe the interdependency between different parameters. This is possible by usingmoderncomputationalintelligencetechniques,suchasgeneticalgorithms,or other techniques such as Petri nets or model order reduction. viii Preface Chapter8,“NonlinearitiesBehavioralModelingandAnalysisofPipelinedADC Building Blocks,” was written by Carlos Silva, Philippe Ayzac, Nuno Horta, and Jorge Guilherme. It presents a high-speed simulation tool for the design and analysis of pipelined analog-to-digital converters implemented using the Python programming language. The development of an ADC simulator requires the behavior modeling of thebasic building blocks and their possibleinterconnections to form the final converter. This chapter presents a Pipeline ADC simulator tool which allows topology selection and digital calibration of the frontend blocks. Several block nonlinearities are included in the simulation, such as thermal noise, capacitor mismatch, gain and offset errors, parasitic capacitances, settling errors, and other error sources. Part II—Radio-Frequency Design Chapter 9, “SMAS: A Generalized and Efficient Framework for Computationally Expensive Electronic Design Optimization Problems,” was written by Bo Liu, FranciscoV.Fernández,Georges Gielen,AmmarKarkar, AlexYakovlev,andVic Grout. Many electronic design automation (EDA) problems encounter computa- tionally expensive simulations, making simulation-based optimization impractical formanypopularsynthesismethods.Notonlyaretheycomputationallyexpensive, butsomeEDAproblemsalsohavedozensofdesignvariables,tightconstraints,and discrete landscapes. Few available computational intelligence methods can solve them effectively and efficiently. This chapter introduces a surrogate model-aware evolutionary search (SMAS) framework, which is able to use much fewer expen- sive exact evaluations with comparable or better solution quality. SMAS-based methods for mm-wave integrated circuit synthesis and network-on-chip parameter design optimization are proposed and are tested on several practical problems. Experimental results show that the developed EDA methods can obtain highly optimized designs within practical time limitations. Chapter 10, “Computational Intelligence Techniques for Determining Optimal Performance Trade-offs for RF Inductors,” was written by Elisenda Roca, Rafael Castro-López,FranciscoV.Fernández,ReinierGonzález-Echevarría,JavierSieiro, Neus Vidal, and José M. López-Villegas. The automatic synthesis of integrated inductorsforradio-frequency(RF)integratedcircuitsisoneofthemostchallenging problemsthatRFdesignershavetoface.Inthischapter,computationalintelligence techniquesareappliedtoautomaticallyobtaintheoptimalperformancetrade-offsof integrated inductors. A methodology is presented that combines a multiobjective evolutionary algorithm with electromagnetic simulation to get highly accurate results.Asetofsizedinductorsisobtainedshowingthebestperformancetrade-offs for a given technology. The methodology is illustrated with a complete set of examples where different inductor trade-offs are obtained. Chapter 11, “RF IC Performance Optimization by Synthesizing Optimum Inductors,”waswrittenbyMladenBožanićandSaurabhSinha.Itreviewsinductor theory and describes various integrated inductor options. It also explains why integrated planar spiral inductors are so useful when it comes to integrated RF circuits. Furthermore, the chapter discusses the theory of spiral inductor design, Preface ix inductor modeling, and how this theory can be used in inductor synthesis. In the centralpartofthechapter,theauthorspresentamethodologyforsynthesisofplanar spiral inductors, where numerous geometries are searched through in order to fit various initial conditions. Chapter12,“OptimizationofRFOn-ChipInductorsUsingGeneticAlgorithms,” was written by Eman Omar Farhat, Kristian Zarb Adami, Owen Casha, and John Abela. It discusses the optimization of the geometry of RF on-chip inductors by means ofageneticalgorithm inorder toachieve adequateperformance. Necessary backgroundtheorytogetherwiththemodelingoftheseinductorsisincludedinorder toaidthediscussion.Asetofguidelinesforthedesignofsuchinductorswithagood quality factor ina standard CMOS process isalso provided. The optimization pro- cess is initialized by using a set of empirical formulae in order to estimate the physicalparametersoftherequiredstructureasconstrainedbythetechnology.Then automated design optimization is executed to further improve its performance by means of dedicated software packages. The authors explain how to use state-of-the-artcomputer-aideddesigntoolsintheoptimizationprocessandhowto efficientlysimulatethe inductor performance using electromagnetic simulators. Chapter 13, “Automated System-Level Design for Reliability: RF Front-End Application,” was written by Pietro Maris Ferreira, Jack Ou, Christophe Gaquière, and Philippe Benabes. Reliability is an important issue for circuits in critical applicationssuchasmilitary,aerospace,energy,andbiomedicalengineering.With the rise in the failure rate in nanometer CMOS, reliability has become critical in recentyears.Existingdesignmethodologiesconsiderclassicalcriteriasuchasarea, speed, and power consumption. They are often implemented using post-synthesis reliabilityanalysisandsimulationtools.Thischapterproposesanautomatedsystem designforreliabilitymethodology.Whileaccountingforacircuit’sreliabilityinthe early design stages, the proposed methodology is capable of identifying an RF front-end optimal design considering reliability as a criterion. Chapter 14, “The Backtracking Search for the Optimal Design of Low-Noise Amplifiers,” was written by Amel Garbaya, Mouna Kotti, Mourad Fakhfakh, and PatrickSiarry.Thebacktrackingsearchalgorithm(BSA)wasrecentlydeveloped.It is an evolutionary algorithm for real-valued optimization problems. The main featureofBSAvis-à-visotherknownevolutionaryalgorithmsisthatithasasingle controlparameter.Ithasalsobeenshownthatithasabetterconvergencebehavior. Inthischapter,theauthorsdealwiththeapplicationofBSAtotheoptimaldesignof RF circuits, namely low-noise amplifiers. BSA performances, viz. robustness and speed, are checked against the widely used particle swarm optimization technique, and other published approaches. ADS simulation results are given to show the viability of the obtained results. Chapter 15, “Design of Telecommunications Receivers Using Computational Intelligence Techniques,” was written by Laura-Nicoleta Ivanciu and Gabriel Oltean. It proposes system-, block- and circuit-level design procedures that use computational intelligence techniques, taking into consideration the specifications for telecommunications receivers. The design process starts with selecting the properarchitecture(topology)ofthesystem,usingafuzzyexpertsolution.Next,at x Preface the block level, the issue of distributing the parameters across the blocks is solved using a hybrid fuzzy–genetic algorithms approach. Finally, multiobjective optimi- zation using genetic algorithms is employed in the circuit-level design. The pro- posed methods were tested under specific conditions and have proved to be robust and trustworthy. Chapter 16, “Enhancing Automation in RF Design Using Hardware Abstraction,” was written by Sabeur Lafi, Ammar Kouki, and Jean Belzile. It presents advances in automating RF design through the adoption of a framework that tackles primarily the issues of automation, complexity reduction, and design collaboration. The proposed framework consists of a design cycle along with a comprehensive RF hardware abstraction strategy. Being a model-centric frame- work,itcaptureseachRFsystemusinganappropriatemodelthatcorrespondstoa given abstraction level and expresses a given design perspective. It also defines a set of mechanisms for the transition between the models defined at different abstraction levels, which contributes to the automation of design stages. The combination of an intensive modeling activity and a clear hardware abstraction strategy through a flexible design cycle introduces intelligence, enabling higher design automation, and agility. Chapter 17, “Optimization Methodology Based on IC Parameter for the Design of Radio-Frequency Circuits in CMOS Technology,” was written by Abdellah Idrissi Ouali, Ahmed El Oualkadi, Mohamed Moussaoui, and Yassin Laaziz. It presents a computational methodology for the design optimization of ultra-low-powerCMOSradio-frequencyfront-endblocks.Themethodologyallows ustoexploreMOStransistors inall regionsof inversion. The powerlevelisset as an input parameter before we begin the computational process involving other aspects of the design performance. The approach consists of trade-offs between power consumption and other radio-frequency performance parameters. This can help designers to seek quickly and accurately the initial sizing of the radio-frequency building blocks while maintaining low levels of power consump- tion. A design example shows that the best trade-offs between the most important low-power radio-frequency performances occur in the moderate inversion region. Volume 2—Computational Intelligence in Digital and Network Designs and Applications Part I—Digital Design Chapter 1, “Sizing Digital Circuits Using Convex Optimization Techniques,” was written by Logan Rakai and Amin Farshidi. It collects recent advances in using convex optimization techniques to perform sizing of digital circuits. Convex opti- mizationtechniquesprovideanundeniablyattractivepromise:Theattainedsolution is the best available. In order to use convex optimization techniques, the target optimization problem must be modeled using convex functions. The gate sizing problem has been modeled in different ways to enable the use of convex optimi- zation techniques, such as linear programming and geometric programming. Statistical and robust sizing methods are included to reflect the importance of