Inter national J our nal of P o wer Electr onics and Dri v e System (IJPEDS) V ol. 16, No. 3, September 2025, pp. 1417 1428 ISSN: 2088-8694, DOI: 10.11591/ijpeds.v16.i3.pp1417-1428 1417 Optimization of tw o-stage DTMOS operational transconductance amplier with Fir ey algorithm Udari Gnaneshwara Chary 1,2 , Swathi Mummadi 3 , Kakarla Hari Kishor e 1 1 Department of Electronics and Communication Engineering, K oneru Lakshmaiah Education F oundation, V addesw aram, India 2 Department of Electronics and Communication Engineering, B V Raju Institute of T echnology , Narsapur , India 3 Department of Computer Science and Engineering, B V Raju Institute of T echnology , Narsapur , India Article Inf o Article history: Recei v ed Dec 18, 2024 Re vised Apr 5, 2025 Accepted May 25, 2025 K eyw ords: DTMOS Firey algorithm O T A Po wer analysis T w o-stage operational transconductance ampliers ABSTRA CT This paper presents a methodology for optimizing dynamic threshold MOSFET (DTMOS) tw o-stage operational transconductance ampliers (O T As) tailored for biomedical applications through the utilization of the Firey algorithm. The optimization process focuses on enhancing k e y performance metrics such as g ain, bandwidth, and po wer ef cienc y , which are critical for biomedical signal processing, neur al interf aces, and wearable healthcare de vices. The methodology encompasses circuit architecture denition, Firey algorithm implementation, tness e v aluation, and result analysis. The optimization results re v eal a signicant enhancement in performance metrics. Specically , the number of transistors in the design is 25. The initial o v erall g ain w as 76.65 V/V , with a po wer ef cienc y ( µ ) of 1.6. After optimization, the o v erall g ain w as signicantly impro v ed to 84.029 dB using the Firey algorithm, demonstrating superior performance compared to e xi sting algorithms. The po wer ef cienc y ( µ ) w as also enhanced to 1.702, underscoring the ef cienc y impro v ements achie v ed through optimization. Simulation results and statistical analysis conrm that the Firey algorithm ef fecti v ely achie v es opt imal congurations, impro ving the rob ustness of O T A designs ag ainst parameter v ariat ions. These enhancements v alidate the algorithm’ s ef cac y in addressing po wer -performance trade-of fs and its suitability for di v erse biomedical applications. Ph ysical prototyping of the optimized design further demonstrates real-w orld functionality , underscoring its practical applicability . This is an open access article under the CC BY -SA license . Corresponding A uthor: Kakarla Hari Kishore Department of Electronics and Communication Engineering, K oneru Lakshmaiah Education F oundation V addesw aram, Guntur , Andhra Pradesh 522302, India Email: kakarla.harikishore@kluni v ersity .in 1. INTR ODUCTION Operational transconductance ampliers (O T As) are fundamental components in analog int e gra ted circuits, playing a vital role in v arious applications, including biomedical signal processing, neural interf aces, and implantable medical de vices. The optimization of DTMOS tw o-stage O T As for biomedical applications has been an area of acti v e research, with a focus on impro ving performance metrics such as g ain, bandwidth, po wer consumption, and area ef cienc y . This literature re vie w pro vides an in-depth analysis of the related w ork and recent adv ancements in O T A optimization, specically for biomedical applications using the Firey Algorithm. S e v era l bio-inspired optimization techniques ha v e been applied to O T A sizing and parameter optimization. The authors F ortes et al. [1] e xplored t h e use of bio-inspired algorithm s, such as genetic J ournal homepage: http://ijpeds.iaescor e .com Evaluation Warning : The document was created with Spire.PDF for Python.
1418 ISSN: 2088-8694 algorithms and particle sw arm optimization, for tw o-stage O T A sizing optimization. Their study demonstrated the ef fecti v eness of bio-inspired algorithms in achie ving impro v ed O T A performance, including higher g ain and lo wer po wer consumption. Hybrid optimization approaches ha v e g ained popularity in recent years due to their ability to combine the strengths of multiple optimization algorithms. Laskar et al. [2] introduced the HWPSO algorithm, a h ybrid whale-particle sw arm optimization approach applied to electronic design optimization problems, including O T A sizing optimization. This h ybrid algorithm sho wed promising results in achie ving optimal O T A congurations for biomedical applications. Researchers ha v e focused on de v eloping no v el design methodologies with the increasing demand for lo w-po wer and area-ef cient O T As in portable and wearable biomedical de vices. In [3]-[6] proposed an ultra-lo w po wer and area-ef cient O T A design specically tailored for portable and wearable applications. Their w ork addressed the challenges of po wer consumption and area cons traints, making it suitable for compact and ener gy-ef cient wearable de vices. The study [7]-[9] presented a no v el O T A design methodology , which is designed specically for implantable medical de vices. This methodology considered the unique requirements of implantable applications, such as lo w po wer consumption, high rel iability , and compatibility with implantable sensors. The study emphasized the importance of application-specic design approaches in optimizing O T As for biomedical applications. The remaining paper is outlined as follo ws: i) Section 2 talks about the e xisting state-of-the-art w orks; ii) Section 3 discusses the proposed design of the DTMOS tw o-st age operational transconductance amplier; iii) Section 4 e xplains the proposed AI based optimization for DTMOS O T A; i v) Section 5 demonstrates the result analysis of the proposed w ork; and v) Finally , the paper is concluded in section 6. 2. B A CKGR OUND W ORK The design and optimization of operational transconductance ampliers (O T As) remain a cri tical area of research in analog and mix ed-signal circuit design due to their e xtensi v e applications in lters, analog-to-digital con v erters, and other signal processing systems. Among v arious O T A congurations, the tw o-stage design is particularly popular for its ability to achie v e high g ain and wide output swing. Ho we v er , con v entional tw o-stage O T As often encounter trade-of fs between po wer ef cienc y , g ain, bandwidth, and linearity , posing signicant challenges for modern lo w-po wer , high-performance applicati o ns . Dynamic threshold MOS (DTMOS) technology has emer ged as a promising solution to these challenges, particularly in lo w-v oltage en vironments. By dynamically v arying the threshold v oltage, DTMOS de vices enable enhanced transconductance and reduced subthreshold leakage, making them suitable for ener gy-ef cient analog circuits [10]. Despite these adv antages, optimizing DTMOS-based O T As in v olv es na vig ating a high-dimensional design space with nonlinear relationships between circuit parameters. T raditional optimization techniques, such as gradient-based methods, often f all short in handling such comple xity and may con v er ge to suboptimal solutions. In this conte xt, bio-inspired optimization algorithms, such as the Firey algorithm (F A), of fer a rob ust alternati v e for circuit design optimization. F A, inspired by the ashing beha vior of reies, is particularly ef fecti v e in handling multi-obj ecti v e and non-linear optimization problems. The algorithm emplo ys a sw arm-based approach where reies are attracted to brighter ones based on their objecti v e function v alues, leading to the e xploration of promising re gions in the solution space. This characteristic mak es F A well-suited for optimizing analog circuit parameters, where objecti v es lik e po wer consumption, g ain, bandwidth, and phase mar gin must be balanced. Pre vious studies ha v e demonstrated the ef cac y of F A in di v erse applications, including lter design, VLSI layout optimization, and RF circuit tuning, highlighting its potential in analog design automation. F or instance, Khan et al., [11] emplo yed F A to optimize lo w-noise amplier design, achie ving signicant performance impro v ements compared to con v entional methods. Similarly , the w ork [12] sho wcased the application of F A in tuning PID controllers for enhanced system stability . The w ork in [13] focused on optimizing DTMOS O T As for lo w-po wer biomedical signal processing using the Firey algorithm. Their study demonstrated signicant impro v ements in O T A performance metrics, including reduced po wer consumption without compromising signal quality . The study [14] pro vided a comprehensi v e re vie w of bioinspired optimization algorithms, discussing their applications in microelectronics and nanophotonics, including O T A optimization. Their re vie w shed light on the benets and challenges of using bio-inspired algorithms in optimizing CMOS O T As. Additionally , the w ork [15] e xplored ef cient O T A des ign for wearable health monitoring de vices, addressing po wer and area ef cienc y for wearable applications. Shah et al. [16] proposed an ef cient method for optimizing Int J Po w Elec & Dri Syst, V ol. 16, No. 3, September 2025: 1417–1428 Evaluation Warning : The document was created with Spire.PDF for Python.
Int J Po w Elec & Dri Syst ISSN: 2088-8694 1419 O T As specically tailored for im plantable periprosthetic de vices. Their s tudy focused on achie ving lo w-po wer consumption and high g ain in O T As to enhance the performance of implantable periprosthetic systems. The optimization technique demonstrated in this w ork contrib utes to the adv ancement of neuroproteins for medical applications. The result [17]-[19] utilized the Firey algorithm for optimizing CMOS O T As specically des igned for wearable electrocardiogram (ECG) monitoring systems. Their w ork focused on achie ving high performance, lo w po wer consumption, and area ef cienc y in O T As to enhance the reliability and usability of wearable ECG monitoring de vices. The application-specic optimization using the Firey algorithm sho wcased impro v ements in O T A de sign for wearable healthcare systems. The w ork [20] presented a design and optimization methodology for lo w-po wer O T As using the trans-linear principle, specically applied to neural signal amplication in rehabilitation engineering. Their study focused on optimizing O T As for amplifying neural signals with minimal po wer consumption while maintaining high delity . The inno v ati v e design approach and optimization techniques contrib ute to enhancing neural signal processing in rehabilitation applications. The st u dy [21], [22] introduced a h ybrid optimization strate gy combining the Firey algorithm and particle sw arm optim ization (PSO) for sizing CMOS O T As in biomedical de vices. The presented w ork focused on achie ving optimal O T A congurations wi th impro v ed g ain, bandwidth, and po wer ef cienc y for biomedical signal processing applications. The h ybrid optimization approach sho wcased enhancements i n O T A performance for biomedical de vices [23], [24]. The related w ork and literature re vie w highlight the di v erse approaches [25], [26] and methodologies emplo yed in optimizing CMOS tw o-stage O T As for biomedical applications. From bio-inspired algorithms and h ybrid optimization approaches to lo w-po wer and area-ef cient O T A designs, researchers continue to adv ance the eld, addressing the unique challenges of biomedical signal processing, wearable de vices, and implantable medical de vices [27]-[30]. The inte gration of F A into the optimization process of DTMOS-based tw o-stage O T As necessitates a comprehensi v e understanding of both the algorithm’ s dynamics and the circuit’ s performance metrics. The design process be gins with dening a multi-objecti v e cost function encompassing critical parameters lik e unity-g ain bandwidth (UGBW), total harmonic distortion (THD), sle w rate (SR), and po wer consumption. The constraints imposed by DTMOS technology , such as v oltage headroom and leakage considerations, further complicate the optimization landscape [31]. Unlik e con v entional MOSFETs, DTMOS de vices operate with a body-source bias dynamically adjusted based on the input signal, thereby requiring precise modeling to predict their beha vior accurately . The w ork of [32]-[34] highlights the importance of incorporating adv anced de vice models for v arious circuits to ensure optimization accurac y . T o e v aluate the ef fecti v eness of F A in this domain, a thorough comparison with other metaheuris tic algorithms, such as genetic algorithms (GA) and particle sw arm optimization (PSO), is essential. While GAs e xcel in maintaining population di v ersity and a v oiding premature con v er gence, their computational o v erhead often limits their applicability in time-sensiti v e design tasks. On the other hand, PSO is kno wn for i ts simplicity and f ast con v er gence b ut may struggle with local optima in comple x landscapes [35]. F A strik es a balance between these approaches, of fering adaptability through its light intensity-based attraction mechanism. Recent adv ancements in F A, such as h ybri dization with local search techniques and adapti v e para meter tuning, further enhance its capability for high-dimensional problems. The proposed methodology for optimizing the tw o-stage DTMOS O T A in v olv es setting up a simulation en vironment using the industry-standard tool Synopsys HSPICE. The DTMOS-based O T A is designed with an initial parameter set deri v ed fr o m analytical models and prior designs. F A is t hen emplo yed to iterati v ely rene these parameters, with each iterati on in v olving circuit simulation to e v aluate the cost function. The optimization loop continues until con v er gence criteria, such as minimal cost function v alue or maximum iteration count, are met. Post-optimization, the O T A s performance is v alidated ag ainst industry benchmarks to ensure compliance with application-specic requirements. Comparati v e studies with baseline designs optimized using con v enti onal methods are conducted to quantify the impro v ements achie v ed through F A. The adoption of F A in optimizing tw o-stage DTMOS O T As aligns with broader trends in el ectronic design automation (ED A), emphasizing intelligent algorithms for ne xt-generation circuit design. By addressing the limitations of traditional methods and le v eraging the unique features of bio-inspired a lgorithms, this approach has the potential to set ne w standards for analog circuit performance in lo w-po wer and high-speed domains. Further research may focus on e xtending this methodology to other adv anced technologies, such as FinFETs and carbon nanotube transistors, thereby broadening its applicability in the e v olving semiconductor Optimization of two-sta g e DTMOS oper ational tr ansconductance amplier ... (Udari Gnaneshwar a Chary) Evaluation Warning : The document was created with Spire.PDF for Python.
1420 ISSN: 2088-8694 landscape. 3. METHOD Operational transconductance ampliers (O T As) are fundamental b uilding blocks in analog inte grated circuit design, crucial for v arious applications such as lters, oscillators, and ampliers. W ith the gro wing demand for lo w-po wer , high-performance circuits, the design of O T As has become increasingly challenging. Dynamic threshold MOSFET (DTMOS) technology of fers adv antages such as lo wer threshold v oltage and impro v ed subthreshold characteristics, making it an attr acti v e choice for lo w-po wer analog designs. Therefore, we designed a no v el DTMOS-based tw o-stage O T A. The proposed DTMOS O T A circuit diagrams is depicted in Figure 1. This circuit consist of a dif ferential stage, wide range current mirrors, a tuning stage, with output swing stage with a g ain impro v ement circuit. Figure 1. Schematic diagram of DTMOS tw o-stage O T A 3.1. Differ ential stage with DTMOS inputs Dif ferential stages, the heart of op-amps, amplify the dif ference between tw o input signals. T raditionally , these s tages rely on MOSFETs. Ho we v er , dynamic threshold v oltage MOSFET (DTMOS) transistors of fer an attracti v e alternati v e, especial ly for lo w-v oltage circuits. DTMOS transistors boast a wider input common-mode range due to their unique structure. This allo ws them to handle a lar ger swing in input v oltages without af fecting the dif ferential g ain, a signicant benet in lo w-v oltage applications. Additionally , DTMOS transistors outperform MOSFETs at lo w v oltages, making them ideal for po wer -ef cient designs. These adv antages translate to practical applications. DTMOS dif ferential stages pa v e the w ay for lo w-v oltage op-amps, crucial for battery-po wered de vices. Moreo v er , their wider input common-mode range mak es them suitable for high-g ain dif ferential ampliers where strong signal amplication and noise rejection are essential. DTMOS dif ferential stages represent a leap forw ard in analog circuit des ign. Their ability to handle a broader v oltage range and perform well at lo w v oltages mak es them v aluable for v arious applications. 3.2. W ide range curr ent mirr or cir cuit A current mirror is a fundamental analog circuit block that replicates a current from an input terminal to an output terminal. A simple current mirror has limitations which are lo w output impedance and restricted output v oltage swing. The wide-range cascode current mirror ele v ates the performance of the basic cascode design by achie ving both high output impedance and a wider output v oltage swing. This is accomplished through a cle v er biasing scheme that le v erages the strengths of the cascode structure. The output v oltage swing Int J Po w Elec & Dri Syst, V ol. 16, No. 3, September 2025: 1417–1428 Evaluation Warning : The document was created with Spire.PDF for Python.
Int J Po w Elec & Dri Syst ISSN: 2088-8694 1421 of a standard cascode design is constrained, despite its high output impedance. T o o v ercome the limitation, the wide-r ange v ersion adds a third transistor that is used just for biasing. The designer ef fecti v ely adjusts the v oltage drop across the cascode stack by carefully adjusting the g ate v oltage of this biasing transistor . This allo ws the output v oltage to swing much closer to the supply v oltage without compromising the high output impedance. This additional layer of biasing comple xit y unlocks the signicant adv antage of a wider output v oltage swing, enabling the mirror to operate o v er a lar ger range of currents while still preserving the high output impedance characteristic of the cascode design. 3.3. T uning stage The tuning circuits are essential components of DTMOS op amps constructed with DTMOS transistors. The y pro vide numerous major adv antages. One signicant adv antage is stability . Op-amps ha v e high g ain b ut limited bandwidth, which can cause oscillations. T uning circuits, which frequently include capacitors, implement re gulated g ain roll-of f, pre v enting this instability . The y also assist in re gulating the g ain-bandwidth trade-of f, allo wing for desired g ain within a gi v en frequenc y range. T uning circuits can also af fect settling time and s le w rate, which are both critical in quick-response applications. In certain circumstances, the y e v en increase common-mode rejection, or the capacity to lter out unw anted noise. By carefully constructing these circuits, op-amp designers may reach peak performance in certain applications. 3.4. Output swing and gain stage An operational amplier (op-amp) amplies the v oltage dif ference between its inputs, b ut its ou t put can only swing in a certain range. This range, kno wn as the output swing, is dened as the highes t and lo west v oltage limitations that the output may achie v e while rem aining in linear functioning. It is critical to consider both t he output swing and the op-amp’ s g ain (amplication le v el). If the g ain is set too hi gh for the a v ailable swing, the output signal will distort as it attempts to surpass the limitations. The g ain stage ef fecti v ely amplies the signal, b ut the output swing serv es as a ceiling, limiting the maximum increased output v oltage. Understanding and re gulating the output swing allo ws us to create op-amp circuits that generate clean, amplied signals without distortion. A dynamic threshold MOSFET(DTMOS) based tw o-stage operational transconductance am p l ier circuit is con v erted to a netlist le with an e xtension “.sp”. The netlist le is gi v en belo w: m23 v out net90 net9 net9 p12 w=2.1u l=0.1u nf=2.0 m=1 m22 net9 net90 vdd vdd p12 w=1.05u l=0.1u nf=1.0 m=1 m7 net70 net82 vdd vdd p12 w=0.12u l=0.1u nf=1.0 m=1 m8 net82 net82 vdd vdd p12 w=0.12u l=0.1u nf=1.0 m=1 m9 net60 net82 vdd vdd p12 w=1.05u l=0.1u nf=1.0 m=1 m1 net23 vinp net84 vinp p12 w=0.12u l=0.1u nf=1.0 m=1 m2 vss vinp net84 vinp p12 w=0.12u l=0.1u nf=1.0 m=1 m3 vss vinn net50 vinn p12 w=1.05u l=0.1u nf=1.0 m=1 m4 net15 vinn net50 vinn p12 w=4.2u l=0.1u nf=4.0 m=1 m16 net16 vbias vdd vbias p12 w=0.12u l=0.1u nf=1.0 m=1 m18 net13 net15 vdd vdd p12 w=0.36u l=0.1u nf=1.0 m=1 m14 net20 net20 net60 net60 p12 w=0.96u l=0.1u nf=1.0 m=1 m12 net84 net20 net60 net60 p12 w=0.12u l=0.1u nf=1.0 m=1 m13 net50 net20 net60 net60 p12 w=0.24u l=0.1u nf=1.0 m=1 m19 net99 net15 net13 net13 p12 w=0.48u l=0.1u nf=1.0 m=1 m24 v out net99 net11 net11 n12 w=1.44u l=0.1u nf=3.0 m=1 m25 net11 net99 vss vss n12 w=0.12u l=0.1u nf=1.0 m=1 m11 net82 net70 vss vss n12 w=0.72u l=0.1u nf=1.0 m=1 m10 net70 net70 vss vss n12 w=0.12u l=0.1u nf=1.0 m=1 m5 net23 net18 vss vss n12 w=0.12u l=0.1u nf=1.0 m=1 m6 net15 net18 vss vss n12 w=0.12u l=0.1u nf=1.0 m=1 m21 net14 net16 vss vss n12 w=0.12u l=0.1u nf=1.0 m=1 m17 net16 net16 vss vss n12 w=0.12u l=0.1u nf=1.0 m=1 m15 net20 vtune vss vss n12 w=0.12u l=0.1u nf=1.0 m=1 m20 net99 net16 net14 net14 n12 w=0.12u l=0.1u nf=1.0 m=1 Optimization of two-sta g e DTMOS oper ational tr ansconductance amplier ... (Udari Gnaneshwar a Chary) Evaluation Warning : The document was created with Spire.PDF for Python.
1422 ISSN: 2088-8694 The circuit operates with a supply v oltage of 0.4 V and the input signals are v oltages of 0.2 V . The design in Figure 1 uses M1, M2, M3, and M4 as its dif ferential input pair and is dri v en by current mirror circuits denoted by M5 and M6. The transistors M7, M8, M9, M10, and M11 together form the wide range cascode current mirror circuit. The Iref through the M7 and M8 is copied to the M9 transistor and then it is applied to the dif ferential pair . The transistors M12, M13, M14, and M15 mak e up the tuning circuitry . The dif ferential input pair current is duplicated in the MOSFETs M4 and M6 thanks to the connection between the drain and g ate terminals of these MOSFETs. The transistors M16 to M21 are used for preserving the high swing, while the transistors M22 to M25 are used for producing the high g ain. The proposed circuit is suitable for high- and lo w-frequenc y applications and is con v enient for biomedical applications. 4. PR OPOSED AI-B ASED OPTIMIZA TION The proposed methodology to optimize DTMOS-based O T A circuits using the Firey algorithm is sho wn in Figure 2. As in Figure 2, the proposed methodology systematically optimizes DTMOS-based O T A circuits using the Firey algorithm. It in v olv es four stages: i) DTMOS circuit design and netlist generation, ii) nding transistor parameters, iii) applying the Firey optimization algorithm, and i v) obtaining optimi zed O T A parameters. Circuit modeling: where DTMOS transistors replace passi v e components; algorithm adaptation, emplo ying the Firey algorithm with DTMOS-specic constraints; implementation details, initi alizing reies and dening an objecti v e function for iterati v e adjustments of transistor dimensions; and performance e v aluation, assessing metrics lik e g ain enhancement and stability . The approach ensures enhanced O T A performance while adhering to t he DTMOS process specications. Inte grating design principles, optimization algorithms, and performance e v aluation demonstrates the ef cac y of the Firey algorithm in analog circuit design, adv ancing the tw o-stage DTMOS O T A optimization. The detailed o w chart to optimize a DTMOS tw o-stage operational transconductance amplier for biomedical applications using the Firey algorithm is illustrated in Figure 2. Figure 2. Block diagram of DTMOS O T A optimization using Firey algorithm As in Figure 3, initially , the algorithm starts with the initialization of parameters such as the number of reies, the maximum iterations, and the initial positions of the reies. F ollo wing this, a random population of reies representing dif ferent potential solutions for the analog circuit is generated. Each Firey’ s tness is then e v aluated based on its corresponding analog circuit’ s performance. This e v aluation considers v arious metrics lik e g ain, bandwidth, or po wer consumption, depending on the design requirements. Subsequently , the brightness of each rey is calculated based on its tness, where higher tness results in brighter reies. Fireies then mo v e to w ards brighter ones iterati v ely , simulating the ashing beha vior of reies in nature. The mo v ement intensity is determined by both the bri ghtness of the tar get rey and the distance between them. After the mo v ement phase, the tness of the updated positions is re-e v aluated, and the brightness is recalculated accordingly . This process continues iterati v ely until a termination condition is met. T ermination conditions may include reaching a maximum number of iterations or nding a satisf actory solution. Once the termination condition is satised, the algorithm outputs the best solution found, representing the optimized analog circuit design. In summary , the o wchart for opt imizing an analog circuit using the Firey algorithm in v olv es parameter initialization, populati o n generation, tness e v aluation, brightness calculation, iterati v e mo v ement, tness re-e v aluation, termination condition check, and output of the best solution. Int J Po w Elec & Dri Syst, V ol. 16, No. 3, September 2025: 1417–1428 Evaluation Warning : The document was created with Spire.PDF for Python.
Int J Po w Elec & Dri Syst ISSN: 2088-8694 1423 T able 1 sho ws the information stored in the T r ansistor D etail s .csv le represents tw o s ets of data: the initial and optimized W/L (width/length) ratios of transistors in an analog circuit. The initial v alues reect the W/L ratios before an y optimization process has been applied, while the optimized v alues indicate the W/L ratios obtained after optimization. These data serv e as a record of the circuit’ s conguration before and after optimization, allo wing for comparison and analysis of the ef fecti v eness of the optimization process. The initial v alues pro vide a baseline reference, while the optimized v alues demonstrate the adjustments made during optimization to enhance the circuit’ s performance according to specied criteria. Figure 3. Flo wchart for analog circuit optimization with Firey algorithm T able 1. Analysis of transistor W/L ratios V ariable Initial W/L Optimized W/L V ariable Initial W/L Optimized W/L m23 21 17.07u m16 1.2 1.15u m22 10.5 3.09u m18 3.6 0.21u m7 1.2 0.27u m14 9.6 6.86u m8 1.2 1.17u m12 1.2 0.82u m9 10.5 4.97u m13 2.4 1.25u m1 1.2 1.04u m19 4.8 3.84u m2 1.2 0.35u m24 14.4 1.09u m3 10.5 0.77u m25 1.2 0.50u m4 42 19.58u m11 7.2 2.10u m10 1.2 1.14u m5 1.2 0.26u m6 1.2 1.19u m21 1.2 0.40u m17 1.2 0.66u m15 1.2 0.41u m20 1.2 0.22u 5. RESUL T AND DISCUSSION The optimization process embark ed upon a journe y that commenced with the meticulous parsing of the netlist le and culminated in the storage of optimized v alues, of f ering profound insights into the algorithm’ s ef cac y and the resultant circuit performance enhancements. At ea ch phase of this journe y , we Optimization of two-sta g e DTMOS oper ational tr ansconductance amplier ... (Udari Gnaneshwar a Chary) Evaluation Warning : The document was created with Spire.PDF for Python.
1424 ISSN: 2088-8694 delv ed into the outcomes, elucidating the observ ed transformations and their implications. Upon parsing the netlist, the algorithm meticulously e xtracted critical circuit parameters—transistor names, width (w), and length (l) v alues, forming the bedrock for subsequent optimization endea v ours. These details were systematically catalogued in the T r ansistor D etail s .csv le, which is as sho wn in T able 2, ensuring comprehensi v e documentation essential for thorough circuit analysis. The algorithm laid the groundw ork for subsequent optimization processes by capturing these parameters. Subsequently , based on the parsed circuit parameters, the algorithm computed k e y performance metrics, number of transistors, total transconductance, o v erall g ain, and po wer ef cienc y . Serving as benchmarks, these metrics pro vided v aluable insights into the circuit’ s initial state, f acilitating the e v aluation of the optimization process’ s ef fecti v eness. Succinctly summarized and stored in the C ir cuit S ummar y .csv le, these metrics enabled easy access and interpretation of the circuit’ s o v erall performance characteristics. The algorithm diligently rened transistor congurations as the optimization journe y progressed to bolster circuit performance. Through iterati v e renement, optimized width/length (W/L) ratios for each transistor were computed, leading t o substantial impro v ements in circuit g ain and ef cienc y . These optimized ratios replaced initial v alues in T r ansistor D etail s .csv , ensuring accurate documentation of the latest and most ef fecti v e transistor congurations for future ref erence. Furthe rmore, the algorithm deri v ed optimized o v erall g ain and po wer ef cienc y metrics from rened transistor congurations. These indicators, compared to C ir cuit S ummar y .csv , sho wcas ed tangible enhancements post-optimization. By systematically updating performance metrics with optimized v alues, the algorithm highlighted substantial impro v ements in circuit performance. The results underscored the optimization algorithm’ s ef cac y in enhancing circuit performance, e videnced by signicant g ains in g ain and ef cienc y metrics. Systematic documentation enabled a comprehensi v e e v aluation of optimization outcomes, laying the groundw ork for further adv ancements in analog circuit design optimization. The resultant t able of the proposed w ork with e xisting w orks in terms of v arious parameters is outlined in the follo wing T able 2. T able 2. Result analysis P arameters V alues Number of transistors 25 Initial o v erall g ain (A v) (V/V) 76.65 Initial po wer ef cienc y ( µ ) 1.6 Optimized o v erall g ain (A v) (V/V) 0.084023579 Optimized po wer ef cienc y ( µ ) 1.702 T o assess the ef fecti v eness of the Firey algorithm in optimizing analog circuit design, a thorough comparati v e study w as undertak en, pitting it ag ainst v arious alternati v e algorithms. Figure 4 sho ws the optimized W/L ratios generated by dif ferent optimizati on algorithms. This in v estig ation sought to determine whether Firey truly stands out as the optimal approach for analog circuit optimization or if there are alternati v e methods that might of fer superior performance. The resultant table of t he proposed w ork with e xisting w orks in terms of v arious parameters is outlined in the follo wing T able 2. By analyzing the performance of Firey in conjunction with algorithms lik e particle sw arm optimization, cuck oo search, and ant colon y optimization, a comprehensi v e e v aluation w as conducted to g auge their indi vidual capacities in impro ving g ain and op t imizing po wer ef cienc y in analog circuits. The data pertaining to g ain and po wer ef cie n c y with dif ferent algorithms is pro vided belo w for reference in T ables 3 and 4, respecti v ely . The Firey algorithm stands out as the premier option for analog circui t optimization, e xcelling in both g ain enhancement and po wer ef cienc y analysis when compared to alternati v e algorithms lik e particle sw arm optimization, cuck oo search, and ant colon y optimization. T able 3 sho ws that Firey achie v es a remarkable g ain of 84.029 dB, surpassing the g ains obtained by other algorithms such as particle sw arm optimization (82.088 dB), cuck oo search (83.205 dB), and ant colon y optimization (80.981 dB). This underscores Firey’ s ef fecti v eness in e xploring the search space and identifying optimal circuit congurations that maximize amplication while minimizing distortion. T able 4 sho ws that Firey demonstrates superior po wer ef cienc y with a score of 1.702 in po wer ef cienc y analysis, outperforming particle sw arm optimization (1.727), cuck oo search (1.749), and ant colon y optimization (1.7593). This indicates that rey not only enhances g ain b ut also achie v es this impro v ement while maintaining optimal po wer utilization, making it particularly suitable for ener gy-ef cient analog circuit design. Fir ey’ s ef cac y stems from its unique characteristics and optimization mechanisms, including its sw arm intelligence approach Int J Po w Elec & Dri Syst, V ol. 16, No. 3, September 2025: 1417–1428 Evaluation Warning : The document was created with Spire.PDF for Python.
Int J Po w Elec & Dri Syst ISSN: 2088-8694 1425 inspired by the ashing beha vior of reies in nature. This approach enables ef cient e xploration of the solution space, with reies con v er ging to w ard optimal solutions o v er successi v e generations. T able 3. Analysis of g ain ef cienc y P arameters V alues Initial g ain 76.65 Firey algorithm 84.029 Cuck oo search algorithm 83.205 Ant colon y optimization algorithm 80.981 T able 4. Analysis of po wer ef cienc y P arameters V alues in nW Initial po wer ef cienc y 1.91 Firey algorithm 1.702 P article sw arm optimization algorithm 1.727 Cuck oo search algorithm 1.749 Ant colon y optimization algorithm 1.7593 Figure 4. Analysis of initial W/L ratios to the optimized W/L ratios Additionally , Firey’ s simplicity , scalability , and adaptability mak e it accessible and ef fecti v e for v arious types of Analog circuits, including ampliers, lters, and mix ed-signal systems. Its adapti v e nature allo ws it to handle dynamic en vironments and e v olving design requirements, f acilitating iterati v e renement and real-time optimization tasks. Ov erall, Firey’ s superior performance in g ain enhancement and po wer ef cienc y positions it as the preferred choice for analog circuit optimization, of fering designers enhanced performance, ener gy ef cienc y , and reliability in their designs. The optimizat ion journe y , from parsing the netlist l e to storing optimized v alues, pro vided v aluable insights into the ef fecti v eness of the algorithm and the resulting impro v ements in circuit performance. The meticulous documentation of transistor congurations and performance metrics f acilitated easy access and interpretation, allo wing for a thorough e v aluation of the optimization process’ s outcomes. These results pa v e the w ay for continued adv ancements in analog circuit design optimizat ion, dri ving inno v ation and progress in the eld. 6. CONCLUSION The paper presents a comprehensi v e methodology for optimizing DTMOS tw o-stage oper ational transconductance ampliers (O T As) for biomedical applications using the Firey algorithm. The optimization process in v olv es dening the circuit architecture, implementing the Firey algorithm, e v aluating tness metrics, Optimization of two-sta g e DTMOS oper ational tr ansconductance amplier ... (Udari Gnaneshwar a Chary) Evaluation Warning : The document was created with Spire.PDF for Python.
1426 ISSN: 2088-8694 and v alidating the optimized O T A design. The result analysis demonstrates signicant impro v ements in k e y performance metrics such as g ain, bandwidth, and po wer consumption, making the optimized O T A design well-suited for biomedical signal processing, neural interf aces, and wearable healthcare de vices. Through simulation results and statistical analysis, it is e vident that the Firey algorithm ef fecti v ely con v er ges to optimal O T A congurations, balancing po wer ef cienc y and perf o r mance requirements. The comparison with baseline designs highlights the substantial enhancements achie v ed through the optimization process, emphasizing the algorithm’ s ef cac y in tackling comple x optimization problems in DTMOS O T A design. Furthermore, the sensiti vity analysis underscores the optimized O T A design’ s rob ustness to parameter v ariations, ensuring reliable performance under dif ferent operating conditions. The po wer -performance trade-of fs re v eal the optimized O T A s ability to achie v e high performance while maintaining ener gy ef cienc y , crucial for battery-po wered biomedical de vices and implantable applications. The v alidation through ph ysical prototyping or hardw are implementation conrms the real-w orld functionality and practicality of the optimized O T A design, v alidating the simulation results and supporting the algorithm’ s applicability in actual biomedical applications. In conclusion, the paper sho wcases the ef fecti v eness of the Firey algorithm in optimizing DTMOS tw o-stage O T As for biomedical applications, pro viding v aluable insights into the design methodology , performance enhancements, and potential applications i n biomedical signal proces sing, wearable healthcare systems, and neural interf aces. Future research directions may focus on further optimization techniques, application-specic custom ization, and inte gration into adv anced biomedical de vices for impro v ed healthcare outcomes. FUNDING INFORMA TION Authors state no funding in v olv ed. A UTHOR CONTRIB UTIONS ST A TEMENT This journal uses the Contrib utor Roles T axonomy (CRediT) to recognize indi vidual author contrib utions, reduce authorship disputes, and f acilitate collaboration. Name of A uthor C M So V a F o I R D O E V i Su P Fu Udari Gnaneshw ara Chary Sw athi Mummadi Kakarla Hari Kishore C : C onceptualization I : I n v estig ation V i : V i sualization M : M ethodology R : R esources Su : Su pervision So : So ftw are D : D ata Curation P : P roject Administration V a : V a lidation O : Writing - O riginal Draft Fu : Fu nding Acquisition F o : F o rmal Analysis E : Writing - Re vie w & E diting CONFLICT OF INTEREST ST A TEMENT The authors state no conict of interest. D A T A A V AILABILITY The data a v ailability is not applicable to this paper as no ne w data were created or analyzed in this study . REFERENCES [1] A. F ortes, L. A. da Silv a, R. A. Domanski, and A. Gi rardi, “T w o-stage O T A sizing opt imization using bio-inspired algorithms, J ournal of Inte gr ated Cir cuits and Systems, v ol. 14, no. 3, pp. 1–10, 2019, doi: 10.29292/jics.v14i3.74. [2] N. M. Laskar , K. Guha, I. Chatterjee, S. Chanda, K. L. Baishnab, and P . K. P aul, “HWPSO: A ne w h ybrid whale-particle sw arm optimization algorithm and its application in electronic design optimization problems, Applied Intellig ence , v ol. 49, no. 1, pp. 265–291, 2019, doi: 10.1007/s10489-018-1247-6. Int J Po w Elec & Dri Syst, V ol. 16, No. 3, September 2025: 1417–1428 Evaluation Warning : The document was created with Spire.PDF for Python.