Fault correcting adder design for low power applications Authors: Pritty Published: 2024 ABSTRACT AbstractField Programmable Gate Arrays are extensively used in space, military, and commercial sectors due to their reprogrammable nature. In high-safety environments, ensuring fault tolerance is crucial to improving the performance of electronic and computational systems. Common fault-tolerant methods include time redundancy, double modular redundancy, triple modular redundancy, hardware redundancy, self-checking, self-repairing, and Operand Width Aware Hardware Reuse. This paper introduces a novel approach based on error correction and detection techniques, aimed at reducing hardware complexity while optimizing for low power consumption, high speed, and minimal transistor count. The proposed technique is applicable to various arithmetic circuits. Fault-correcting and fault-detecting adders were designed using a combination of components, including a full adder (comprising one XOR gate, one NOT gate, and two 2:1 multiplexers), three XOR gates, three XNOR gates, one functional unit, two inverters, and two multiplexers (2:1 MUX). To assess the fault tolerance of the design, the technique was applied to an adder circuit, with its performance in terms of power consumption, hardware usage, energy efficiency, and delay simulated using Cadence Virtuoso @90 nm technology. Pre-layout and post-layout simulation results showed a 98% reduction in power consumption, 82% energy savings, 36.5 and 55.95% transistor savings, and a 215% reduction in area overhead compared to existing fault-tolerant adders. Additionally, multiplier designs were tested to validate the fault-correcting adder design.Keywords: Full adder, XOR gate, XNOR gate, Fault detecting adder, Fault correcting adder, Transistor overhead, Multiplexer (MUX)Subject terms: Electrical and electronic engineering, Other nanotechnology, Techniques and instrumentation, Engineering Introduction Today, advancements in integrated circuit (IC) technology have led to a rapid reduction in the number and sizes of devices, along with an increase in the density of MOSFETs. VLSI circuits that consume a significant amount of power can result in chip failures in both digital and analog systems. High-speed and low-power operations are essential features in the latest designs of various microelectronic components. The adder is a key element in this context, extensively used in compact devices such as mobile phones, remote sensors, virtual reality systems, and other applications that require minimal power dissipation. Adders are integral to numerous computational circuits designed for energy efficiency and play a vital role in the arithmetic units of devices. They are commonly employed in various multiplier designs and numerous circuits and systems1. Additionally, adders are utilized in digital signal processors, microprocessors, and mobile and portable devices for performing arithmetic operations. An adder circuit optimized for high-speed using 90 nm CMOS technology has been documented in2. However, as the transistors or components are enhanced, the likelihood of failures also increases3. Fault tolerance refers to the ability of an electronic system or network to maintain continuous operation even when one or more components fail. Fault-tolerant circuitry addresses interruptions caused by logic errors and other types of failures. The likelihood of faults occurring in specific applications, such as medical monitoring systems, defense surveillance, and space applications, can severely disrupt system functionality4. Faults are identified using the self-checking technique within the circuit. Most of the approaches reported require repeated execution of instructions for fault detection5. The self-checking technique enables high accuracy through parallel checking, utilizing additional circuitry that is more cost-effective than duplicating circuits for monitoring and allowing precise error detection6. A device achieves total self-checking (TSC) when it is self-testing and operates without faults7. This technique has been applied in various applications, including self-testing arithmetic logic units (ALUs) and self-healing networks8. Circuits and systems employing self-checking have been designed to identify different types of faults, such as transient (temporary) and permanent faults9. Currently, numerous methods have been developed for fault removal in various adder designs10–16. Other techniques for fault detection and repair include the OWHR self-repairing method and self-checking methods17. State-of-the-art research indicates that various works have been published to achieve energy-delay-product (EDP) efficient adders and multipliers18. Fault-tolerant adders may also incorporate specialized types of FETs19, quantum-dot cellular automata20,21, and reversible logic methods22. Additionally, they can be designed using parallel prefix adders23, carry-select adders24,25, and other techniques26–33. In reference34, an adder was created using either a fault-tolerant full adder (FTFA) or an exact full adder (EFA) employing 90 nm technology. This design achieved a low count of Look-Up Tables (LUTs), but power consumption still fell within the milliwatt (mW) range. It also attained minimal power consumption in the nanowatt (nW) range and occupied an area of 13 µm2; however, the reduction in power-delay product (PDP) compared to other adders was minimal. In reference35, the Optimized Fault-Tolerant Adder (OFTA), Selector-Based Fault-Tolerant Adder-II (SBFTA-II), and Selector-Based Fault-Tolerant Adder-I (SBFTA-I) were developed, resulting in reduced power and delay, though they required a larger area. Self-checking adders have been utilized to create fault-tolerant adders designed for applications that demand minimal area and operate at low voltage using nano meter technology36–38. This paper presents a high-speed, fault-tolerant full adder designed with a straightforward architecture using 90 nm CMOS technology, which can be further scaled. The adder is initially constructed using a multiplexer and an inverter. Subsequently, the proposed fault-detecting and fault-correcting adders are designed to detect and repair errors. The proposed adder employs a 2:1 MUX and a NOT gate, focusing on minimizing the transistor count. Fault detection in the adder is managed by a fault-detecting block, while corrections are made by a fault-correcting block. The fault-detecting block features XNOR gates, a functional unit, and XOR gates, utilizing fewer transistors and operating at low power and voltage. The paper is organized into several sections: Section “State-of-art techniques” elaborates on state-of-the-art fault-detecting and fault-correcting techniques, Section “Proposed fault-tolerant adder design” discusses the proposed fault-detecting and fault-correcting adder designs, Section “RC model for full adder” analyzes the output of the proposed design along with tables, graphs, and comparisons to relevant state-of-the-art adder designs, and Section “Result and discussion” concludes the discussion on the proposed fault-correcting and fault-detecting adder designs. State-of-art techniques State-of-art fault-tolerant adder designs are reported in4,5,17. They have been designed using several techniques. Most designs have used some type of redundancy: (a) Time redundancy, (b) Double modular redundancy, (c) triple-modular redundancy, (d) hardware redundancy, (e) Operand width aware hardware reuse (OWHR). State-of-art and adder design techniques are been based on these types of redundancy. Some adders are able to detect faults and some have ability to correct the fault, while others can do both. Four fault correction and detection adder techniques are explained below. Fault detection techniques Self-checking adder using XNOR, full adder and Boolean expression The self-checking adder was developed by Vasudevan17. This design distinguishes between fault-free conditions and stuck-at faults in the adder circuit. The fault-checking system includes XOR gates, a fault-repairing multiplexer, two adders, and two rail checkers. Figure 1a illustrates the self-checking adder from17. Z1 and Z2 are the checker outputs, and if they are the same, the circuit is fault-free. For a fault-free case, both Z1 and Z2 are either zero or both are one. If Z1 is low and Z2 is high, or vice versa, the circuit is also considered fault-free. While this design can detect single-net faults, it cannot pinpoint the fault location in the adder. Fault repair is also not feasible, as the entire unit must be replaced. Additional limitations include its inability to detect double faults or faulty carry propagation, and the design is hardware-expensive. (a) Self-checking adder using XNOR, full adder and Boolean expression (b) Self-checking full adder using XNOR, full adder and function unit; (c) self-repairing (SR) adder using full adder and Multiplexers; (d) self-repairing adder using self-checking full adder, multiplexers and NOT gates. Self-checking full adder using XNOR, full adder and function unit The self-checking full adder circuit is designed to detect single or multiple faults and accurately identify their location. Both the carry select adder (CSA) and ripple carry adder (RCA) have been implemented. The CSA is the fastest adder and requires fewer transistors compared to other designs, while the RCA is simpler to implement but requires more time for computation. In this design, the carry and sum outputs are repeatedly measured and checked to detect any faults in the outputs. The detailed diagram of the self-checking (SC) adder design is shown in Fig. 1b. The fault-checking circuit equations for the adder are provided. Fault-checking in sum output of adder, Fault-checking in carrying output of adder, Equivalent functional units (FU) and XNOR gates are employed for eliminating the faulty situation at output carry. XNOR1 and output of FU have furnished inputs to XNOR-5 to discover the faulty condition. Fault correction techniques Self-repairing (SR) adder using full adder and multiplexers The SR adder, reported by Muhammad Ali Akbar5, detects faults and identifies the location of single and multiple faults. Some of the limitations from a previous adder design17 have been addressed. This design includes a rail checker, a full adder, and two XNOR gates for fault detection. It uses two circuits of adders to correct faults in the SR adder. The XNOR gate (X1) compares the outputs of the carry and sum of the full adder. When all the inputs are the same, the carry and sum outputs match. If one input differs from the others, the carry and sum outputs will be different. The second XNOR gate (X2) ensures fault-free operation. The fault in the adder design is represented by Ef, as shown in Fig. 1a. The fault-detecting adder can detect two or more errors, while the fault-correcting adder identifies these errors and requires replacing the faulty full adder circuit with other full adders, as shown in Fig. 1c. This design requires additional full adders to achieve fault-free operation. However, it cannot detect faults when two faults occur simultaneously, leading the fault-correcting adder to show a fault-free output (Ef) despite the presence of errors. The SR adder5 is insufficient in eliminating these faults, making the design less reliable. The occurrence of double faults is a key limitation of this design. Self-repairing adder using self-checking full adder, multiplexers and NOT gates The Fault-Tolerant (FT) adder is used to restore fault-free operation, detecting faults through the self-checking technique of the carry select adder (CSA). The FT adder design ensures the removal of various faults, making the design fault-free. Compared to other approaches, the self-repairing adder design requires minimal area17. This method operates on the following principle: the carry and sum outputs are either one or zero, depending on the inputs to the adder schematic. If Fs equals one, a fault has occurred in the adder design, and the fault is corrected using a multiplexer, which selects the inverted sum output. Similarly, when Fc equals one, a fault has affected the carry output, and the multiplexer selects the inverted carry output. The equations for this design are Similarly, The operation of the advanced adder design relies on the outputs Fs and Fc given by the SC adder cell. If Fs is zero, then it shows a fault-free condition in sum output. The ultimate Sum output is selected by MUX. Multiplexers are designed with transmission gates. If Fc is zero, then it tells that there is a fault-free condition achieved in the carry-out of the adder. The 2:1 multiplexer, NOT gate of adder design is utilized for reverse carry, fault-free output. The last output is obtained as displayed in Fig. 1d. In4, a self-checking adder from Fig. 1b is utilized as shown in Fig. 1d. In4, if there is a fault in the adder circuit’s output, such as when the carry output for a 1-bit full adder should be 00010111 for inputs A = 00001111, B = 00110011, and Cin = 01010101, but instead we get 00101011 due to faults, it indicates a fault in the carry output while the sum remains unaffected. For inputs A, B, and Cin like 010, 011, 100, and 101, the sum is correct, but the carry is faulty. According to Table 1 and Fig. 1b, it is observed that Fc = 11000011 and Fs = 10000001 for the faulty carry output Cout = 00101011. However, this contradicts the condition where Fc = 1 or Fs = 1 indicates a fault. The proposed FC adder design overcomes this limitation. Table 1Truth table for fault tolerant adder4.ABCinSumCoutG1G2G3F1FcFs0000011111100110011010010110001000110000010010011000100101000001001100101101011111111111 Truth table for fault tolerant adder4. Proposed fault-tolerant adder design Adder design Adder is a fundamental portion of electronic systems. Its primary aim is to add two n-bit numbers and input carry. For e.g. An … A3 A2 A1 + Bn…B3 B2 B1 = C0 Sn … S3 S2 S1. For example, a 1-bit adder has inputs A, B, and a carry input C, with SUM and CARRY as outputs. Typically, adder circuits use XOR gates for the sum and AND-OR gates for the carry. In this adder design, one 2:1 MUX takes inputs B and B′ (NOT of B), while another 2:1 MUX for the SUM takes inputs A and A′' (NOT of A). The 2:1 MUX for the carry takes inputs A and B. The output from one 2:1 MUX is used as the select line input for both the SUM and CARRY 2:1 MUXes. The result from the SUM 2:1 MUX is the final sum. For instance, if A = 1, B = 1, and C = 0, the output of the SUM 2:1 MUX is B or 1. The select line input for the CARRY 2:1 MUX is high (1). When the output of the SUM 2:1 MUX is low (0), the output of the CARRY 2:1 MUX is high (1). This process is how all the outputs of the adder are obtained. The adder schematic is shown in Fig. 2a. The W/L ratios for the N-MOSFETs and P-MOSFETs are 0.180/0.09 µm and 0.360/0.09 µm, respectively. (a) Schematic of full adder; (b) blocks of proposed adder with 1-bit adder, fault detecting and fault correcting adder. Proposed fault detecting adder The proposed fault-correcting adder circuit incorporates a self-checking technique that allows it to detect both single and double faults from the outputs. XNOR, XOR gates, and functional circuitry are used to identify errors at the output ports. The red line highlights the differences observed with the FD adder design in4,17. The outputs of the functional unit’s [F4] and [F3] XOR2 gates are compared using the XOR3 gate, which identifies faults in the carry output, as shown in the Fault Detecting block of Fig. 2b. The output of XOR3, If FC is zero, it denotes the error-free/ faults-less condition. If FC is 1, it shows output Carry of a full adder is faulty. Similarly, XNOR1 output F1 and XNOR2 output F2 has compared using the XOR1 gate to detect the fault in Sum output as displayed in Fig. 2b. And we obtained output FS from XOR1 gate such as If output FS is zero, then error-free condition is obtained. If FS is one, it exhibits the ‘Sum’ of adder design is faulty. Proposed fault-correcting (FC) adder Additional circuitry is used in the fault-detecting adder to correct faults occurring in the circuit. The fault-tolerant adder reported in17 was designed solely with XNOR gates, while the proposed FC/FD adder is built using a combination of XOR and XNOR gates. The functional unit in this design utilizes fewer transistors than4, and the functional unit produces different outputs as proposed in this paper. The red line illustrates the differences between the FC adder designs from4,17. Two 2:1 MUXs are employed in the fault-correcting adder block—one for the sum and another for the carry. The output from the fault-detecting adder serves as the select line input for the fault-correcting adder. In a fault-free condition (FC = 0), the MUX selects the carry output; if a fault occurs (FC = 1), the MUX selects the inverted carry. Similarly, for the sum, FS = 0 (fault-free) selects the sum, while FS = 1 (faulty) selects the inverted sum. The outputs, SumF and CarryF, show the auto-corrected values even after faults affect the sum and carry terminals of the adder. Figure 2b displays the block diagram of the proposed fault-correcting adder. For instance, if the carry output for a 1-bit full adder is 00010111 with inputs A = 00001111, B = 00110011, and C = 01010101, but due to a fault, the carry output is 00101011, the self-detecting adder design can detect the fault, and the self-correcting adder design can correct it. The values of Fc and Fs indicate the specific fault in the adder output. The self-correcting adder block successfully removes the fault, as shown in Table 2. In the table, red-colored 0 s or 1 s indicate faults in the carry, while blue-colored 0 s or 1 s indicate faults in the sum. The layout designs are presented in Fig. 3, with Fig. 3a showing the full adder, Fig. 3b the fault-detecting adder, and Fig. 3c the fault-correcting adder. The layout for the proposed adder, featuring both fault-detecting and fault-correcting capabilities, is shown in Fig. 3d. Table 2Fault-tolerant adder table with self-correcting technique.ABCSumCarryFcFsCarryF = Fc ⨁ carrySumF = Fs ⨁sum000000000001011101010100001011010010100100001101010010110111010111110011 Fault-tolerant adder table with self-correcting technique. (a) 1-bit full adder; (b) proposed fault detecting adder block by layout; (c) layout of fault correcting block; (d) fault adder with adder, FD and FC adders. RC model for full adder The full adder design is asymmetrical, with the gate capacitance being directly proportional to the transistor’s width. The adder is constructed using basic digital logic components like XNOR, NOT, and MUX gates. Capacitance within the adder can be calculated using the internal nodes (X and Y). By reducing the width-to-length ratio (W/L) of the transistors, the capacitance load at these nodes is minimized, which in turn decreases the delay for various input values. The capacitance values at internal nodes X and Y are The capacitance values at output node (S) sum and (C) carry are The RC model of the full adder is shown in Fig. 4. This full adder circuit has been used to design fault-correcting adder because of its advantages of low delay, power, and PDP values. Total delay in full adder at different input combinations for worst case are RC model for different input cases of full adder (a) ABC = 000 (b) ABC = 001 (c) ABC = 010 (d) ABC = 011 (e) ABC = 100 (f) ABC = 101 (g) ABC = 110 (h) ABC = 111. Result and discussion The schematic, layout, and simulations for all adder designs were performed using SPICE with 90 nm CMOS technology. Figure 5a displays the input and output waveforms of the fault-free adder design, while Fig. 5b presents the output waveform for the fault-detecting adder block. Figure 5c illustrates the output waveform for the fault-correcting adder block. Similarly, Fig. 6a shows the input and output waveforms of the faulty adder design. Figure 6b depicts the output for the fault-detecting adder block, and Fig. 6c presents the output waveform for the fault-correcting adder block. Waveforms for (a) full adder (b) proposed FD adder (c) proposed FC adder when no error. Waveforms for (a) full adder (b) proposed FD adder (c) proposed FC adder when error introduced. The output waveforms of Sum and SumF are identical, indicating fault-free conditions, while the output waveforms of Carry and CarryF also confirm that the circuit is fault-free. In the presence of a fault, the outputs differ, and either FC = 1 or FS = 1. Table 2 provides the time delay, power, and energy measurements for various components, including XOR, XNOR, a functional unit (FU), (P) adder, the proposed fault-detecting (FD) adder, and the proposed fault-correcting (FC) adder, all measured at VDD = 0.8 V. Circuit performance comparisons are illustrated in Fig. 7a. Figure 7b,c compare the proposed fault-correcting (FC) adder with the previous self-repairing adder and fault-tolerant adder designs, respectively. Figure 7d presents a comparison of adder designs with state-of-the-art fault-tolerant adders across various parameters. The number of transistors and transistor overhead metrics are used for comparison with previous designs4,17. Figure 8b,c depict the power consumption and energy (EDP) of multiplier designs using the proposed 1-bit adder. Tables 3, 4 and 5 provide comparative results. Fault-tolerant adders are typically simulated using 90 nm, 180 nm, or 250 nm technology, but this paper utilizes 90 nm technology to design the fault-detecting and fault-correcting adders. This technology reduces power consumption, minimizes delay, decreases power delay product/energy, and lowers area overhead and transistor count compared to designs in4,17. The use of 90 nm technology offers several scaling advantages. Comparison of (a) power, delay and energy of proposed designs (b) number of transistors (c) transistor overhead (d) adder designs with fault tolerant adder designs4 using different parameters. (a) 4 bit multiplier design using 1 bit fault correcting adder4; comparison of 4-bit multiplier using (b) power and (c) energy parameters. Table 3Power, delay and energy of proposed designs.Time delay (in ns)Power (in nW)Energy (in aJ)XOR9.1692.0819.071XNOR0.0030.3513.511FU20.0083.67173.449Adder19.2422.96749.394FD adder14.111.36819.302FC adder11.3184.01445.43 Power, delay and energy of proposed designs. Table 4Table for number of transistors.Number of transistors comparisonSelf repairing adder17Fault tolerant adder4Proposed FC adderInverter2222:1 MUX442XOR664XNOR664Functional unit(FU)12127Adder282810Proposed FD adder807047Proposed FC adder1128257 Table for number of transistors. Table 5Table for transistor overhead measurement.1 BIT adder (number of transistors)Total no. of transistorsTransistor overheadArea overhead (in %)Proposed FC adder105747185Self-repairing adder172811284300Fault tolerant adder4288254192.85736206242210255611256300 Table for transistor overhead measurement. In this design, inverter MOSFETs and the body terminals of MOSFETs are connected to VDD or ground (GND) terminals. The proposed FC and FD adder designs have fewer connections to VDD and GND compared to4,17, making them less vulnerable to voltage fluctuations. As a result, the proposed FC and FD adders are highly reliable. The full adder is implemented with a reduced number of transistors (57 transistors in the proposed design) compared to the 82-transistor design reported in4, as shown in Tables 4 and 5. This reduction in transistor count contributes to power savings in the proposed FC adder by eliminating unnecessary switching, unlike state-of-the-art designs. The design in4, implemented at 55 nm, experiences significant leakage power, while the proposed design at 90 nm does not suffer from the same level of leakage. The reduction in leakage power dissipation and transistor counts together result in substantial power savings in the proposed FC and FD adders. Adder and multiplier designs from4 were re-evaluated at 90 nm in the current evaluation environment, and the results were compared with the proposed adder and multiplier, as shown in Figs. 7 and 8. The MUX-based adder in the proposed design is least affected by voltage fluctuations, as all connections to VDD/GND are routed through inverters. These inverters are designed to provide noise margin and other characteristics that make the circuit nearly immune to voltage fluctuations, whereas state-of-the-art designs, which directly connect to VDD/GND, are more susceptible to voltage variations. Table 6 presents a comparison between the proposed adder designs and state-of-the-art adder designs. A fault-tolerant adder has been introduced at the transistor level, incorporating fault-detection and fault-correction mechanisms. The full adder design achieves at least 66% power savings and reduces the power-delay product (PDP) by approximately 50–99%, with a maximum reduction of 70% in transistor count compared to designs referenced in30–33,37,38. The proposed fault-detecting adder reduces the transistor count by 30–49.107% compared to related adders in17,36. The fault-correcting (FC) adder shows a 30.5% reduction in transistor count compared to fault-tolerant adders in4,36. In terms of transistor overhead, there is a 55.95% reduction compared to the self-repairing adder in17 and a 31.481% reduction compared to the fault-tolerant adder in4. Table 6Power, delay and energy of proposed designs w.r.t state-of-the-art adders.1-bit full adderTime delay (in ns)Power (in nW)Energy (in aJ)Transistor countsTechnology (nm)Proposed full adder19.2422.96749.39410902021300.4563078140335452018310.04362502730306520203261.133332034610180202133–1471–990202037–5162–16–2023380.101990100.4910–2017221.728.735189.721016Fault detecting/fault correcting adderTime delay (in ns)Power (in nW)Energy (in aJ)Voltage (in volt)Technology (nm)Proposed FD adder14.111.36819.3020.890Proposed FC adder11.3184.01445.430.8902019341.811789032390–902018391.7921300381500.990SBFTA-II35 20221.091583017320190SBFTA-I35 20221.651241020460190FTA-I36 20183.95882503485871.0528FTA-II36 20184.71819019895791.265OFTA35 20221.751284022450190 Power, delay and energy of proposed designs w.r.t state-of-the-art adders. Regarding area overhead, a 215% reduction is achieved compared to the self-repairing adder in17 and a 7.85% reduction compared to the fault-tolerant adder in 4. The proposed design also provides substantial power savings over existing state-of-the-art designs. When compared to conventional adder designs4,17,36, the proposed fault-correcting (FC) and fault-detecting (FD) adders show a transistor overhead reduction of 31.481–55.95%. In terms of power dissipation and energy efficiency, the proposed design achieves power savings ranging from 66 to 99%, with an energy reduction of approximately 82.678–100% compared to existing designs4,34–36,39. Fault correcting multiplier design The fault-correcting adder design has been used to develop a fault-correcting multiplier. The proposed fault-correcting (FC) and fault-detecting (FD) adder designs are illustrated in Figs. 2 and 3, while the output waveforms of the adder circuits are shown in Figs. 5 and 6. Simulations using 90 nm technology at 0.8 V were conducted to analyze the performance of the fault-correcting multiplier. Power consumption and energy (EDP) were estimated using Spice simulation. Figure 8a presents the block diagram of the multiplier design, and Fig. 8b,c provide a graphical comparison of the multiplier design. A mathematical comparison of power consumption and power-delay product (PDP) is provided in Table 7. Table 7Table for power and energy comparisons.Power (in µW)Energy (in aJ)Proposed 4-bit multiplier1.26712163.23Fault tolerant multiplier494.3856467.554292.6139262305 Table for power and energy comparisons. Conclusion This work presents a highly reliable, fault-tolerant, and power-efficient design for a full adder. The full adder and fault-correcting block are designed using a 2 × 1 multiplexer (MUX) and a NOT gate with a minimal number of transistors. The fault-detection block incorporates XNOR gates, a functional unit, and XOR gates to ensure low power dissipation. The proposed fault-detecting and fault-correcting adder minimizes the transistor count, enabling the adder circuit to detect and correct single and multiple faults, effectively making it fault-free. This design significantly reduces area overhead, transistor overhead, and other parameters. The proposed fault-detecting adder shows a notable decrease in the number of transistors. The fault-correcting (FC) adder was also evaluated and demonstrated a substantial reduction in transistor count compared to state-of-the-art designs. In terms of transistor overhead, an approximate reduction of 50% was achieved. When comparing area overhead, the proposed design offers double the area savings compared to other adder designs. Additionally, massive power savings were observed, with the proposed design showing a 100% reduction in power consumption and a significant decrease in energy consumption compared to existing fault-tolerant designs. Acknowledgements I am very grateful for resources and support from USICT, Guru Gobind Singh Indraprastha University. Author contributions D.P.: Conceptualization, Methodology, Software, Data curation, Writing—Original draft preparation, Visualization, Writing—Reviewing and Editing, Investigation, Validation. Data availability All data generated or analyzed during this study are included in this published article. Declarations Competing interests The authors declare no competing interests. Footnotes The original online version of this Article was revised: In the original version of this Article Pritty was incorrectly affiliated with ‘USIC&T, Guru Gobind Singh Indraprastha University, New Delhi, 110075, India’. The correct affiliation is: NIMS institute of Engineering and Technology (NIET), NIMS University Rajasthan, Jaipur, India, 303121. Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Change history 3/13/2025 A Correction to this paper has been published: 10.1038/s41598-025-93099-x References Associated Data Data Availability Statement All data generated or analyzed during this study are included in this published article.