Turkish Journal of Computer and Mathematics Education Vol.12 No.10 (2021), 2610-2620
Research Article
2610
FinFET based Add and Shift Multiplier for Wireless Communication
Alluri Navaneetha
1, Kalagadda Bikshalu
21Mahatma Gandhi Institute of Technology, Hyderabad-500075, India. 2KU College of Engineering and Technology, Warangal-506009, India. 1neethaalluri@gmail.com, 2kalagaddaashu@gmail.com
Article History: Received: 10 January 2021; Revised: 12 February 2021; Accepted: 27 March 2021; Published
online: 28 April 2021
Abstract :- In this paper FinFET based low power multiplier for IOT applications is proposed. At present society is need of
low power IOT devices such as Bluetooth, Wi-Fi, RFID and Zig-Bee to connect the things to real world. FinFET BSIM CMG model files for the feature size 7 nm, 10 nm, 16 nm and 20 nm are used to analyze FinFET based conventional multiplier using licensed SymicaDE software tool. Static and dynamic power dissipation is investigated for FinFET based conventional multiplier. As the technology node decreases simulation results proved that power dissipation of multiplier circuit reduces and its value is less compared to MOSFET based multiplier circuits, FinFET based Array multiplier, Vedic multiplier circuits. The proposed FinFET multiplier can be used for low power IOT devices such as wireless sensor networks and RFIDs which operates with low power batteries with more battery life.
Keywords: Low Power (LP),Internet of things(IOT),FinFET,Metal oxide semiconductor Field effect transistor
(MOSFET),Complementary metal oxide semiconductor (CMOS).
1. Introduction
At present development of every section of society takes place with the usage of IOT technology is used in wide area of applications such as Embedded Systems, mobile computing, smart energy a, smart grid, smart health, smart transportation and smart environment requires devices and circuits which operate at low power (LP) design. Future technologies such as cloud computing, Fog computing, big data, distributed computing utilizing IOT technology need circuits which work on high secure, high speed, low area, low power. The high-performance circuits can be designed by using advanced devices such as FinFET compared to MOSFET. Low leakage power and miniaturization FinFET is considered as emerging transistor compared to MOSFET due to control of channel by the gates. Portability is an important aspect for IOT devices along with low battery power and high security. Portability is achieved by scaling of device. At reduced technology below 20 nm FinFET has less energy compared to other technology nodes which presented by Dinesh Kumar in research article [1].At the technology node below 20 nm FinFET emerged as leading technology as it overcomes short channel effect compared to MOSFET [2].Senthil Kumar in research article shown that and gates, Vedic multiplier designed by FinFET has low power compared to MOSFET [3].
In digital circuits such as CPU, multipliers are important component in which speed and power requirement is main concern. To reduce the power dissipation in multiplier switching activity has to reduced [4]. Many Researchers have done work to reduce the power dissipation for different multipliers [[5]-[7]]. Rajshree Shanmugam shown in research article on comparative analysis of various multipliers in that array Multiplier has more power and Baugh Woolley multiplier has low power and high speed [8] which was implemented in Xilinx ISE 8.1 for synthesis making used of VHDL programming. Large areas occupied by tree multipliers when used in high speed applications. Radix base multiplier such as carry select adder-based radix multiplier requires more transistors which consumes more power. Less area and simple in design is obtained by using multipliers based on radix and shift architecture [9]. Already authors have done research on conventional multiplier and BZFAD multiplier using 130 nm MOSFET CMOS technology [4]. The result showed BZFAD has less power compared to conventional at technology node 7 nm. Analysis of FinFET node from 20 nm to 7 nm which can be used for the design of FIR Filter for fabricating DSP processors which is used in many IOT applications such as wireless sensor networks, RFID etc
Turkish Journal of Computer and Mathematics Education Vol.12 No.10 (2021), 2610-2620
Research Article
2611
The main objective is towards low power multiplier circuits compared to the existing FinFET Array Multiplier [10], FinFET Vedic multiplier [3] and MOSFET conventional multiplier & BZFAD ((By pass zero feed A directly) [4].Heat dissipation is proportional to power. More heat dissipation results due to more power which decreases speed and reliability. So, the requirement for IOT devices such as smart environment, smart health and smart transportation invokes to make the devices for low power. M.Mottaghi-Basterji [4] presented a paper based on low power structure based on add and shift architecture, using 130 nm CMOS technology .V.M. Senthil Kumar [3]presented a paper implementation of Vedic mathematics based on discrete wavelet transform for biomedical and signal processing.MAC unit, RAM or ROM, Adder, multiplier is included in the discrete wavelet transform the existing CMOS has more power than FinFET DWT which is implemented using 32 nm BSIM Files . Anubhuthi Mittal [11] presented the design of low power and high speed 16 order FIR filter. In the 16 order FIR filter[11] add & shift Wallace tree (WT) and Vedic multiplier are used for the multiplication. Reduced area and delay is obtained by using add & shift multiplier implemented on FPGA Spartan 3 using Xilinx ISE 8.1.1.1 Organization of paper
Section 2 presents the background of FinFET device. Section 3 presents the Proposed FinFET based conventional design of multiplier which is implemented by using 7nm BSIM CMG FinFET files. Section 4 presents comparison of proposed FinFET multiplier with existing multiplier. The proposed FinFET multiplier shown low power as compared to the existing one. Section 5 presents comparison of FinFET multipliers from 20 nm to 7 nm shown the result that as the technology node decreases the static power dissipation and dynamic power dissipation decreases. Section 6 concludes the paper
2. Background of FinFET device
FinFET device is a 3D structure which has a silicon body perpendicular to the plain of wafer [1]. The gate of the FinFET is wrapped on channel. FinFET can have one, two, three and four gates wrapped on channel. Channel is controlled by gates which reduce short channel effects [12]. Compared to MOSFET, FinFET has better gate control which results faster switching speed, higher on current, lower leakage. FinFET can be operated in 2 modes a) Independent gate mode (LP) b) shorted gate mode (SG).In the Independent Gate Mode four terminals are present in FinFET. Independent gate mode also called LP mode which is used to reduce threshold voltage. In this mode back and front gate are connected to different inputs. In the Shorted Gate mode (SG) 3 terminals are present in FinFET. Front gate and back gate both are tied together. At technology node below 20 nm MOSFET suffer from short channel effects such as more leakage power, threshold variations [13].To overcome short channel effects FinFET is proposed. FinFET is formed on thin silicon insulator termed as fin [3]. Gate work function of the FinFET can be used to reduce power and adjust the threshold voltage. FinFET is emerged as promising device in literature to provide solution to the problem takingplace due to reduced technology node[14].FinFET fabrication is compatible with MOSFET for rapid deployment[15]-[16].
Figure 1. FinFET fabrication 3.FinFET based Add and Shift Multiplier
Conventional Shift-and-add multiplication is similar to the multiplication performed by paper and pencil. To multiply two N-bit numbers, the algorithm takes one bit of the multiplier at a time from right to left, multiplying the multiplicand by this single bit of the multiplier and placing the partial product to the left of the earlier results. At every clock cycle the multiplier bit is shifted to its LSR value. If the multiplier LSR [0] value is “0” the accumulator is added with MSR register and zeros then 1-bit right shifting operation is performed. If the multiplier LSR [0] value is “1” then added to the accumulator is added with MSR register and A input then 1-bit right shift operation is performed. After all the multiplier bits have been tested the product is in Partial Product Register (MSR, LSR). The conventional
Turkish Journal of Computer and Mathematics Education Vol.12 No.10 (2021), 2610-2620
Research Article
2612
multiplier consists of the following components.1. Vectored 2X1 Mux of k bit each input.
2. Ripple carry adder k-bit with k+1 bit.
3. Two special purpose shift Registers with parallel in parallel out and right shift register used to store the Product.
4. Control logic circuits has outputs such as counter output signal and shift/load signal. The counter output tells which partial product is generated for the input given to the multiplier.
Figure 2. Architecture of Conventional Shift and Add Multiplier
Table 1: Analysis of Conventional Add and Shift Multiplier by Mathematical method input A=0011,B=0011 Rese t Cloc k Shift/Loa d (1/0) Adder = MSR+A Partial Product MSR LSR 1 0 00011 0000 0000 0 0 00011 0000 0011 0 1 00110 0011 0011 0 0 00100 0001 1001 0 1 00111 0100 1001 0 0 00010 0010 0100 0 1 00010 0010 0100 0 0 00001 0001 0010 0 1 00011 0001 0010
Turkish Journal of Computer and Mathematics Education Vol.12 No.10 (2021), 2610-2620
Research Article2613
0 0 00011 0000 1001Turkish Journal of Computer and Mathematics Education Vol.12 No.10 (2021), 2610-2620
Research Article
2614
Figure 4. Simulation Results of 4bit Conventional Add and shift Multiplier for A=0011, B=0011 andOutput=00001001
Figure 5. Schematic of 4-bit FinFET Conventional Multiplier
Figure 6. Power simulation for conventional Add and shift Multiplier
4. Results and Discussion
Simulation results of FinFET based conventional multiplier using 7 nm BSIM Files [15] shown that the static power dissipation in conventional multiplier is less than existing multipliers with MOSFET at different technology nodes and its static and dynamic power dissipation decreases as technology node reduces. Table 5 data results show that comparison of static power dissipation of FinFET conventional multiplier has low power dissipation compared to the existing MOSFET conventional and FinFET array and Vedic multiplier.
Turkish Journal of Computer and Mathematics Education Vol.12 No.10 (2021), 2610-2620
Research Article2615
Technology 20 nm 16 nm 10 nm 7 nm Supply Voltage (V) 0.9 0.85 0.75 0.7 Conventiona l 325 187.4 85.07 1 3.90Table 3 : Dynamic Power (µW) Comparison
Technology 20 nm 16 nm 10 nm 7 nm Supply Voltage (V) 0.9 0.85 0.75 0.7 Convention al 1836.8 9 1278. 65 1125. 8 815.3 6 Table 4: Transistor count and clock cycles of Proposed multiplier
Figure 7. Comparison of static power dissipation of FinFET conventional multiplier at different technology nodes. Dynamic Power (µW) Technology(nm) 10 16 20 500 815.36 1000 1125.8 1278.65 1500 1836.89 2000
Dynamic Power
7 3.9 10 16 20COMPARSION OF STATIC POWER
Technology(nm) Static Power (µW)
85.07 187.4
Turkish Journal of Computer and Mathematics Education Vol.12 No.10 (2021), 2610-2620
Research Article
2616
Figure 8. comparison of dynamic power dissipation of FinFET conventional multiplier at different technology nodes.Parameter Transistor count Clock Cycles /
Multiplication
Turkish Journal of Computer and Mathematics Education Vol.12 No.10 (2021), 2610-2620
Research Article
2617
Table 5: Comparison of static power dissipation of Existing multipliers and Proposed 7nm FinFETConventional multiplier
Multiplier Width Technology Static Power
(µW) Ref. Paper No. Publis hed Year
Conventional 16-bit 0.13 µm (MOS) 10.824 [4] 2008
BZFAD 16-bit 0.13 µm (MOS) 7.576 [4] 2008
VEDIC 4-bit 32 nm (FinFET) 175.68 [3] 2019
Designed Conventional
4-bit 7 nm (FinFET) 3.90
Figure 9. Comparison of static power of Existing multiplier and proposed 7nm FinFET conventional multiplier
Figure 10. Transistors count and clock cycles per multiplication of 7nm FinFET conventional multiplier
Comparsion of static power of existing and
designed multiplier
175.78 200 150 100 50 10.82 7.57 3.9 CLOCK CYCLES PER MULTIPLICATIO TRANSISTOR COUNT 150 0 100 0 202 2 250 0 200FinFET based conventional
Multiplier
Turkish Journal of Computer and Mathematics Education Vol.12 No.10 (2021), 2610-2620
Research Article
2618
Conclusion
The Multipliers are extensively used in Digital Signal Processing (DSP). With the advancement in VLSI technology as the DSP has become increasingly popular over the year, the high speed realization of multiplier with less power consumption has become much more demanding. The operation of FIR filter is mainly depended only on the multiplier design. Implemented Conventional multiplier using FinFET 7 nm, 10 nm, 16 nm and 20 nm technologies accomplished by using BSIMCMG files and licensed SymicaDE Tool. Simulated multiplier shown the results that the FinFET based conventional multiplier has low power compared to the mosfet based conventional and BZFAD multiplier. The power dissipation, of FinFET based conventional multiplier is found less than the FinFET based Vedic and array multiplier. By observing these comparisons, we can state that conventional multiplier reduces the power dissipation. To improve the multiplier performance the FinFET based conventional multiplier is best compared with mosfet based other multipliers for implementation.
Acknowledgement
A part of research work is contributed by K. Murali Mohan undergraduate student of Mahatma Gandhi Institute of Technology. The authors would also like to thank Predictive technology model research team at Arizona state University and Symica software Company for providing licensed SymicaDE software tool.
References
1. S. Dinesh Kumar, Himanshu Thapliyal, Azhar Mohammad “Fin SAL: FinFET-Based Secure Adiabatic logic for Energy-Efficient and Resistant IOT Device” IEEE Transactions on computer-Aided design of integrated circuits and systems, vol.37,N0.1,January 2018 pp.110-
2. 111.
3. Joseph Whitehouse and Eugene John “Leakage and Delay Analysis in FinFET Array Multiplier Circuits” IEEE 978-1-4799-4132- 2/14/ pp.909-911
4. V.M. Senthil Kumar, S. Ravindra Kumar, D. Nithya, N.V. Kousik “A Vedic mathematics-based processor core for discrete wavelet transform using FinFET and CNTFET technology for biomedical signal processing” Microelectronics Journal 20 August 2019 pp.2-4.
5. M. Mottaghi-Dastjerdi, A. Afzali-Kusha, and M. Pedram “BZ- FAD: A Low- Power Low-Area Multiplier based on Shift-and- Add Architecture” IEEE Trans. on VLSI Systems,
6. vol.17,N0.2,February 2009 pp.302-305.
7. Chandrakasan and R. Brodersen, “Low-power CMOS digital design,” IEEE J. Solid-State Circuits, vol. 27, no. 4, pp. 473–484 Apr. 1992.
8. N.YShenaid O.T.-Chen “Low power multipliers by minimizing switching activities of partial products”,in Proc.IEEE Int .Sysmn
9. .Circuits Syst,May 2002,Vol.4,pp.93-96.
10. O. T. Chen, S. Wang, and Y.-W. Wu, “Minimization of switching activities of partial products for designing low-power multipliers,” IEEE Trans.Very Large Scale Integr.(VLSI) Syst.,vol.11,no.3,pp.418–
11.
433, Jun. 200312. Rajshree Shanmugaratnam, Kathrivel Brindhadevi “Comparative Analysis of various types of Multipliers for effective lowpower” Microelectronics Journal vol.214, June 2019, Pages 28-37
13. D.Hismotoetal.,” FinFET a self-aligned double-gate MOSFET scalable to 20 nm”, IEEE Trans.Electronic Devices, vol.47, no.12, pp2320-2325, Dec,2000
14. Z. Weinman , G, F. Jerry , M. Leo , D. Yang , Physical insights regarding design and performance of independent gate FinFETs,IEEE Trans. Electron Dev. 52 (10) (2005) 2198–2206
15. T. Rudenko , V. Kilchytska , N. Collaert , A. Nazarov , M. Jurczak , VM. Senthil ku-mar , S. Saravanan , Design of a reduced carry chain propagation adder using FinFET, Asian J. Inf. Technol. 15
16. (11) (2016) 1670–1677
17. M. Prateek, M. Anish, K, J. Niraj, in: Gate Sizing: FinFET’s vs 32 nm
18. Bulk MOSFETs, Nano electronic Circuit Design, 2011, pp. 23–54, doi: 10.1007/ 9781441976093 _ 2
19. Yogesh Singh Chauhan, Darsen D. Lu, Sriram Kumar Venugopal FinFET Modeling for IC Simulation and Design Using the BSIM-CMG ,Academy press,2015.
Turkish Journal of Computer and Mathematics Education Vol.12 No.10 (2021), 2610-2620
Research Article
2619
multiplier In 32nm Technology, Springer-International Conference on Soft Computing and Signal 21. Processing, 2018.Turkish Journal of Computer and Mathematics Education Vol.12 No.10 (2021), 2610-2620
Research Article