A Novel Design of a 4 Bit Reversible ALU using Kogge-Stone Adder

ISSN No: 2456 -
International
Research
A Novel Design of a four bit Reversible ALU using
an emphasized Carry Look
Swetha Potharla
M.Tech , Dept. of ECE, CITS Warangal.
ABSTRACT
Reversible circuits are one promising direction
with applications in the field of low
design or quantum computation. However, no
real design flow for this new kind of circuits exists
so far. Significant contributions have been made
in the literature towards the design of reversible
logic gate structures and arithmetic units,
however, there are not many efforts directed
towards the design of reversible ALUs. In this
paper, a novel programmable reversible Kogge
Stone adder is presented and verified, and its
implementation in the design of a
Arithmetic Logic Unit is demonstrated. Then,
reversible implementations of ripple
Kogge-Stone carry look-ahead adders are
analyzed and compared in terms of delay. The
proposed design consists of the reversible Fredkin,
Feynman, MG, HNG, PG and RKSC gates. The
performance characteristics analysis is carried out
in Xilinx environment.
Keywords: Arithmetic Logic Unit, Carry Look
Ahead Adder, Emerging Technologies, Low
Power, Nanotechnology, Reversible Logic,
Ripple-Carry Adder
I. INTRODUCTION
In reversible logic there exists a one to on
mapping between the inputs and the outputs
vectors. In an irreversible circuit erasing a bit is
equivalent to dissipation of kTln2 joules of heat
energy where k is the Boltzmann’s constant and T
is the absolute temperature of environment which
is demonstrated by Landauer [1]. This resulting
dissipated heat also causes noise in the remaining
circuitry, which results in computing errors.
Bennett showed that the dissipated energy directly
correlated to the number of lost bits, and that
computers can be logically reversible, maintain
- 6470 | www.ijtsrd.com | Volume -
International Journal of Trend in Scientific
Research and Development (IJTSRD)
International Open Access Journal
an emphasized Carry Look-ahead Adder
CITS Warangal.
Subhash. R R. Rajkumar
Asst. Prof., Warangal , Telangana
ing direction
of low-power
design or quantum computation. However, no
real design flow for this new kind of circuits exists
so far. Significant contributions have been made
in the literature towards the design of reversible
logic gate structures and arithmetic units,
er, there are not many efforts directed
towards the design of reversible ALUs. In this
paper, a novel programmable reversible Kogge-
Stone adder is presented and verified, and its
lementation in the design of a reversible
trated. Then,
reversible implementations of ripple-carry,
ahead adders are
analyzed and compared in terms of delay. The
proposed design consists of the reversible Fredkin,
Feynman, MG, HNG, PG and RKSC gates. The
ristics analysis is carried out
Arithmetic Logic Unit, Carry Look-
Ahead Adder, Emerging Technologies, Low
Power, Nanotechnology, Reversible Logic,
one to one
mapping between the inputs and the outputs
vectors. In an irreversible circuit erasing a bit is
equivalent to dissipation of kTln2 joules of heat
energy where k is the Boltzmann’s constant and T
is the absolute temperature of environment which
rated by Landauer [1]. This resulting
dissipated heat also causes noise in the remaining
circuitry, which results in computing errors.
Bennett showed that the dissipated energy directly
correlated to the number of lost bits, and that
be logically reversible, maintain
their simplicity and provide accurate calculations
at practical speeds [2]. Resultantly, a new paradigm
in computer design arose with the goal of reducing
the entropy increase and subsequent energy
dissipation. Such a logical structure must possess the
same number of inputs and outputs and a one
mapping between the input and output states. Any
device designed to these constraints is known as a
reversible logic device.
Section II, we outline the goals of programmable
reversible logic design and some
gates. In Section III, a novel 5*5 programmable MG
gate is proposed that may be utilized in an
arithmetic logic unit requiring the calculation of
AND, NAND, OR, NOR, XOR and XNOR results.
In Section IV, previous work in prog
reversible arithmetic logic units is reviewed, and
the proposed MG gate is implemented in the design
of a novel programmable arithmetic logic unit.
adder designs are presented for ripple kogge
adder.
In Section VI, a reversible implementation of the
carry logic presented in the Kogge
and verified. In section VII, a 4
implemented with ripple carry and kogge
adders and these designs are functionally verified
and compared in terms of delay.
II. REVERSIBLE LOGIC
A. Programmable Reversible Design Goals
The three major design goals of reversible logic are
as follows. First, minimization of the quantum
cost – the number of 1*1 and 2*2 reversible
calculations necessary to generate the logical
output [6] will reduce the device’s computational
complexity. Second, minimization of the delay
the logical depth of the device [7] will improve the
- 1 | Issue – 6
Scientific
(IJTSRD)
International Open Access Journal
dder
R. Rajkumar
Warangal , Telangana, India.
simplicity and provide accurate calculations
at practical speeds [2]. Resultantly, a new paradigm
in computer design arose with the goal of reducing
the entropy increase and subsequent energy
dissipation. Such a logical structure must possess the
me number of inputs and outputs and a one-to one
mapping between the input and output states. Any
device designed to these constraints is known as a
Section II, we outline the goals of programmable
reversible logic design and some fundamental logic
gates. In Section III, a novel 5*5 programmable MG
gate is proposed that may be utilized in an
arithmetic logic unit requiring the calculation of
AND, NAND, OR, NOR, XOR and XNOR results.
In Section IV, previous work in programmable
reversible arithmetic logic units is reviewed, and
the proposed MG gate is implemented in the design
of a novel programmable arithmetic logic unit. In
presented for ripple kogge-stone
implementation of the
carry logic presented in the Kogge-Stone is proposed
and verified. In section VII, a 4-bit ALU is
implemented with ripple carry and kogge-stone
adders and these designs are functionally verified
A. Programmable Reversible Design Goals
The three major design goals of reversible logic are
as follows. First, minimization of the quantum
the number of 1*1 and 2*2 reversible
calculations necessary to generate the logical
output [6] will reduce the device’s computational
minimization of the delay -
the logical depth of the device [7] will improve the

International Journal of Trend in Scientific Research and Development (IJTSRD) ISSN: 2456-6470
ancillary inputs and garbage outputs - inputs
and outputs not implemented in the design of the
gate and only serve to maintain reversibility of the
device will improve the design space require to
implement the logic.
A programmable reversible logic gate is defined
in [8] as a logic structure which possesses a
bijection between input and output states and an
equal number of inputs and outputs wherein a
subset of the inputs are fixed select lines, and a
fixed subset of the outputs produce guaranteed
logical calculations. An ideal programmable
reversible logic gate with j inputs and outputs has
a quantity of fixed select inputs m, fixed select
outputs n, data inputs d and propagated outputs p
such that |d-p|=|m-n| [8]. In addition, an ideal
programmable reversible logic gate with m select
inputs may produce at maximum n*2^m logical
calculations on the n logical outputs.
B. Fundamental Logic Gates
There are three types of fundamental 2*2
reversible logic gates. First, the square-root-of-not
gates utilize the unitary operators to produce
reversible logic calculations. The Controlled-V
and the Controlled-V+ gates are the two types of
square-root-of-not gates. In both of these gates,
when the control input is 0, the second input is
propagated to the output. The same holds for two
Controlled-V+ gates in series.
When a Controlled-V and Controlled-V+ gate are
activated in series, they act as an identity.
The second type of fundamental 2*2 reversible-
logic gate is the Feynman gate, or the Controlled-
Not gate. Proposed in [9] by Feynman, it is
configured such that its outputs states correlate to
the input states in the following manner: P=A and
Q=A B. The resulting value of the second output
corresponds to the result of a conventional XOR
gate. Since fanout is expressively forbidden in
reversible logic, since a fanout has one input and
two outputs, the Feynman gate may be used to
duplicate a signal when B is equal to 0. Its
quantum configuration is shown in Fig 2.
Fig 1: Quantum Representation of Feynman
gate
The third type of fundamental 2*2 reversible logic
gate is the integrated qubit gate. This gate is
implemented with a Feynman gate with either a
Controlled-V or Controlled V+ gate. The XOR
output of the Feynman gate is used as the control
signal for the Controlled-V or V+ gate it is coupled
with. The quantum cost of the integrated qubit gate
is 1 and its worst-case delay is 1. The quantum
configurations of these gates are shown below in
Fig. 3.
Fig 2: Quantum Representations of Integrated
Qubit Gates
The 3*3 Peres gate was proposed by Peres in [12].
The Peres gate has a quantum cost of 4 and a worst-
case delay of 3. The quantum representation is
shown in Fig. 6. The output states map to the inputs
in this manner: P=A, Q=A B and R=AB C.
There are three 3*3 fundamental reversible logic
gates. The first was proposed in [10] by Fredkin and
Toffoli. The Fredkin gate’s outputs states map to the
inputs as follows:
P=A, Q=A B AC and R=AB A C.
Therefore, the outputs serve as a multiplexed output
of the two data inputs based on the control input. It is
realized using 2 Feynman gates, a Controlled-V gate
and two integrated qubit gates.
Toffoli proposed the second fundamental 3*3
reversible logic gate in [11]. The output states of the
Toffoli gate map to the inputs in this manner: P=A,
Q=B and R=AB C. The quantum cost is 5 and the
worst-case delay is 5. The quantum representation is
shown below in Fig. 5.
Fig 3: Quantum Representation of Fredkin gate
Fig 4: Quantum Representation of Toffoli gate

The 3*3 Peres gate was proposed by Peres in [12].
The Peres gate has a quantum cost of 4 and a
worst-case delay of 3. The quantum representation
is shown in Fig. 6. The output states map to the
inputs in this manner: P=A, Q=A B and R=AB
C.
Fig 5: Quantum Representation of Peres gate
The HNG gate, also designed by Hagparast and
Navi and proposed in [13], is a full adder with
inputs A,B,C,D producing outputs P=A, Q=B,
R=(A B) C and S=(A B)C (AB D) and is
shown in Fig. 6.
Fig. 6: Hagparast-Navi (HNG) Gate
The Peres And-Or (PAOG) gate, proposed in [8],
produces the outputs P = A, Q =A B, R=AB
C and S=(AB C) ((A B) D). Fig. 7
shows the quantum representation of the PAOG
gate. This gate is an extension of the Peres gate
for ALU realization. When the PAOG is utilized
as a programmable reversible logic gate with two
select inputs, it will calculate four logical
calculations on those two logical outputs: AND,
NAND, NOR and OR.
The Morrison-Ranganathan (MRG) gate,
proposed in [8], has a quantum cost of 6, since it
consists of three XOR gates, 2 Controller-V and
one Controller-V+ gate. The worst-case delay of
the MRG gate is 6. The quantum representation of
the MRG gate is shown in Fig. 8 below. When the
MRG is utilized as a programmable reversible
logic gate with two select inputs, it will calculate
four logical calculations on those two logical
Fig. 8: Quantum Representation of the MRG Gate
III. PROPOSED 5*5 PROGRAMMABLE
REVERSIBLE MG GATE
Next, we propose the design of a 5*5 programmable
reversible logic gate structure utilized in the
implementation of an ALU. Fig. 9 shows the block
diagram of the MG, and the logical calculations
based on the programmable inputs are presented in
Table 2. The cost of the MG is 7, and the worst-case
delay is 7. The design for the programmable MG
was verified and simulated using VHDL in Xilinx.
Fig 9: Quantum Representation of Proposed MG.
TABLE II
Mg Programmable Inputs And Logical Outputs
C D E R S T
0 0 0 A B AB A+B
0 0 1 A B AB (A+B)'
0 1 0 A B (AB)' (A+B)'
0 1 1 A B (AB)' A+B
1 0 0 A B AB A+B
1 0 1 A B AB (A+B)'
1 1 0 A B (AB)' (A+B)'
1 1 1 A B (AB)' A+B
IV. MODIFIED ALU DESIGN WITH MG
GATE
Two 1-bit ALUs were proposed in [8]. The first
utilizes the MRG gate and HNG gate to produce six
logical calculations: ADD, SUB, XOR, XNOR, OR
and NOR. The ALU has 8 inputs and 8 outputs. The
inputs consist of three data inputs (A, B and Cin) and
five fixed input select lines. The eight outputs are: A,
S0, S3 and S4 propagated to the output, A B,
SUM, Cout, Overflow and Result. The cost of this 1-
bit ALU is 24, and the worst-case delay is 16. For n
bit ALU devices, an addition cost of 2 is incurred per
bit in order to propagate S1 and S2 to other bits.

The MG gate is utilized in the implementation of a
novel arithmetic logic unit based on those
proposed in [8]. The ALU, in addition to
producing the same logical calculations as the
MG, is able to perform addition and subtraction
by utilizing the HNG gate and store less-than
operation. The cost an n-bit ALU is 37n-3 and had
a worst case delay of 4n+13. The proposed ALU
is shown in Fig. 10
V. BASIC ADDER DESIGNS
The adders implemented are the ripple carry adder
and Kogge-Stone adder. The ripple carry adder is
one of the simplest adders. It consists of a
cascaded series of full adders. The 4-bit ALU
implemented by cascading four 1-bit ALUs. The
delay for an N-bit adder is given by
tadder = (N-1)tcarry + tsum
where, tcarry is the carry propagation delay for one
stage and tsum is the time required to compute the
sum bit for one stage.
The Kogge-Stone adder is classified as a parallel
prefix adder since the generate and the propagate
signals are pre computed. In a tree-based adder,
carries are generated in tree and fast computation
is obtained at the expense of increased area and
power. The parallel-prefix adder becomes more
favorable in terms of speed due to the O(log2n)
delay through the carry path compared to O(n) for
the RCA.
Fig.11. 4-bit Kogge-Stone Adder
The 4-bit Kogge-Stone adder is built from
generate and propagate (GP) blocks, black cell
(BC) blocks, and buffer blocks as shown in the
Figure 11. The expressions for the output signals
g,p generated by the black cell are given by g =
gi+pi gj and p = pi pj. The expressions for the
output signals g, p obtained by the buffer block
are given as g=gi and p=pi .
The expressions for the output signals g, p
obtained by the GP block are given as g=a.b and
p=a xor b.
Reversible Kogge-Stone Cumulate Logic
A reversible carry look-ahead adder based on the
Kogge-Stone adder is presented [9]. First, a RKS
Cumulate utilized in the calculation of the carry out
signal is designed and verified. The logical structure
of the RKSC is shown in Fig. 12. The cost of the
RKSC is 14 and it has a worst-case delay of 4.
VI. IMPLEMENTATION OF REVERSIBLE
KOGGE-STONEADDER
A reversible Kogge-Stone adder is implemented by
using Peres gate as GP block, Feynman gate as
Buffer block and RKSC gate as BC block as shown
in Fig.13. The design was verified and simulated
using VHDL in Xilinx 9.2i.
Fig. 12: Reversible Kogge-Stone Cumulate (RKSC)
Logical Layout
VII. IMPLEMENTATION OF REVERSIBLE
KOGGE-STONE ADDER
A reversible Kogge-Stone adder is implemented by
using Peres gate as GP block, Feynman gate as
Buffer block and RKSC gate as BC block as shown
in Fig.13. The design was verified and simulated
using VHDL in Xilinx 9.2i.
Fig 13: 4-bit Reversible Kogge-Stone adder ALU
Design with Reversible Kogge-Stone adder
The 4-bit ALU was implemented using reversible
Kogge-Stone adder as shown in Fig.14. This design
was verified and simulated using VHDL in
Xilinx9.2i.

VIII. COMPARISION AND RESULTS
The comparison is carried out in between the
reversible Ripple Carry and Kogge-Stone adders
with respect to 4-bit ALU design and for those the
code was written in VHDL and that was verified
and simulated using Xilinx 9.2i.
TABLE IV
Comparison of ALUs of ripple carry and
kogge-stone adders
Parameter 4-bit ALU
with
4-bit ALU
with
ripple carry
adder
kogge-stone
adder
Delay(ns) 16.856 15.536
LUT’s 37 31
Fig.14. 4-bit reversible ALU with kogge-stone
adder
Fig.15. 1-bit reversible ALU
Fig.16. 4-bit reversible kogge-stone adder
Fig.17. 4-bit reversible ALU with ripple carry
adder
Fig.18. 4-bit reversible ALU with kogge-stone
adder
IX. CONCLUSION
A novel 5*5 programmable MG gate was proposed
and verified that may calculate of AND, NAND,
OR, NOR, XOR and XNOR depending on the inputs
from the programmer. The proposed MG gate was
implemented in the design of a novel programmable
arithmetic logic unit. The novel 1-bit ALU required
only minimal increase in quantum cost and delay due
to the MG design, which also allowed for increased
functionality for the programmer.
Next, we presented reversible implementations of
ripple carry adder (RCA) and kogge-stone adder. A
eversible implementation of the carry logic presented
in the Kogge tone was presented and verified. The
proposed design of 4-bit ALU with RCA and Kogge-
Stone adder were compared in terms of delay.
REFERENCES
1) R. Landauer, "Irreversibility and Heat
Generation in the Computational Process," IBM
Journal of Research and Development, vol. 5,
1961, pp. 183-91.
2) C. Bennett, "Logical Reversibility of
Computation," IBM Journal of Research and
Development, vol. 17, 1973, pp. 525-532.
3) R. Feynman, "Simulating Physics with
Computers," International Journal of Theoretical
Physics, 1982.
4) R. Feynman, "Quantum Mechanical
Computers," Foundations of Physics, vol. 16,
iss. 6, 1986,
5) E. Fredkin and T. Toffoli, "Conservative
Logic," International Journal of Theoretical
Physics, vol. 21, 1980, pp. 219-53. [6] T. Toffoli,
"Reversible Computing," Technical Report
MIT/LCS/TM-151, 1980.
6) A. Peres, “Reversible Logic and Quantum

@ IJTSRD | Available Online @ www.ijtsrd.com | Volume – 1 | Issue – 6 | Sep - Oct 2017 Page: 1301
1985, pp. 3266-3267.
7) M. Morrison and N. Ranganathan, "Design of
a Reversible ALU Based on Novel
Programmable Reversible Logic Gate
Structures," To Appear in the IEEE
International Symposium on VLSI, 2011.
8) R. Feynman, "Quantum Mechanical
Computers," Foundations of Physics, vol. 16,
iss. 6, 1986,
9) E. Fredkin and T. Toffoli, "Conservative
Logic," International Journal of Theoretical
Physics, vol. 21, 1980, pp 219-53
10) T. Toffoli, "Reversible Computing," Technical
Report MIT/LCS/TM- 151 , 1980.
11) A. Peres, “Reversible Logic and Quantum
Computers,” Physical Review, vol. 32, iss. 6,
1985, pp 326 .

A Novel Design of a 4 Bit Reversible ALU using Kogge-Stone Adder

More Related Content

What's hot (18)

Similar to A Novel Design of a 4 Bit Reversible ALU using Kogge-Stone Adder (20)

More from ijtsrd (20)

Recently uploaded (20)

A Novel Design of a 4 Bit Reversible ALU using Kogge-Stone Adder