A Bus Encoding Method for Crosstalk and Power Reduction in RC Coupled VLSI Interconnects

International Journal of VLSI design & Communication Systems (VLSICS) Vol.3, No.2, April 2012
DOI : 10.5121/vlsic.2012.3203 29
A Bus Encoding Method for Crosstalk and Power
Reduction in RC Coupled VLSI Interconnects
S.K.Verma1
and B.K.Kaushik2
1
Department of Computer Science and Engineering,
G. B. Pant Engineering College, Pauri-Garhwal, INDIA
skverma.gbpec@rediffmail.com
2
Department of Electronics and Computer Engineering,
Indian Institute of Technology, Roorkee, INDIA
bkk23fec@iitr.ernet.in
ABSTRACT
The performance factors such as propagation delay, power dissipation and crosstalk in RC modelled
interconnects are major design issues for the System on-chip (SoC) designs in current Deep Submicron
(DSM) era. The crosstalk effect is a consequence of coupling and switching activities that is encountered
when there is a transition as compared to previous state of wire and or when there are transitions in
adjacent wires. Therefore, minimization or elimination of switching and coupling activities is crucial in
enhancing the performance of SoC designs. There are several methods for the reduction of power
dissipation, crosstalk and delay. The encoding method is most effective and popular method for enhancing
the behaviour of on-chip buses. This paper proposes encoding scheme to achieve overall reduction in
transitions. The reduction in transition improves the performance in terms of reduced power dissipation,
coupling activity and delay in on-chip buses. This encoding method is implemented using VHDL. The result
evidently demonstrates reduction in transitions which consequently improves the overall performance of
on-chip buses.
Keywords
Coupling, VLSI, SoC, Bus Encoding, Interconnects
1. INTRODUCTION
The feature size of integrated circuits has been consistently reduced in the pursuit of improved
speed, power, silicon area and cost characteristics. Semiconductor technologies with feature sizes
of several tens of nanometres are currently in development. As per International Technology
Roadmap for Semiconductors (ITRS), the future nanometre scale circuits will contain more than a
billion transistors and operate at clock speeds well over 10GHz. Distribution of robust and
reliable power and ground lines; clock; data and address; and other control signals through
interconnects in such a high-speed, highly-complex environment, is a challenging task. The
function of wiring systems or interconnects is to distribute clock and other signals and to provide
power/ground to and among the various circuits/systems functions on the chip. The performance
parameters i.e. time delay and power dissipation of a high-speed chip is highly dependent on
interconnects, which connect different macro cells within a VLSI chip.

30
In current DSM (Deep Submicron) technology, coupling and self capacitance plays an important
role for deciding the behaviour of on-chip interconnects. Due to the coupling and self
capacitance, crosstalk, delay and power consumption problems will arise. The above problems
are highly dependent on the frequency of signal used, length of interconnects, data signal etc.
Interconnects can be modelled as RC or RLC transmission line. The effect of inductance plays an
important role as the length and used frequency signal increases. [1-4, 8-16] This paper considers
RC network for the implementation and study of the behaviour of interconnects due to coupling
capacitance.
The components that affect the behavior of the on-chip bus are internal parasitic capacitances of
the transistors, interconnect capacitances and input capacitances of the fan-out gates.
The most common methods to reduce crosstalk, propagation delay and power are:
• Insertion of repeaters
• Insertion of shielding between adjacent wires
• Minimizing spacing between signal and ground lines.
• Isolating clocks and other critical signals from other lines (larger line spacing) or isolation
with ground traces.
• In backplane or wire-wrap applications, use twisted pair for sensitive applications such as
clocks and asynchronous set or clear functions. While using ribbon or flat cable, make every
other line a ground line.
• Introduction of intentional delay among coupled signal transmission.
• Bus Encoding methods
• The use of tight geometry in most systems can reduce crosstalk significantly although it
cannot eliminate it entirely. Some preventive design measures can be used to minimize
crosstalk.
• Using maximum allowable spacing between signal lines.
• Terminating signal lines into their characteristic impedance.
The Bus encoding method is widely used technique to reduce dynamic switching power and the
effects of crosstalk (signal noise, delay) during data transmission on buses [15]. Low power
encoding techniques aim to transform the data being transmitted on buses in such a manner so
that the self and coupling switching activity on buses are reduced. Crosstalk aware encoding
techniques can also modify the switching patterns of a group of wires to reduce crosstalk
coupling effect. These techniques are quite effective in reducing power consumption, improving
transmission reliability, and increasing system performance. For any encoding scheme, the
encoder and decoder functions are the inverse of each other. Bus encoding schemes can be
classified according to several criteria, such as the amount of extra information needed for coding
(redundant or irredundant coding), and the method of encoding implementation (hardware,
software, or a combination of the two), Type of code used (algebraic, permutation, or probability
based), the degree of encoding adaptability (static or dynamically adaptable encoding), the
targeted capacitance for switching reduction (self, coupling or both). Encoding techniques are

31
often aimed at power reduction, signal transmission delay reduction and reliability improvement,
or a combination of these due to the reduction in the transition. Certain optimizations such as
crosstalk reduction can have multiple benefits associated with them such as power reduction,
signal delay reduction and noise reduction.
This paper presents an encoding method for the reduction of coupling transitions. The crosstalk
is classified into different types [1, 2, 15, 16] depending upon the transitions of the signal in the
wire. The proposed encoding scheme also reduces data lines hence, reduces the redundancy and
power consumption. It also considers the worst case crosstalk effects due to transitions in the
group of lines.
2. Estimation of Power and Crosstalk in RC Bus Model
The total power consumption in the VLSI chip comprises of dynamic power, short circuit power,
static power and leakage power. It can be simply described as summation of all these components
i.e.
ܲ஽௜௦௦௜௣௔௧௜௢௡ = ܲௌ௧௔௧௜௖ + ܲ஽௬௡௔௠௜௖ + ܲ௅௘௔௞௔௚௘ + ܲௌ௛௢௥௧௖௞௧ (1)
The capacitance of interconnect can be classified as coupling capacitance and self-capacitance.
The coupling capacitance is the capacitance between the adjacent data lines while the self-
capacitance refers to the capacitance between the substrate and the wire itself. The dynamic
power in VLSI chip decides the behaviour of chip and is highly dependent on the load
capacitance and coupling capacitance i.e. bus line signal transitions. Dynamic power dissipation
on a coded bus thus can be defined as following equation
ܲ஽,஼௢ௗ௘ௗ = ሺߙ௖௟ × ‫ܥ‬௅ + ߙ௖௖ × ‫ܥ‬஼ሻ × ܸ஽஽
ଶ
× ݂
= ሺߙ௖௟ + ߙ௖௖ × ߣሻ × ‫ܥ‬௅ × ܸ஽஽
ଶ
× ݂ሻ (2)
where, CL, Cc, VDD, f and λ is the load capacitance, coupling capacitance, supplying voltage,
clock frequency and capacitance ratio defined as: λ = Cc/CL respectively. λ is dependent on the
technology which is being used, hence its value depends on the physical parameters. αcl is the
value of average switching activity for the load capacitance. For un-encoded buses αcl is 1. αcc is
the value of average coupling activity for the coupling capacitance. For un-encoded buses αcc is 1.
Hence for un-encoded data the power dissipation can be defined by following equation
ܲ஽,௨௡ି௖௢ௗ௘ௗ = ሺ1 + ߣሻ × ‫ܥ‬௅ × ܸ஽஽
ଶ
× ݂ (3)
Effective crosstalk capacitance is determined by
‫ܥ‬௘௙௙ = ‫ܥ‬௖ ×
௱௏మି௱௏భ
௏ವವ
+ ‫ܥ‬஼ ×
௱௏మି௱௏య
௏ವವ
(4)

32
where, ∆V2 is voltage variation of the centre wire. ∆V1 and ∆V3 are voltage variations in the
adjacent wires. ܸ஽஽ is power supply voltage which equals rail-to-rail signal voltage in CMOS
circuits. ‫ܥ‬௘௙௙ is effective coupling capacitance variation [1, 2].
The coupling between the groups of the three wires is classified into five types depending upon
the nature of transitions of signals in the wire that are Type-0, Type-1, Type-2, Type-3 and Type-
4. The Type-0 coupling occurs when all of the 3-bit wires are in the same state transition. A
transition from 000 to 111 (i.e. ↑↑↑) causes a Type-0 coupling. For Type-0, coupling capacitance
is zero. Type-1 coupling occurs when there is a transition in one or the two wires (including the
centre wire) and the third wire remains quite. There are eight possibilities by which Type-1
condition occurs. The coupling capacitance in this case is ‫ܥ‬஼. A Type-2 coupling occurs when the
centre wire is in the opposite state transition with one of its adjacent wires while the other wires
undergo the same state transition as the centre wire i.e.100 to 011. Ten different conditions are
possible for Type-2 coupling. The coupling capacitance is 2‫ܥ‬஼ in this case. A Type-3 coupling
occurs when the central wire undergoes the opposite state transition with one of the two wires
while the other wires are quiet i.e. 010 to 001. Coupling capacitance in the case of Type-3
coupling is 3‫ܥ‬஼ and there are four possibilities that cause Type-3 coupling. For a Type-4 coupling,
all three wire transitions are in the opposite states with respect to each other. Two conditions
cause Type-4 coupling with a coupling capacitance effect of 4‫ܥ‬஼. All the five Types of couplings
are shown in Table-1 [2].
Table.1. Crosstalk Types for a 3-Bit Bus Considering RC Model of Interconnect [2].
↑: switch from “0” to “1”, ↓: switch from “1” to “0”, - : no transition
It can be concluded from the above description that power and crosstalk is highly dependend on
the transitions of the signal in the wires. If the number of transitions is reduced by encoding
methods, the dynamic power dissipation as well as crosstalk will also be reduced. Different
encoding methods are proposed by different researchers. The efficient encoding proposed by
Burleson et al. [3], which include the concept of counting the number of transitions, with respect
to the previous states of group of lines. Data is transmitted in original form or in inverted form
depending on the number of state transition of the lines as compared to previous states.
Type-0 Type-1 Type-2 Type-3 Type-4

33
3. Encoding Method
Fig. 1. Block Diagram of Encoding Method
The block diagram of proposed encoder is shown in Fig.1 that consists of counter, controller,
comparator and three 3-bit register. Each module of the encoding method is explained in later
sections. The encoder transforms the bus line signals for the reduction or elimination of the worst
case crosstalk by reducing the 7-bit lines into 4-bit lines. The encoder deals with the coupling
transitions among the group of seven bits. It identifies the higher number of 0’s and 1’s in counter
and controller modules. Output line of controller is set according to higher number of 1’s or 0’s.
The comparator module compares the input 7-bit data lines with the output line of the controller
to identify the flips in the line.
Firstly, the number of 1’s and 0’s are counted using a counter module. If the number of 1’s is
more than the number of 0’s, the comparator module sets output line in high state i.e. 1. If the
number of 0’s are more than the number of 1’s, the comparator module sets output line in low
state i.e. 0. There are two best cases possible; either all lines are in one state or all lines are in zero
state. Maximum three flips are possible in comparator module due to the counter and controller
modules. These flip positions are stored in three 3-bit registers.

34
The single output line of controller is compared with initial 7-bit input line and finds the flipped
bit positions. The number of flipped bits could be 0 i.e. best case when all the inputs are either ‘0’
or ‘1’, or 1 or 2 or 3 i.e. worst case. After identification of the flipped bits positions, these
positions are stored in the 3-bit registers. The number of the registers is chosen to be three in
order to work for the best as well as the worst cases. The best case is one when all the input lines
are ‘0’ or ‘1’. In this case all the registers are empty as there is no change in any of the bits. Thus,
there is significant reduction in crosstalk as all three registers have same null value. Otherwise,
the value of the registers can vary from zero to maximum three. The seven bits data lines are
taken as inputs to the encoder, so, three bits have taken to indicate the flip bit positions in the
seven lines. All the equivalent positions for these 8 combinations are shown in Table.2. Each of
the register value sends in three different clock cycles. At each clock cycle, contents of one
register are sent to the decoder. Next register contents are sent in the next clock cycle and so on.
Since the maximum flips that can be possible are three so maximum three clock cycles are
required for decoding the 7-bit line at decoder side.
Table 2. Register Values Indicating the Positions of Flipping
Register Value Flipped I/P Line Position
000 0
001 1
010 2
011 3
100 4
101 5
110 6
111 7
3.1 Counter Module
The counter module is for the determination of the number of 0’s and 1’s in the input sequence of
seven bit lines. The input to the counter module is 7-bit input. Each of these lines is selected and
then compared with ‘0’ and ‘1’. Two variables are initialized for storing the values of the number
of 0’s and 1’s in the input sequence. The first variable stores the value of the number of 0’s and
the other one store the number of the 1’s in the input sequence. These two lines are then fed into
the controller module along with two variables and are passed to the controller.
3.2 Controller
As seen in Fig.1 the two output lines of the counter represents the number of 1’s and 0’s. These
two inputs are fed to the controller. The Controller decides for the number of occurrences of 0’s
and 1’s as provided by counter module. This module decides the output as ‘0’ or ‘1’ based on the
input value. If number of 0’s is higher as compared to the number of 1’s the output of the
controller will be ‘0’. If number of 1’s is higher as compared to the number of 0’s the output of
the controller will be ‘1’.

35
3.3 Comparator
The single output line is fed to the comparator along with the initial 7-bit inputs. This single
output line is compared with each of the initial 7-bits. This is to compare whether the value of the
bit is the same or different from the single line. The output signal from the controller is XORed
with each of the seven input lines to identify the flipping. This module identifies the flipped
positions and stored in the 3-bit register for the decoding purpose.
3.4 Registers
This module is used for storing the position of flipped bit identified by the comparator module.
This ascertains the proper interpretation of the input bits where a 000 value in the register stands
for the 0th
bit position, 001 represents the 1st bit and likewise 111 represents 7th
position as shown
in the Table.2. These three registers are meant for the simultaneous storage of the flipped bit
positions. These lines along with the single input equivalent line from the controller are
transmitted to the decoder. The encoder provides the output in four lines for the decoding purpose
i.e. one line from the controller and the other three lines from the register. Three clock cycles are
needed for complete decoding of the input sequence.
4. Decoder
The decoder of the proposed encoding method is shown in Fig.2. The decoder of the proposed
encoding method consists of the three 3-bit registers, splitter, inversion module and line identifier.
The four lines from the encoder are fed to the decoder. In the first clock cycle the contents of the
first register from encoder are stored in the first register of the decoder and so on. As we know the
maximum flips that can be possible are three, so maximum three clock cycles are required for
complete decoding of the 7-bit data. Although, we can also use single 3-bit register in the encoder
and decoder side instead of three 3-bit registers, three registers are used for the compensation of
delay generated by the clock cycles for the transmission of flipped bit positions to decoder side.

36
Fig. 2. Decoder of Encoding Method
The output line taken from the comparater of encoder side is fed to the spliter module of the
decoder. The splitter module splits the 1-bit input line to seven bits of output lines. All the seven
output lines have similar value as that of the input i.e. if the input to the spliter is ‘1’ then output
of the splitter is of seven lines having value of each line as ‘1’. Similarily when input of the
splitter is ‘0’ all the seven lines will have output value ’0’. The line identifier module takes the
input from the three 3-bit registers sequentially. The sequence is managed by the clock cycle.
Line identifier gets the content of 3-bit register and identifies the line to be flipped as per Table 2.
These identification indications are then fed to the inversion module. The inversion module
inverts the indentified line. In next turn this process is repeated. And after three iterations the final
decoded data of seven bits is taken as the output of the decoder.
5. Simulation and Results
Encoder and decoder are implemented in VHDL[5]. The encoder described in the previous
sections had a 7-bit input sequence, primarily targetting for reduction in crosstalk and power
dissipation. The designed encoder reduces the 7-bit input sequence to a 4-bit output sequence
which reduces the redundancy. The number of wires is reduced by 57.14%. Therefore,
redundancy is also reduced by 57.14%. However, in other encoding schemes, redundancy
increases by 25%. The proposed encoding scheme not only reduced the redundancy but also
reduces the power consumption. On implementation of encoding mechanism, it is observed that
the crosstalk reduces by 35 to 40%. The encoder is tested on the random 7-bit sequences and it
was found that the worst case crosstalk reduced by 40%. This encoding method reduces the
power dissipation and crosstalk. Whereas, a three clock delay is introduced. This delay can
therefore be allowed as the overall implementation is better than the initial input sequence
transmission method. Hence, it is evident that this method is acceptable. This encoder considers
only the coupling transitions in the input 7-bit sequence.
7-bit
output
data
Output
from
Encoder
(ED)
7
3
3
3
3
3
3
70/1
Splitter Inversion
Module
Line
Identifier
3 bit
Reg
3 bit
Reg
3 bit
Reg

37
6. Conclusion
The proposed method of bus encoder significantly eliminates or reduces the worst case crosstalk.
Reduction of crosstalk is because of the reduction in the number of lines from 7 bits to 4 bits as
an output of the Encoder. This reduces the crosstalk effectively by 35 to 40 percent. The
transitions in the state of buses decide the behaviour of the switching and coupling activities. It is
shown that reduction in the coupling and switching activity reduces the dynamic power
dissipation and crosstalk. Thereby, the power dissipation is also reduced as compared to the initial
input sequence. Hence, the overall performance of the VLSI chip is improved. For all possible
128 combinations of the input sequence the output is seen manually and via the implemented
mechanism. It is finally found that the method is implemented successfully and serves its purpose
to a great extent.
7. ACKNOWLEDGEMENTS
Special thanks to Mr. Kanwaljeet Kathait, Ms. Sushila Mishra, and their group members for their
valuable efforts and cooperation.
REFERENCES
[1] Duan C., Calle V.H.C. & Khatri S.P., (2009) “Efficient On-Chip Crosstalk Avoidance CODEC
Design”, IEEE Transactions on Very Large Scale Integration (VLSI) Systems, ISSN: 1063-8210,
Vol. 17, Issue 4,pp551-560.
[2] Khan Z., Arslan T. & Erdogan A. T., (2005) “A Low Power System on Chip Bus Encoding Scheme
with Crosstalk Noise Reduction Capability”, IEEE Proc. Computer & Digital Techniques.
[3] Stan M. R., Burleson W.P., (1995) “Bus-invert coding for low power I/O”, IEEE Transactions on
Very Large Scale Integration Systems (TVLSI), pp49–58.
[4] Shin Y., Chae S. & Choi K., (1998) “Partial Bus-Invert Coding for Power Optimization of System
Level Bus”, Proceedings of the 1998 International Symposium on Low Power Electronics and
Design, pp127-129.
[5] VHDL Reference Guide, Xilinx, Inc., 1999 (http://toolbox.xilinx.com/docsan/).
[6] Perry D. L., “VHDL”,Programming by example, 4th
edition, McGraw-Hill.
[7] Mano M., Kime C., (2001) “Logic and Computer Design Fundamentals”, 2nd
Edition, Prentice Hall,
Upper Saddle River.
[8] Aghaghiri Y., Fallah F. & Pedram M., (2004) “Transition reduction in memory buses using sector
based encoding techniques”,IEEE Transactions on Computer-Aided Design of Integrated Circuits
and Systems, Vol. 23, No. 8,pp1174–1184.
[9] Satyanarayana N., VinayaBabu A. & Mutyam M., (2009) “ Delay-efficient bus encoding
techniques”, Microprocessors & Microsystems, Vol. 33, No.5-6,pp365-373.
[10] Lin T., Shang T. & Jou J., (2007) “On-Chip Bus Encoding for Power Minimization under Delay
Constraint”, IEEE International Symposium on VLSI Design, Automation, and Test (VLSI-DAT).

38
[11] Kaushik B.K., Sarkar S., (2008) “Crosstalk Analysis for a CMOS-Gate-Driven Coupled
Interconnects”, IEEE Trans. on Computer-Aided Design of Integrated Circuits and Systems, Vol. 27,
No. 6, pp1150-1154.
[12] Kaushik B.K., Sarkar S. & Agarwal R.P., (2007) “Waveform Analysis and Delay Prediction for a
CMOS Gate Driving RLC Interconnect Load”,Integration, the VLSI Journal, Elsevier
Pub.,Netherlands, Vol. 40, No. 4, pp394-405.
[13] Kaushik B.K., Sarkar S., Agarwal R. P. & Joshi R. C., (2009) “Crosstalk Analysis of Simultaneously
Switching Interconnects”, International Journal of Electronics, Taylor and Francis (UK), Vol. 96,
No.10, pp1095-1114.
[14] Kaushik B. K., Sarkar S., Agarwal R. P. & Joshi R.C., (2007) “Effect of Line Resistance and Driver
Width on Crosstalk in Coupled VLSI Interconnects.”, Microelectronics International, Emerald Pub.,
U.K., vol. 24, No. 3, pp42-45.
[15] Verma S.K., Kaushik B.K., (2010) “Encoding Schemes for the Reduction of Power Dissipation,
Crosstalk and Delay”, International Journal of Recent Trends in Engineering [ISSN 1797-9617],
Academy Publishers, Finland, vol. 3, No. 4, pp74-79.
[16] Verma S.K., Kaushik B.K., (2010) “Reduction of transitions in RC coupled systems in VLSI
interconnects”, International Conference on Emerging Trends in Engineering and Technology
ICETET, IEEE computer Society [ISBN 978-0-7695-4246-1], pp735-740.
Authors
Shashi Kant Verma received B.E. degree in Electronics and Communication
Engineering from G. B. Pant Engineering College, Pauri –Garhwal,
Uttarakhand, India in 1999; and M-Tech degree in Computer Science and
Engineering from MNNIT-Allahabad, UP, India in 2002. He is currently
pursuing PhD degree in Computer Science and Engineering from Uttarakhand
Technical University Dehradun, Uttarakhand, India. Since 2002, he has worked
in various institutions in the field of teaching and research work. Presently, he is
serving as Assistant Professor in Department of Computer Science and
Engineering, G.B.Pant Engineering College, Pauri Garhwal. His research interests include Encoders in
VLSI Interconnects, Low Power Digital Circuit, Design and Implementation of Signal Processing Systems.
Brajesh Kumar Kaushik received his BE degree in Electronics and
Communication Engineering from C. R. State College of Engineering, Murthal,
Haryana in 1994. He received M.Tech in Engineering Systems, from Dayalbagh
Educational Institute-Agra, in 1997. He obtained PhD under AICTE-QIP scheme
from IIT Roorkee, India. He served Vinytics Peripherals Pvt. Ltd., Delhi as
Research & Development Engineer. He joined Department of Electronics and
Communication Engineering, G.B.Pant Engineering College, Pauri Garhwal, as
Lecturer in 1998, where he later served as Assistant Professor from May, 2005 to
Dec, 2009. Presently, he is serving as Assistant Professor in Department of Electronics and Computer
Engineering, Indian Institute of Technology-Roorkee. He has published more than 100 research papers in
national and international journals and conferences. His research interests are in VLSI Interconnects, Low
power VLSI Design, Carbon Nanotube based Designs, Organic Thin Film Transistor Design and
Modelling, Microprocessor based system design and Digital Signal Processing. He is reviewer of many
international journals such as IEEE Transactions on Circuits and Systems, IEE Journals, International

39
Journal of Electronics, Elsevier Journals, Emerald etc. He has delivered many keynote talks in international
and national conferences. He also holds the position of Editor of Journal of Electrical and Electronics
Engineering Research, Academic Journals. He has received many awards and recognitions from
International Biographical Center (IBC), Cambridge, Great Britain such Top 100 Scientists in World- 2008,
Leading Scientist of the World- 2008, International Scientist of the Year for 2008, IBC Foremost Scientists
of the World-2008, Leading Educators of the World 2008, IBC Foremost Educators of the World 2008,
IBC International Educator of the Year 2008. His name has been listed in Marquis Who’s Who in Science
and Engineering® and Marquis Who’s Who in the World®.

A Bus Encoding Method for Crosstalk and Power Reduction in RC Coupled VLSI Interconnects

More Related Content

What's hot (18)

Similar to A Bus Encoding Method for Crosstalk and Power Reduction in RC Coupled VLSI Interconnects (20)

Recently uploaded (20)

A Bus Encoding Method for Crosstalk and Power Reduction in RC Coupled VLSI Interconnects