Design and Performance Analysis of 8 x 8 Network on Chip Router

International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 06 | June 2022 www.irjet.net p-ISSN: 2395-0072
© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 456
Design and Performance Analysis of 8 x 8 Network on Chip Router
Preethi T S1, Jayanth K Holalkeri2, Manasa Dev G J3, Manoj M Biradar4, Nagesh R Shanbhag5
1Assistant Professor, Dept. of Electronics and Communication Engineering, Sapthagiri College of Engineering,
Karnataka, India
2345 Student, Dept. of Electronics and Communication Engineering, Sapthagiri College of Engineering, Karnataka,
India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract – Network on chip is a new prototype model in
complex system-on-chips that provide an effective on chip
communication networks. The information is transmitted
through the interconnected modules in terms of packets. The
main task of router is routing the data. So, the router
architecture must be efficient with lower latency and higher
throughput. In this project, we design and implement a 2D
Network-on-Chip router mesh having four ports connected to
other ports in four directions, namely North, South, West, East,
and the fifth connected to the Local processing element
through a network interface. This project is aimed to improve
“Quality-of-Service” by employing algorithms like wormhole,
routing, arbiter and crossbar switching.
Key Words: Bandwidth reservation, Flitting, Label
switching, Networks-on-chip, Quality of Service,
Streaming applications, System-on-Chip.
1. INTRODUCTION
The recent technological advancements in the domain of
semiconductor integration technologyistofill thenumberof
working transistors on an available chip area forfabrication,
satisfying the Moore’s law as shown in Fig-1, which resulted
in developments of techniques and technologies in the field
of System on Chip, hence this has resulted with new arena to
the work space of miniaturization and this has highly
influenced by gate level circuit association for the
Application Specific Integrated Circuit (ASIC). Then
technologies have to be developed in-line with the demands
of market space and design time for execution, all these
criteria of innovations to meet the needs of communication
in the existing domain which lead the scope for the evolving
of networking on the same silicon chip for communicates
leading to the era of Network on Chip.
Fig -1: Evolution of transistor integration on a chip
Network on-chip is one such innovation known for its
adaptation and flexibility of operation over the chip by
satisfying the factor of power, area and number ofgates.The
ASIC and SoC design approaches are well suited for various
integration of peripherals with its approaches meeting the
architectural requirement for the minimization of the delay
and effective communication among the devices in both
simplex and duplex communicationcriteria.Thisleadsto the
simplification of various redundant interconnections among
various peripherals in the connectionplatform whichshould
be fault tolerant as well for successiveimplementationof the
desired system. But in the present scenario, for the effective
on-chip communication design, SoC is the best suited option
but not fair enough for the effective communication among
the peripherals. Hence, it requires reframing of the
architecture to achieve minimum delay with maximum
reliability, which can be obtained by interfacing NoC with
SoC.
NoC based systems also playa vital rolefordataexchange.To
meet the needs and requirements of data handling over the
different types of topology and algorithms for memory or
buffer-based datatransmissiononanFPGAdevice,itrequires
Dynamically Reconfigurable NoC(DRNOC).Inthisprojectwe
are bound to design a router under two-dimensional
structure. Later the design is proposed in generalizing to N x
N structure for the improvement of the properties of the
FPGA viz, delay, throughput, LUT and flip flops using Xilinx
ISE platform for simulation and testing of the properties
comparatively over the existing systems.
2. LITERATURE SURVEY
The paper [1] proposes switch model design and
implementation forstudents’laboratoryactivities. Thethree
staged switching architecture comprising of software and
hardware systems which arebasedonmodular3x3switches,
with controller as an Arduino microcontroller. The switch
uses a few stages of smaller NxN crossbar switches to help in
producing non-blocking connections and forming an NxN
network. As the tiniest component within which a switching
system exists, is a squared cross-bar switch. This switch has
to be designed in such a way that it should be able to connect
all inputs to all its outputs. The blocking is due to busy
destination as the experiment conducted also it showcases
94% successful connections while 6% blocking is not caused
due to busy switches, but due to busy destinations.

The paper [6] examines the impact of failures on the
reliability of the chips and develops corelating passive
counter protocols either to prevent or recover from them. In
this view, several schemes were adapted in order to reduce
various kinds of error symptoms,whileareaandpowerareat
a minimum. Over a normal transmissionscheme,itrequiresa
3-flitting re-transmission buffer per virtual, since a flit must
be kept for 3 cycles after it leaves the current node. The least
complex type of transmission buffer is a Queue working on
the principle of First-In First-Out (FIFO) buffer [25]. Such
implementation type has one input port and one outputport,
and has a control logic. The concept of deadlock has been
enormously stressed upon. Some of errors mainly
encountered in the router are VC allocator error, crossbar
error, handshaking error. The schemeturnedouttobehighly
effective both in terms of latency and power even under high
error rates. All of these mechanisms described in the work
helps in keeping the critical path of the Network on Chip
router intact.
The work proposed in [9] describes the design of on chip
routers based on power consumed and its area occupied by
chip. Proposedarchitectureof on-chip routerinthisgivesthe
results in which power consumption is reduced and silicon
area is also minimized. This router for NoC increases
throughput, and introduces architecture which shows an
improvement in Figure of Merit. But increases the area and
power due to extra crossbar and arbiter scheme, and got up
to 94% of throughput, whilepowerconsumptionisincreased
by the factor of 1.28. [10] This approach reduces power
consumption. It also pointsatthebufferutilizationbymaking
the channels bidirectionalandshowsdrasticimprovementin
the performance of system. A router architecture with
Reliability Aware Virtual Router allocates more memory to
the busy channels and less to the idle channels. This solution
is delay efficient but not area and power efficient.
The paper [20] presents how a NoC topology, routing
algorithm and switching are vital in communication. In any
NoC architecture, routing algorithm is an important key
factor. The path chosen by the packet from source to
destination is defined by a routing algorithm. XY algorithmis
one of the simplest routing algorithms used in NOC. The
paper reviews a different XY routing algorithm that is
implemented. The use of this algorithm is totally dependent
on the application and environment. The selection of which
totally depends on the applicationandthetrafficofpacketsin
the network. Because the simplicity in implementation is
important in all architecture, XY routing algorithm is widely
used. The Dynamic-XY routing algorithm achieves better
balance in load distribution and also provides deadlock-free
and livelock-free facility where the user can’t compromise
with accuracy of data received, they go for the fault-tolerant
routing like XYX. Finally, we can conclude that the choice of
XY routing algorithm is totally dependent upon
environmental condition of NOC architecture.
The paper [14] proposed an approach to find the shortest
path distance between two nodes implementing on FPGA
technology. The FPGA articulates parallelism to significantly
reduce steps as compared to sequential effort. In this paper,
A* algorithm is chosen for the shortest path distance
calculation based on its heuristic behavior since it achieves
superior time running. Shortest path algorithms are applied
for applications like transportation and networking. If the
shortest path is calculated on general purpose processor, to
read the input there are separate instructions, and then
compute the result to produce the output which slows down
the overall performance. The research work reducesdelayin
finding the shortest path. This research [14] contributes in
finding the shortest path with reduced delay by taking
advantage of parallel processing in FPGA. A new design in
coprocessor implementationisproposedusingA*algorithms
technique, and using this brings in the conceptofparallelism.
3. SCOPE OF WORK
Based on the literature survey and various studies, in this
paper we are bound to design a Label Switching Network on
Chip router comprising of following:
1. Crossbarswitch to establish a connection from each
input to each output.
2. Arbiter module to check port validity and to flit.
3. A NoC manager used for bandwidth allocation.
4. A LS router that checks if the router is valid or not.
We design and implement all the modules mentioned above,
using Xilinx ISE 14.7and instantiateitinatopmoduletoform
a single router. Further we implement power optimization
techniques to reduce the power and improve QoS.
4. METHODOLOGY
The following steps are carried out to form an 8x8 NoC
Router.
4.1 To Design a Crossbar Switch
A Crossbar switch is designed in Xilinx ISE asshowninFig-2.
1. Every on chip router has 5 ports namely, Local,East,
West, North and South.
2. Every port needs to have one multiplexer, therefore
for five ports we need five MUX.
3. The crossbar module has seven input ports and five
output ports and five wires for intermediate
connections.

4. The 5bit enable bus is used to enable the particular
MUX and 3bit select line is used to connect inputtoa
particular output.
5. Every possible condition of enable and select lines
are specified and based on that condition the
incoming ports and output ports are connected.
6. The select line bits are determined by the arbiter
module discussed below.
Fig -2: Crossbar Switch using MUX
4.2 To Design an Arbiter module
An arbiter module is designed in Xilinx ISE and the RTL
schematic is shown in Fig-3.
1. An arbiter module is designed using arbiter
algorithm and it consists of 5 input ports and 6
output ports.
2. This module helps to flit the message bits.
3. Based on the condition of the destination ID and
label, it checks the validity of the port.
4. Inferring the select conditions of all the five ports,
the output port is allocated with the select line bits
used in crossbar module.
Fig -3: RTL Schematic of Arbiter module
4.3 To Design an NoC Manager
An NoC manager is designed andtheRTLschematicisshown
in Fig-4.
1. NoC manager is used to monitor the used and the
unused bandwidth.
2. It mainly works on the principle of edge detection.
3. Edge detection between two correspondingrouters
is done by considering LSB bits as shown in Fig-5.
4. Every port has one NoC manager, so for five ports
there are five NoC manager present in a single
router.
5. Scaling factor is set to 5 between two slots and 2
between the packets in one slot.
6. Only if the bandwidth is greater than 2, the port is
considered to be available.
Fig -4: RTL Schematic of NoC manager
Fig -5: Edge Detection occurring in NoC manager

4.4 To Design a Label Switched Router
Label switched router is designed and the RTL schematic
is shown in Fig-6.
1. A label-based switching router is connected to all
the five ports.
2. It also consists of a Queue module which
implements FIFO operation.
3. It is used to generate labels and to check whether
port is valid or not.
4. It checks whether received packet is valid or not
and whether the packet is successfully transmitted
or not.
5. It also checks congestion of the port, if the port is in
use the packet is buffered in FIFO module and
whether the FIFO is full or not.
Fig -6: RTL Schematic of LS Router
5. RESULTS OF THE PROJECT
5.1 Simulation of the project.
Crossbar Switch and Arbiter module are simulated and
verified its operation successfully. The simulation results of
crossbar and arbiter module are shown in Fig-7 and Fig-8
respectively.
Fig -7: Simulation results of Crossbar module
Fig -8: Simulation results of Arbiter module
The specifications for this project is mentioned in Table-1.
Table -1: Design Specification
Sl. Specification Value
1. Design Topology 2D Mesh
2. Size 8x8
3. Switching Mechanism Crossbar
4. Packet Format Header, Message
5. Routing Algorithm Deterministic XY
6. Packet Size 22 bits
7. Simulation Tool Xilinx ISE 14.7
8. Kit FPGA Spartan 6
Once all the sub-blocks are designed and verified, we
instantiate all the modules in the top module to form a single
router. A packet format of 22 bits is considered of which 6
bits denotethe source and destinationaddress,remaining16
bits denotes the message. Then the singlerouterismultiplied
9 times to form a 3x3 router, and 64 times to form a 8x8
router. To verify the functionality of 8x8 router, initially a
test bench is written andsimulated. The simulationresultsof
the 8x8 router are shown in Fig-9.

Fig -9: Simulation results of 8 x 8 router
5.2 Power and Area Optimization
Before implementing it on the FPGA, we have written an
algorithm to optimize the power. The algorithm is aimed to
reduce the transitions in the message bits. For this we
adopted the encoding and decoding techniques. Before
transmitting the message is encoded using the bits b1 and
b2. Any combination of b1 and b2 can be taken, but in this
project, we have taken b1 as 0 and b2 as 1. All the even bits
are XOR with b1 and odd bits are XOR with b1 and b2. The
same decoding technique is used at the receiver to retrieve
the original message. An example is shown in Fig-10. From
the figure we can see that the number of transitions is
reduced from 6 to 3, which reduces the power consumption
during the process of transmitting.
Fig -10: Power Optimization Technique
After successful verification of 8x8 router, it is implemented
on FPGA Spartan 6. The power consumption is observed to
be 0.085W as shown in Fig-11. The area details in the design
summary are shown in Table -2.
Fig -11: Power Consumption Report
Table -2: Design Summary
Logic Utilization Used Available Utilization
No. of Slice registers 21,325 126,576 16%
No. of Slice LUTs 28,005 63,288 44%
No. of occupied Slices 9,974 15,822 63%
No. of IOBs 40 296 13%
ChipScope Tool is used to implement the project and
additional files required toimplement are writtenandadded
as a source to the top module. Data port width is taken as 57
and classified 16 ports for sending data, 16 ports for
receiving data, 16 bits for the data, 4 bits for expected and
arrived port and 1 port for reset. Then trigger setup is done
and waveform is added to the window. The implementation
and its results are shown in Fig-1.
Fig -13: Implementation results using ChipScope
6. CONCLUSION
The various challenges in SoC design made the researchers
look for better alternatives that made a way for Network on
chip technology. The NoC is currently still in its initial stages
and emerging research area.NoChasaremarkableimpacton
the SoC design of next generation and multicore
architectures. For this project, various surveys and recent

research aspects of NoC, routing algorithms and network
topology details were explored. An attempt was made to
contribute in the research of NoC by exploring the space of
NoC router design which is a predominant component of the
network. The main focus of this project is to design an
efficient router since it is one of the most important
components that determines various network parameters
like latency, throughput and delay. The designing and
simulation have been done using the hardware description
language Verilog HDL in Xilinx ISE tool and implemented on
Field Programmable Gate Array, using ChipScope Tool.
REFERENCES
[1] Sani. A. and Fauzi. R, “Crossbar switch module
implementation for laboratory purpose”, In Journal of
Physics: Conference Series, Vol. 1235, No. 1, p. 012042,
IOP Publishing, June 2019.
[2] Islam, M.A., Arafath, M.Y. and Hasan, M.J., “Design of
DDR4 SDRAM controller”, In 8th International
Conference on Electrical and Computer Engineering,
IEEE, pp 148- 151, December 2014.
[3] Nikologiannis, A., Papaefstathiou, I., Kornaros, G. and
Kachris, C., “An FPGA-based queue management system
for high speed networking devices”, Microprocessors
and Microsystems, 28(5-6), pp.223-236, 2004.
[4] Agarwal, A., Iskander, C. and Shankar, R., “Survey of
network on chip (noc) architectures & contributions”,
Journal of engineering, Computing and Architecture,
3(1), pp.21-27, 2009.
[5] Hecht, R., Kubisch, S., Herrholtz, A. and Timmermann, D,
“Dynamic Reconfiguration with hardwired Networks-
on-Chip on future FPGA”, In International Conferenceon
Field Programmable Logic and Applications,2005,IEEE,
pp 527-530, August 2005.
[6] Park, D., Nicopoulos, C., Kim, J., Vijaykrishnan, N. and
Das, C.R., “Exploring faulttolerant network-on-chip
architectures”, In International Conference on
Dependable Systems and Networks (DSN'06), IEEE, pp
93-104, June 2006.
[7] Tsai, W.C., Lan, Y.C., Hu, Y.H. and Chen, S.J., “Networkson
chips: structure and design methodologies”, Journal of
Electrical and Computer Engineering, 2012.
[8] Corbetta, S., Rana, V., Santambrogio, M.D. and Sciuto, D.,
“A light-weight Network-onChip architecture for
dynamically reconfigurable systems”, In 2008
International Conference on Embedded Computer
Systems: Architectures, Modeling, and Simulation,IEEE,
pp 49-56, July 2008.
[9] Kim, J., Park, D., Nicopoulos, C., Vijaykrishnan, N. and
Das, C.R., “Design and analysis of an NoC architecture
from performance, reliability and energy perspective”,
In 2005 Symposium on Architectures for Networking
and Communications Systems (ANCS), IEEE, pp. 173-
182, October 2005.
[10] Valinataj, M., Mohammadi, S. and Safari, S., “Fault-aware
and reconfigurable routingalgorithmsforNetworks-on-
Chip”, IETE Journal of Research, 57(3), pp 215-223,
2011.
[11] Shilpa. K. Gowda, “On-Chip Network Design with
Dynamic Reconfiguration”, International Journal of
Scientific & Engineering Research, Volume VI, Issue VII,
pp 1326-1329, July 2015.
[12] Paramasivam, K., "Network On-Chip and Its Research
Challenges", ICTACT Journal on Microelectronics(1.02),
2015.
[13] Seema Sreekumar, "Network on Chip configuration and
associated Algorithms", Journal of Emerging
Technologies and Innovative Research (JETIR), Volume
VIII, Issue III. (ISSN-2349-5162), March 2021.
[14] Idris, M.Y.I., Bakar, S.A., Tamil, E.M., Razak, Z. and Noor,
N.M, “High-speed shortest path co-processor design”,
IEEE In 2009 Third Asia International Conference on
Modelling & Simulation, pp 626-631, May 2009.
[15] Sao, P., Agrawal, R., “A Survey of Router Design
Algorithms for NoC Platform and their Performance
Analysis”, International Journal of Latest Technology in
Engineering, Management & AppliedScience(IJLTEMS),
Volume II, Issue VIII, pp 73- 76, 2013.
[16] Cesário, W.O., Lyonnard, D., Nicolescu, G., Paviot, Y.,Yoo,
S., Jerraya, A.A., Gauthier, L. and Diaz-Nava, M.,
“Multiprocessor SoC platforms: a component-based
design approach”, IEEE Design & Test of Computers,
19(6), pp 52-63, 2002.
[17] Sgroi, M., Sheets, M., Mihal, A., Keutzer, K., Malik, S.,
Rabaey, J. and SangiovanniVincentelli, A., “Addressing
the system-on-a-chip interconnect woes through
communication-based design”, In Proceedings of the
38th Design Automation Conference (IEEE Cat. No.
01CH37232), pp 667-672, June 2001.
[18] Yoo, Sungjoo, Gabriela Nicolescu, Damien Lyonnard,
Amer Baghdadi, and Ahmed A. Jerraya, "A generic
wrapper architecture for multi-processor SoC co-
simulation and design.", In Proceedings of the ninth
international symposium on Hardware/software co-
design, pp 195-200, 2001.
[19] Bahn, J.H., Lee, S.E. and Bagherzadeh, N., “On design and
analysis of a feasible network-on-chip (noc)
architecture”, In Fourth International Conference on
InformationTechnology(ITNG'07),IEEE,pp1033-1038,
April 2007.

[20] Chawade, S.D., Gaikwad,M.A. andPatrikar,R.M.,“Review
of XY routing algorithm for network-on-chip
architecture”, International Journal of Computer
Applications(IJCE), 43(21), pp.975-8887, 2012.
[21] Sayankar, B.B., Agrawal, P. and Dorle, S.S., “Routing
Algorithms for NoC Architecture:ARelativeAnalysis”,In
2013 6th International Conference on Emerging Trends
in Engineering and Technology, IEEE, pp 105-106,
December 2013.
[22] Sani, A, "Multistage switching hardware and software
implementations for student experiment purpose", In
IOP Conference Series: Materials Science and
Engineering, Vol. 309, no. 1, p. 012097, 2018.
[23] Talwar, B. and Amrutur, B, “Traffic engineered NoC for
streaming applications”, Microprocessors and
Microsystems, 37(3), pp 333-344, 2013.
[24] Kendaganna, S.S., Jatti, A. and Uma, B.V., “Design and
implementation of secured agent based NoC using
shortest path routing algorithm”, International Journal
of Electrical and Computer Engineering(IJECE), 9(2),
p.950, 2019.
[25] Yang, Z.J., Kumar, A. and Ha, Y., “An area-efficient
dynamically reconfigurablespatial divisionmultiplexing
network-on-chip with static throughput guarantee”, In
2010 International Conference on Field-Programmable
Technology, IEEE, pp. 389-392, December 2010.
BIOGRAPHIES
Preethi T S,
Assistant Professor, Dept. of ECE
Sapthagiri College of Engineering
Email :
ts.preethi@gmail.com
Jayanth K Holalkeri,
Student, Dept. of ECE
Email:
jayanth.holalkeri@gmail.com
Manasa Dev G J,
Email:
manasadev30@gmail.com
Manoj Mahadev Biradar,
Email:
mmbiradar1906@gmail.com
Nagesh R Shanbhag,
Email:
nageshrshanbhag619@gmail.com

Design and Performance Analysis of 8 x 8 Network on Chip Router

Recommended

More Related Content

Similar to Design and Performance Analysis of 8 x 8 Network on Chip Router (20)

More from IRJET Journal (20)

Recently uploaded (20)

Design and Performance Analysis of 8 x 8 Network on Chip Router