Fault Injection Approach for Network on Chip

IJSRD - International Journal for Scientific Research & Development| Vol. 1, Issue 2, 2013 | ISSN (online): 2321-0613
All rights reserved by www.ijsrd.com 272
Fault Injection Approach for Network on Chip
Tapas Patel1
1
Gujarat Technological University Ahmedabad, Gujarat.
Abstract— Packet-based on-chip interconnection networks,
or Network-on-Chips (NoCs) are progressively replacing
global on-chip interconnections in Multi-processor System-
on-Chips (MP-SoCs) thanks to better performances and
lower power consumption. However, modern generations of
MP-SoCs have an increasing sensitivity to faults due to the
progressive shrinking technology. Consequently, in order to
evaluate the fault sensitivity in NoC architectures, there is
the need of accurate test solution which allows evaluating the
fault tolerance capability of NoCs. Presents an innovative
test architecture based on a dual-processor system which is
able to extensively test mesh based NoCs. The proposed
solution improves previously developed methods since it is
based on a NoC physical implementation which allows
investigating the effects induced by several kind of faults
thanks to the execution of on-line fault injection within all
the network interface and router resources during NoC run-
time operations. The solution has been physically
implemented on an FPGA platform using a NoC emulation
model adopting standard communication protocols. The
obtained results demonstrated the effectiveness of the
developed solution in term of testability and diagnostic
capabilities and make our solutions suitable for testing large
scale.
I. INTRODUCTION
Nevertheless, NoCs are characterized by high performances
and low power consumption; one of the open problems that
afflict research activities on NoCs is the evaluation of their
fault tolerance capabilities. Indeed, as specified in, faults are
increasing happening and a very large set of effects must be
considered in new generation of NoCs. The larger
occurrence of fault is mostly due to the technology scaling
of the new chip generations that result more susceptible to
fault appearance induced by different phenomena such as
single-event upsets, cross-talk, age-related degradation or
process variability. Generally, test method of NoC
infrastructures address two issues: testing the switch blocks
and testing the interconnection segments layout including
the logic routers logic resources. Different testing
techniques have been proposed in order to evaluate fault
effects on NoC. Several NoC test methods have been
focused on testing of the functional IP cores using Test
Access Mechanism (TAM). Other authors assumed specific
fault model for NoC fabric and subsequently adopting it to
test the data transportation and the functional blocks.
Vice versa, dedicated TAM based on specific on-chip
network is adopted by functional test solutions on SoCs
multi-cores. The solution of our work improves previously
proposed method by developing a flexible and accurate fault
injection environment, which can be adopted in order to
evaluate the fault tolerance capability of different NoC
architectures. The main advantage of the proposed solution
rely on the possibility to apply different fault models that
can emulate the effective faults affecting NoC architectures,
besides the proposed solutions has a full controllability an
observability of the NoC under test, since interconnections
values and routers functional behavior can be directly
observed during the test operations, feature which is
extremely reduced for test solutions applied directly to the
manufactured chip, since NoC interconnections are deeply
embedded and spread across the chip, therefore adding of
probe interconnections results inapplicable. The
implementation of our solution relies on the main idea
illustrated in Figure 1.
The fault injection method we developed is innovative since
it is based on single reconfigurable chip, such a Static RAM-
based Field Programmable Gate Array (SRAM-based
FPGAs) where thanks to a suitable architecture, faults can
be injected and evaluated. As illustrated in Figure 1, the
architecture consists of two processors: the processor 1 is
devoted to the application of the test pattern to the NoC
under test while the processor 2 performs the injection of the
faults. The execution of the fault injection does not require
the insertion of intrusive module into the NoC architecture,
since modifying the configuration memory bit of the FPGA
device thus physically inserting the desired fault performs
the injection. This operation is performed thanks to the
availability of the Configuration Access Port which is
located internally to the device and can be controlled by a
logic core; in our case by the processor 2.
Thanks to our solution the fault injection into the NoC
architecture can be performed without intrusive modules
affecting the real behavior and with optimal performances,

Fault injection approach for network on chip
(IJSRD/Vol. 1/Issue 2/2013/0054)
since the working frequency of the NoC is not drastically
degraded by the fault insertion. Figure1. The main
architectural scheme of the proposed approach. The rest of
this paper is organized as follows. In Section II we present
the main background concepts behind the implementation
and testing of NoC. The description of the fault injection
method, the developed architecture and the NoC fault
models are presented in Section III. Section IV describes the
testing routines developed for testing the router logic blocks
and the NoC switches and interconnections. Fault injection
results performed on a real NoC case study are presented in
Section V. Finally, conclusions and future works are drawn
in Section IV.
II. BACKGROUND
A Network on Chip is an interconnection
architecture consisting of a 2-D mesh of routers each of
which connected to a set of interconnection resources. An
example of NoC structure is illustrated by the NoC under
Test box in Figure 1. Each router may be connected to its
four neighboring routers and to a set of other routers not
located on its same row and column. Various NoC
architectures have been proposed basing on the 2-D matrix
of resources. ll of them are characterized by several aspects
including topology, routing algorithms, switching and flow
control mechanisms. Several NoC architectures have been
proposed in the past. The typical NoC design is based on the
data exchange between the functional IP cores in the form of
data packets. Depending on the adopted communication
scheme, data packets are transmitted through routers and
interconnections from the source to the destination IP port.
Different schemes have been proposed for NoC
communication such as communication switching, virtual
cut-through and wormhole switching. While communication
switching is more applicable to small-size networks, virtual
cut-through and wormhole switching are appropriate for
large MPSoCs since data have to traverse large distances and
packets are buffered by halfway routers on the routing path
from the source to the destination. NoC architecture embeds
two main resources: interconnections and routers.
Interconnections are wire segments that link the various
routers inside to the NoC array architecture, while routers are
the most complex part of the NoC architecture. The details of
a generic router are illustrated in the following section.
The architecture of a NoC router, which an example
is illustrated in Figure 2, consists of a set of FIFO buffers
forming the input and output data ports, a crossbar switch
controlled by a logic circuitry that allow to implement the
transportation methodology. A router generally consists of
multiple buffer stages connected through the input and
output ports to the routers placed on the same row (SR),
column (SC) or on other locations (OL) of the considered
router. Each input/output buffer stage form an input/output
queue which is internally connected to a crossbar switch.
The crossbar switch is the core of the router since it allows to
form the links between the input and output queues in all the
possible combinations. The connection between the input
and output queues are managed by an arbitration circuitry,
which on the basis of the adopted communication protocol
routes the data packets.
III. THE PRAPOSED FAULT INJECTION METHOD
The environment of the proposed method consists of a host-
PC which communicates to the FPGA board through a
JTAG connection and a RS-232 serial cable. The method
consists of the preliminary generation of the fault list
performed on the host-PC by parsing the NoC net list and
creating all the possible fault locations according to the used
synthesis model. The fault locations are consequently
transferred through the JTAG cable to the processor 2
dedicated memories. The fault injection execution flow is
illustrated in Figure 3. It consists on the following principal
modules:
Fault Application: the fault application module isA.
executed by the processor 2 and consists on all the
operations dedicated to the insertion of a fault into the NoC
architecture. At first a fault location is read from the
memory module, secondly, the fault location is converted
into the correspondent FPGA’s configuration memory
coordinates that control the selected NoC resource. Finally,
the processor 2 accesses to the FPGA’s internal
configuration access port facility and activate the
reconfiguration of the configuration memory bits related to

the selected coordinates. During the modification of the
FPGA’s configuration memory, the NoC functionality is
temporary freeze. This has not any impact on performance
degradation of the NoC, since the freezing is synchronized
with the microprocessor control clock.
Configuration memory coordinates: The configurationB.
memory coordinates are the physical identification of the
FPGA’s configuration memory bits. Each configuration
memory coordinate is associated to a correspondent
sequence of FPGA’s configuration memory bits.
FPGA configuration memory: The FPGA’sC.
configuration memory contains millions of SRAM cells that
configure the behavior of the circuit mapped on the FPGA
device. The main idea of the present work consists in acting
on the configuration memory cells related to the behavior of
the NoC architecture in order to insert a fault. Therefore, in
order to perform the right fault injection it is necessary to
know the exact correspondence between each configuration
memory bit and the controlled FPGA’s resource. The
modification of the configuration memory is performed by
the Internal Configuration Access Port (ICAP) controlled by
the software running on the Processor 2. Through the ICAP
port it is possible to modify selectively a single FPGA
configuration frame in a small amount of time (dependent
from the kind of FPGA device adopted).
Patterns generator: The patterns generator aims atD.
generating the test stimuli applied to the NoC. The stimuli
are generated in the form of data packets transmitted from a
source to a destination IP considering a NoC using a
Wormhole transport methodology. The patterns are
generated by the software running on the Processor 1. Each
pattern consists of a signature composed by the
identification of the source and destination IPs and by the
complete data packet. The application of the test patterns is
executed by the Processor 2 by executing the following
steps: at first the signature is decoded individuating the
source IP; secondly the Processor 2 access through the On-
Chip Peripheral Bus (OPB) to the source IP and send the
data packet; third, the signature is decoded individuating the
destination IP and finally, the Processor 2 reads through the
OPB bus the data packet received by all the IP (except to the
source one).
Fault classification: The fault classification is executedE.
by the Processor 2. The first action consists on capturing all
the outputs of the IP destination ports. Secondly, the pattern
signature is decoded individuating the destination IP.
Network-on-Chip under Test: It consists on theF.
architecture of the NoC under evaluation. The NoC model
should be described in HDL language in order to facilitate
the integration in our environment. The integration into the
testing environment is performed by linking each NoC’s
port to the platform IP port.
IV. SIMULATION
Simulation without fault injectionA.
Simulation for testing with fault injectionB.
V. ADVANTAGE
The approach has several advantages: flexibility,
observability and test speed. The approach has been
implemented on an FPGA platform and a real NoC
architecture using standard communication protocol has
been tested. The experimental results demonstrated the
effectiveness of the developed approach and make our
approach suitable for testing large scale NoC design. ICs
have been designed with dedicated point-to-point
connections, with one wire dedicated to each signal. For
large designs; in particular, this has several limitations from
a physical design viewpoint. The wires occupy much of the
area of the chip, and in nanometer CMOS technology,
interconnects dominate both performance and dynamic
power dissipation, as signal propagation in wires across the
chip requires multiple clock cycles. NoC links can reduce
the complexity of designing wires for predictable speed,
power, noise, reliability, etc., Multi-core processor systems;
a network is a natural architectural choice. A NoC can
provide separation
Between computation and communication, support
modularity and IP reuse via standard interfaces,
handle synchronization issues, serve as
a platform for system test, and, hence, increase
engineering productivity.
VI. APPLICATION MULTI-CORE
PROCESSOR SYSTEMS: A NETWORK IS A NATURAL
ARCHITECTURAL CHOICE
Test approach based on a dual-processor system
implemented on SRAM based FPGA which is able to test

mesh-based Network on Chips. Applicable for large scale
network on chip.
REFERENCES
[1] S. Borkar, “Designing reliable systems from unreliable
components: the challenges of transistor variability
and degradation”, IEEE Micro, Vol. 25, No. 6, pp.
10-16, November 2005.
[2] L. Chunsheng, V. Iyengar, S. Jiangfan, E. Cota, “Power-
Aware Test Scheduling in Network-on-Chip
Using Variable-Rate On Chip Clocking”, IEEE
International Very Large Scale of Integration Test
Symposium , 2005, pp. 349 - 354
[3] B. Vermeulen, J. Dielissen, K. Goossens, C. Ciordas,
“Bringing communications networks on a chip: test
and verification implications”, IEEE
Communications Magazine, Volume 41, September,
2003, pp. 74 - 81.
[4] A. Kohler, G. Schley, M. Radetzki, “Fault
Tolerant Network on Chip Switching With Graceful
Performance Degradation”, IEEE Transactions on
Computer-Aided Design of Integrated Circuits and
Systems, Vol. 29, No. 6, June 2010, pp. 883 - 896.
[5] M. Nahvi, A. Ivanov, “Indirect Test Architecture for
SoC Testing”, IEEE Transactions on Computer-Aided
Design of Integrated Circuits and Systems, Volume 23,
Issue 7,July 2004, pp. 1128 - 1142.
[6] Xilinx Product Specification, “Logic ORE IP XPS
HIWCAP”, Product Spec., DS586, June 22, 2011.
[7] T. Bjerregaard and S. Mahadevan, “A survey of
research and practices of network-on-chip”, ACM
Computing Surveys, Vol. 38, pp. 1-51, March, 2006.
[8] J. Duato, S. Yalamanchili, L. Ni, “Interconnection
Networks - An Engineering Approach”, Morgan
Kaufmann, 2002.
[9] C. Grecu, P. Pande, A. Ivanov, R. Saleh, “Timing
Analysis of Network-on-chip Architectures for MP-
SoC
[10] Platforms”, Elsevier Microelectronics Journal, XX [10]
L. Benini, D. Bertozzi, “Xpipes: Network on-chip
architecture for gigascale systems-on-chip”,
IEEE Circuits and Systems Magazine, Volume 4,
Issue 2, 2004, pp. 18-31.
[11] S. Pasricha, Y. Zou, D. Connors, H. J. Siegel,
“OE+IOE: A novel turn model based fault tolerant
routing scheme for networks-on-chip”,
IEEE/ACM/IFIP International Conference on
Hardware/Software Codesign and System Synthesis,
2010, pp. 85 - 93
[12] H .Zimmer, A. Jantsch, “A fault model notation
and error-control scheme for switch-to-switch buses
in a network-on-chip”, IEEE/ACM/IFIP
International Conference on Hardware/Software
Codesign and System Synthesis, 2003, pp. 188 - 193.
[13] E. Cota,”Power aware NoC Reuse on the Testing of
Core-Based Systems”, proceedings of the
IEEE International Test Conference 2003, pp. 612 -
621.
[14] Yung-Chang Chang, Ching-Te Chiu, Shih-Yin Lin,
Chung-Kai Liu, “On the Design and Analysis of
Fault Tolerant NoC Architecture Using Spare
Routers”, International DAC Conference, 2011, pp. 431
- 436.
[15] L.Sterpone, M. Violante, “A new analytical approach
to estimate the effects of SEUs in TMR
architectures implemented through SRAM-based
FPGAs”, IEEE Transactions on Nuclear Science, Vol.
52, Issue 6, 2005,pp. 2217 - 2223.
[16] A. Perreira, J. P. Teixeir and M. Santos, “A Novel
Approach to FPGA-based Hardware Fault Modeling
and Simulation”, IEEE International Workshop on
Design and Diagnostic of Electronic Circuits and
Systems, 2003, pp. 17 - 24.
[17] Reference to the Network-on-Chip emulator on
www.opencores.org

Fault Injection Approach for Network on Chip

More Related Content

What's hot (20)

Similar to Fault Injection Approach for Network on Chip (20)

More from ijsrd.com (20)

Recently uploaded (20)

Fault Injection Approach for Network on Chip