Overview of Radiation Hardening Techniques for IC Design
With the development of astronautic techniques, the radiation effects on Integrated Circuits (ICs) have been cognized by people. Environments with high levels of ionizing radiation create special design challenges for ICs. To ensure the proper operation of such systems, manufacturers of integrated circuits and sensors intended for the military aerospace markets adopt various methods of radiation hardening. An overview of radiation hardening techniques for IC design is given in this study. First, seven major radiation damage sources, two fundamental damage mechanisms, five sorts of end-user effects and six types of single-event effects are introduced, followed by the brief introduction of radiation hardening techniques. Secondly, typical physical radiation hardening techniques are introduced. Thirdly, typical logical radiation hardening techniques are introduced. Fourthly, we propose our radiation hardening scheme for microwave power amplifier chip design. Here, a Radio-Frequency (RF) Power Amplifier (PA) is a sort of electronic amplifier employed to convert a low-power radio-frequency signal into a larger signal of significant power, typically for driving the antenna of a transmitter. Finally, we concluded the whole study.
Received: January 28, 2010;
Accepted: March 07, 2010;
Published: June 10, 2010
Environments with large amount of ionizing radiation create special design
challenges for integrated circuits, typically for RF power amplifier chips (Barbara
et al., 1990). A single charged particle can knock thousands of electrons
loose, resulting in electronic noise and signal spikes. In the case of digital
circuits, this can lead to results that are inaccurate or unintelligible. It
is a particularly serious problem in designing artificial satellites, spacecraft,
military aircraft, nuclear power stations and nuclear weapons. Typical sources
of exposure of electronics to ionizing radiation are the Van Allen radiation
belts for satellites, nuclear reactors in power plants for sensors and control
circuits, residual radiation from isotopes in chip packaging materials, cosmic
radiation for spacecraft and high-altitude aircraft and nuclear explosions for
potentially all military and civilian electronics (Holmes-Siedle
and Adams, 2002). Seven familiar sources are: (1) Cosmic rays (Gaisser,
1990). They consist of approximate 85% protons, 14% alpha particles and
1% heavy ions with X-ray radiation. Most effects result from particles with
energies between 108 and 2x1010 eV. The atmosphere filters
most of them, so they are mainly considered for spacecraft and high-altitude
aircraft. (2) Solar particle events (Lantos and Fuller,
2003). They come from the sun and consist of a large flux of high-energy
protons and heavy ions accompanied by X-ray radiation. (3) Van Allen radiation
belts (Snyder, 1959). They include electrons and protons
trapped in the geomagnetic field. Depending on the actual conditions of the
sun and the magnetosphere, the particle flux in the regions farther from the
Earth can vary wildly. They mainly affect satellites. (4) Secondary particles
(Badhwara et al., 1995). They result from interaction
of other kinds of radiation with structures around the electronic devices. (5)
Nuclear reactors. They produce Gamma radiation and neutron radiation, which
can affect sensor and control circuits in nuclear power plants. (6) Nuclear
explosions. They produce a short and extremely intense surge through the entire
spectrum of electromagnetic radiation, an electromagnetic pulse (EMP), neutron
radiation and a flux of both primary and secondary charged particles. In case
of a nuclear war they pose a potential concern for all civilian and military
electronics. (7) Chip packaging materials (Baumann et
al., 1995). They are a kind of insidious source of radiation that was
found to cause soft errors in new DRAM chips in the 1970s. Traces of radioactive
elements in the chip packaging will produce alpha particles, which discharge
occasionally some of the capacitors used to store the DRAM data bits. These
effects have been reduced today by using purer packaging materials and employing
error-correcting codes to detect and often correct DRAM errors.
With the development of astronautic techniques, the radiation effects have
been cognized by people. The research about the radiation effects on semiconductors
has been carried out since 1960s. Later, the research about the radiation effects
on electronic elements and circuits was carried out also. With regard to radiation
effects on electronics, two fundamental damage mechanisms (Van
Lint et al., 1980) can be described as follows. One is called lattice
displacement. It is caused by neutrons, protons, alpha particles, heavy ions
and very high energy gamma photons. They change the arrangement of the atoms
in the crystal lattice, create lasting damages, increase the number of recombination
centers, deplete the minority carriers and worsen the analog properties of the
affected semiconductor junctions. This damage is especially important for bipolar
transistors that depend on minority carriers in their base regions and increased
losses caused by recombination will cause loss of the transistor gain. The other
is called ionization effects. They are mainly caused by charged particles, including
the ones with energy too low to cause lattice effects. They are usually transient,
creating glitches and soft errors, but can lead to destruction of the device
if they trigger other damage mechanisms. Gradual accumulation of holes in the
oxide layer in MOSFET transistors results in performance degradation, up to
device failure when the dose is high enough. The effects can vary wildly dependent
on the type of radiation, total dose, the radiation flux, combination of types
of radiation and even the kind of the device load, which makes thorough testing
difficult and time-consuming and require a lot of test samples.
Based on above damage mechanisms, the resultant end-user radiation effects
can be characterized into following five groups: (1) Neutron effects (Arimura,
1982). A neutron interacting with the semiconductor lattice will displace
its atoms. This leads to an increase in the count of recombination centers and
deep-level defects, a decrease in the lifetime of minority carriers, which influences
bipolar devices more than CMOS ones. There is also the risk of induced radioactivity
caused by neutron activation, which is a major source of noise in high energy
astrophysics instruments. Induced radiation, together with residual radiation
from impurities in materials, can introduce all sorts of single-event problems
during the devices lifetime. GaAs LEDs are very sensitive to neutrons.
The lattice damage affects the frequency of crystal oscillators. Kinetic energy
effects of charged particles also belong to this category. (2)Total ionizing
dose effects (Pease, 2003). It is the cumulative damage
of the semiconductor lattice introduced by ionizing radiation over the exposition
time. In CMOS devices, the radiation creates electron-hole pairs in the gate
insulation layers, which introduce photocurrents during their recombination
and the holes trapped in the lattice defects in the insulator create a persistent
gate biasing and affect the transistors threshold voltage, making the
N-type MOSFET transistors easier and the P-type ones more difficult to switch
on. The accumulated charge can be high enough to keep the transistors permanently
open (or closed), leading to device failure. Crystal oscillators are somewhat
sensitive to radiation dose, which alters their frequency, but the sensitivity
can be greatly reduced by using swept quartz. (3) Transient dose effects (Meulenberg
et al., 1988). It refers to the short-time high-intensity pulse of
radiation, typically occurring during a nuclear explosion. The high radiation
flux creates photocurrents in the entire body of the semiconductor, which makes
transistors randomly open and alters logical states of flip-flops and memory
cells. Permanent damage may occur if the pulse duration lasts too long or if
the pulse causes junction damage or causes a latchup. Crystal oscillators may
stop oscillating during the flash due to prompt photoconductivity induced in
quartz. (4) System-Generated EMP effects (SGEMP) (Higgins
et al., 1978). They are caused by the radiation flash traveling through
the equipment, which causes local ionization and electric currents in the material
of the chips, circuit boards, cables and cases. (5) Single-Event Effects (SEE)
(Dodd, 2005). They mostly affect digital devices and
the following paragraph gives an overview for SEE.
Among above effects, single-event effects (SEE), mostly affecting only digital
devices, were not studied extensively until relatively recently. When a high-energy
particle travels through a semiconductor, it leaves an ionized track behind.
This ionization may introduce a highly localized effect similar to the transient
dose one. SEE are important for electronics in satellites, aircraft and other
both civilian and military aerospace applications. SEE can be classified into
following 6 categories: (1) Single-Event Upsets (SEU) (Kuznetsov
and Nymmik, 1996). They are state changes of memory or register bits caused
by a single ion interacting with the chip. In some very sensitive devices, a
single ion may introduce a Multiple-Bit Upset (MBU) in several adjacent memory
cells. SEUs can become Single-Event Functional Interrupts (SEFI) when they upset
control circuits, which would then require a reset or a power cycle to recover.
(2) Single-Event Latchup (SEL) (Schrimpf et al.,
2007). It may occur in any chip with a parasitic PNPN structure. A heavy
ion or a high-energy proton passing through one of the two inner-transistor
junctions can turn on the thyristor-like structure, which will stay shorted
until the device is power-cycled. As the effect can occur between the power
source and substrate, destructively high current may be involved and the part
may fail to work. (3) Single-Event Transient (SET) (Adell
et al., 2005). It occurs when the charge collected from an ionization
event discharges in the form of a spurious signal traveling through the circuit.
In fact, this is the effect of an electrostatic discharge. (4) Single-event
snapback (Walsh et al., 2001). It is similar
to SEL but not requiring the PNPN structure. It can be induced in N-channel
MOS transistors switching large currents, when an ion hits near the drain junction
and causes avalanche multiplication of the charge carriers. The transistor then
opens and stays opened. (5) Single-event induced burnout (SEB) (Albadri
et al., 2006). It may occurs in power MOSFETs if the substrate right
under the source region gets forward-biased and the drain-source voltage is
higher than the breakdown voltage of the parasitic structures. The resulting
high current and local overheating then may destroy the device. (6) Single-Event
Gate Rupture (SEGR) (Badila et al., 2001). It
was observed in power MOSFETs if a heavy ion hits the gate region while a high
voltage is applied to the gate. A local breakdown then occurs in the insulating
layer of silicon dioxide, causing local overheat and destruction of the gate
region. It can happen even in EEPROM cells during writing or erasing if the
cells are subjected to a comparatively high voltage.
To ensure the proper operation of such systems, manufacturers of integrated
circuits and sensors intended for the military aerospace markets adopt various
methods of radiation hardening. Radiation hardening (Johansson,
1977) is a technique to design and test electronic components and systems
to make them resistant to damage or malfunctions caused by ionizing radiation
such as particle radiation and high-energy electromagnetic radiation, which
would be encountered in the outer space, high-altitude flight, around nuclear
reactors, or during nuclear accidents or nuclear warfare. The resulting systems
are called to be radiation-hardened or rad-hard. Most radiation-hardened chips
are based on their commercial equivalents, with some manufacturing and design
variations that reduce the susceptibility to interference from electromagnetic
radiation. Due to the extensive development and testing required to design a
radiation-tolerant microelectronic chip, radiation-hardened chips tend to lag
behind the cutting-edge of developments.
Since, the early 1980s, the Defense Advanced Research Projects Agency (DARPA)
of America has started the research on digital radiation hardening techniques
for digital GaAs circuits (Naber, 1995). At the same time,
the Department of Defense (DoD) of America also proposed a project to design
microwave/millimeter wave single-chip integrated circuits in order to develop
GaAs microwave circuits and this project was supported by the Strategic Defense
Initiative (SDI) plan, which mainly aims at the application of elements in the
outer space. Therefore, with the rapid development of GaAs integrated circuits,
the research on total ionizing dose effects and neutron effects has been deeply
researched. GaAs has some electronic properties which are superior to those
of silicon. It has a higher saturated electron velocity and higher electron
mobility, allowing transistors made from it to function at frequencies in excess
of 250 GHz. Another advantage of GaAs is that it has a wide bandgap, which means
that it is highly resistive to ionization effects. Combined with the high dielectric
constant, this property makes GaAs a very good electrical substrate and unlike
Si provides natural isolation between devices and circuits. Thus, as a wide
direct band gap material with high breakdown voltage and resulting resistance
to radiation damage, GaAs is an excellent material for space and optical windows
in high power applications. Currently, aiming at various radiation effects,
people mainly provide solutions in improving material performance and element
structure. For example, we can adopt GaAs process because it can reduce the
total ionizing dose effects. Alternatively, we can adopt the HBT (Heterojunction
Bipolar Transistor) process (Torvik et al., 2000)
because its special structure avoids the production of parasitical BJT (Bipolar
Junction Transistor) and makes the work area be far from the surface of elements
and thus it can resist to much more cosmic radiation. Basically speaking, the
radiation hardening techniques can be classified into two categories, i.e.,
physical solutions and logical solutions. In the remainder of this study, we
will overview these two kinds of hardening techniques separately and then propose
our radiation hardening scheme in power amplification chip design.
PHYSICAL RADIATION-HARDENING TECHNIQUES
Physical radiation-hardening techniques use various physical means, such as using insulating substrates, utilizing bipolar integrated circuits, adopting radiation-tolerant SRAM, etc., to realize the hardening purpose.
Hardened chips are often manufactured on insulating substrates instead of the usual semiconductor wafers. Silicon on Insulator (SOI) and Silicon on Sapphire (SOS) are commonly adopted.
Silicon on Insulator (SOI) technique (Simoen et al.,
2007) adopts a layered silicon-insulator-silicon substrate instead of conventional
silicon substrates in semiconductor manufacturing, especially microelectronics,
to reduce parasitic device capacitance and thereby improve its performance.
The differences between SOI-based devices and conventional silicon-built devices
lie in that the silicon junction is above an electrical insulator. The first
industrial implementation of SOI was announced by IBM in August 1998. The benefits
of the SOI technique compared to conventional silicon processing include: (1)
Lower parasitic capacitance due to isolation from the bulk silicon. (2) Resistance
to latchup due to complete isolation of the n-and p-well structures. From a
manufacturing viewpoint, SOI substrates are compatible with most conventional
processes. The primary barrier to SOI implementation is the drastic increase
in substrate cost, which contributes an estimated 10-15% increase to total manufacturing
Silicon on Sapphire (SOS) (Roig et al., 2004)
is a hetero-epitaxial process for integrated circuit manufacturing consisting
of a thin layer of silicon grown on a sapphire wafer. SOS is part of the Silicon
on Insulator (SOI) family of CMOS technologies. The SOS is mainly used in aerospace
and military applications because of its inherent resistance to radiation. The
first advantage of sapphire lies in that it is an excellent electrical insulator,
preventing stray currents caused by radiation from spreading to nearby circuit
elements. The second advantage of silicon on sapphire over exotic technologies
is that it is manufactured in the same factories that produce common bulk silicon
wafers. A further advantage is that, because of its better performance, it can
be manufactured in a less advanced factory than similar devices in bulk silicon.
Where silicon on sapphire has disadvantages over bulk silicon is that it is
by nature a more complex process. Sapphire substrates are expensive. SOS has
seen little commercial use to date because of difficulties in fabricating the
very small transistors used in modern high-density applications. They are physically
heavy, causing problems with manufacturing machines not designed for their mass.
With regards to radiation tolerance, while normal commercial-grade chips can withstand between 5 and 10 krad, space-grade SOI and SOS chips can survive doses many orders of magnitude greater. At one time many 4000 series chips were available in radiation-hardened versions.
Bipolar integrated circuits:
Bipolar integrated circuits contain Bipolar Junction Transistors (BJT) as
their principle elements. A Bipolar Junction Transistor (BJT) is a three-terminal
electronic device constructed of doped semiconductor material and may be used
in amplifying or switching applications. Bipolar transistors are so named because
their operation involves both electrons and holes. Charge flow in a BJT is due
to bidirectional diffusion of charge carriers across a junction between two
regions of different charge concentrations. This mode of operation is contrasted
with unipolar transistors, such as field-effect transistors, in which only one
carrier type is involved in charge flow due to drift. By design, most of the
BJT collector current is due to the flow of charges injected from a high-concentration
emitter into the base where they are minority carriers that diffuse toward the
collector and so BJTs are classified as minority-carrier devices. A new silicon
power device concept based on the Super Junction (SJ) principle for power electronics
in a broad spectrum of consumer, industrial and other energy conversion applications
is presented by Bauer (2004). This new concept can help
to sustain the trend towards ultra low loss switching the past, present and
future dominant driving force in the development of silicon high power switches.
The Super Junction Bipolar Transistor (SJBT) shares many similarities with the
super junction MOSFET. After several decades of development, the GaAs Pseudomorphic
High Electron Mobility Transistor (PHEMT) has emerged as a high performance,
low cost and manufacturable device. Not only does it exhibit excellent noise
properties, but it is also the premier power device for frequencies ranging
from low microwaves through millimeter waves. Accordingly, it is the ideal device
for applications that require high performance, including digital point-to-point
radio, future cellular, LMDS and satellite communication. Based on this consideration,
a 4 W K-band AlGaAs/InGaAs/GaAs Pseudomorphic High Electron Mobility Transistor
(PHEMT) Monolithic Microwave Integrated Circuit (MMIC) high Power Amplifier
(PA) was reported by the authors of this study (Huang et
With regards to radiation tolerance, bipolar integrated circuits generally have higher radiation tolerance than CMOS circuits. It was reported that the low-power Schottky (LS) 5400 series can withstand 1000 krad and many ECL devices can withstand 10 000 krad.
To tolerate radiation, capacitor-based DRAM is often replaced by more rugged
(but larger and more expensive) SRAM. A SRAM device used mainly in the read
state such as configuration RAM in an FPGA can be hardened against radiation
effects to a very high level by adding a large value resistor. A SRAM device
used mainly in the read state is usually written only once on power-up to define
the function of the integrated circuit and in most applications it is never
changed after power up. Recently, He et al. (2008)
presented the practical issues encountered in designing SRAM cell design
on partially depleted SOI, including the effects of floating-body potential
and parasitic bipolar. It also discussed the characteristics of Single-Event
Upsets (SEU) hardening and total-dose radiation hardening of SOI SRAM.
Wide band-gap substrate:
It can give higher tolerance to deep-level defects by using wide band-gap
substrate (Szmidt, 1999). The magnitude of the coulombic
potential determines the bandgap of a material and the size of atoms and electronegativities
are two factors that determine the bandgap. Materials with small atoms and strong,
electronegative atomic bonds are associated with wide bandgaps. Wide Band Gap
Semiconductors such as Gallium Nitride (GaN) and Silicon Carbide (SiC) have
emerged as one of the most promising materials for future electronic components.
They offer tremendous advantages in terms of power capability (DC and microwave),
radiation insensitivity, high temperature and high frequency operation, optical
properties and even low noise capability. Therefore, wide band gap components
are strategically important for the development of next generation spaceborne
systems. Although, impressive results have already been demonstrated, a large
amount of research and development work still remains to be carried out. In
particular, improvements are needed in the quality of the basic crystal materials
through to fabrication of complete devices with enhanced performance and reliability.
Further research work is required to better understand the semiconductor physics,
to improve materials growth and to optimize device performance. In addition,
work is also needed to develop advanced packaging techniques and to understand
the benefits offered to space systems by undertaking detailed application assessment.
Shielding the package against radioactivity:
Obviously, this is an intuitional scheme to reduce exposure of the bare
device. Recently, Miller et al. (2009) studied
the lunar soil as shielding against space radiation. The measurements and model
calculations indicated that a modest amount of lunar soil affords substantial
protection against primary Galactic Cosmic Radiation (GCR) nuclei and Solar
Particle Event (SPE), with only modest residual dose from surviving charged
fragments of the heavy beams. Cherng et al. (2007)
studied two representative spacecraft-shielding materials: aluminum representing
low/medium-Z material and tungsten representing high-Z material. Calculation
results indicate that, for the radiation attenuation required for typical electronics
used in a Jupiter mission, the low-Z material and the low/high-Z combination
are a less-efficient shield per the same areal mass than the high-Z material
in the Jovian radiation environment. When massive shielding >10 g cm-2
is required to protect very radiation-sensitive electronics, then the low- /high-Z
combination is a better shield per the same areal mass.
Shielding the chips by using depleted boron:
The depleted boron consists only of isotope boron-11. Cosmic radiation will
produce secondary neutrons if it hits spacecraft structures; and neutrons cause
fission in Boron-10 if it is present in the spacecraft's semiconductors, producing
a gamma ray, an alpha particle and a lithium ion. The resultant fission products
may then dump charge into nearby semiconductor chip structures, causing data
loss (bit flipping, or single event upset). In radiation hardened semiconductor
designs, one countermeasure is to use depleted boron which is greatly enriched
in Boron-11 and contains almost no Boron-10. Boron-11 is largely immune to radiation
damage and it is a by-product of the nuclear industry. In general, the depleted
boron is used in the borophosilicate glass passivation layer to protect the
chips. Here, borophosphosilicate glass, commonly known as BPSG, is a type of
silicate glass that includes additives of both boron and phosphorus (Kern
and Smeltzer, 1986).
LOGICAL RADIATION-HARDENING TECHNIQUES
Logical radiation-hardening techniques adopt various logical means, such as using error correcting memory, utilizing redundant elements, adopting a watchdog timer etc., to realize the hardening purpose. Finally, the reliability evaluation problem is introduced.
Error correcting memory:
In general, DRAM memory can afford increased protection against soft errors based on error correcting codes. The error-correcting memory, known as ECC (Error Correcting Codes) or EDAC (Error Detection and Correction)-protected memory, is especially suitable for high fault-tolerant applications, such as servers, as well as deep-space applications due to cosmic radiation. It utilizes extra parity bits to check for and possibly correct corrupted data. Since, radiation effects may destroy the memory content even if the system is not accessing the RAM, a so-called scrubber circuit should be used to continuously sweep the RAM. Typically, the following three steps are involved:
||Reading out the data
|| Checking the parity for data errors
|| Writing back any corrections to the RAM
Traditional error-correcting memory controllers adopt Hamming codes, although,
some may use triple modular redundancy (TMD). Interleaving allows us to distribute
the effect of a single cosmic ray that potentially upsets multiple physically
neighboring bits over multiple words by associating neighboring bits to different
words. As long as a Single Event Upset (SEU) is not larger than the error threshold
in any particular word between accesses, it can be corrected and the illusion
of an error-free memory system can be maintained.
Error-correcting schemes have been widely applied in both memory architectures
and communication beginning with Von Neumanns seminal work on repetition
codes. However, state-of-the-art CMOS and disk technologies have very small
error rates that may be only in order of one in a billion and thus rigorous
error correction is not always necessary. Recently, Jeffery
et al. (2004) proposed a 3-level error correcting memory architecture
for nanoscale memory utilizing single- or double-error correcting codes. For
high error rates, however, stronger and multiple error correcting codes such
as BCH codes are required for nano-scale devices (Sun and
Zhang, 2006). Ou and Yang (2004) proposed hardware
design for the decoding and encoding routines of Hamming codes, where the memory
reliability is increased at the cost of only 5ns delay in the memory access
time. Although, Hamming codes are capable of correcting a single error in the
block of physical bits used in the encoding, they become less productive for
high error rates. In practical applications, the BCH (250, 32, 45) code can
provide 99.9956% correctness at 10% bit error rate in memory, but 1 byte error
in every 711 bytes is expected to be defective. In general, if we only use error
correcting codes, we will need very strong and complex error correction codes
resulting in large overhead in area and latency and thus we will lose all the
benefits of using nanoscale memory.
According to above description, besides active error correction through encoding,
we require using defect maps to store the locations of the faulty bits in memory
devices (Vollrath et al., 2001). For reconfigurable
architectures, tile-based memory units have been proposed by storing the defect
map in a distributed fashion (He et al., 2005;
Ziegler and Stan, 2003). However, the drawback of using
defect maps in the bit-level lies in that the storage overhead is generally
very high. To reduce the size of the required defect map, Tahoori proposed a
defect unaware design flow (Tahoori, 2005) that identifies
universal defect free subsets within the partially defective chips, while Wang
et al. (2006) proposed the use of bloom filters for storing defect
maps in nanoscale devices. However, hashing for every bit is computationally
expensive and may significantly increase the number of memory access times.
Therefore, Sun and Zhang (2006) proposed the use of CMOS
memory for storing metadata to identify good parts of the memory based on two
||A two level hierarchy of CMOS and nano-device memory
||A bootstrapping technique to store the reliable block information in some
good part of the non-reliable memory and storing this index in the reliable
We should note that the amount of memory to store the ranges increases with
the sparseness of faulty memory bits. It can be shown that when the error rate
is close to 10%, the number of entries in the list is very large.
In a word, error correcting codes reduce the defect rate of memory at the cost of additional computation and redundancy. For example, strong error correcting codes (e.g., BCH (250, 32, 45)) are computationally expensive. The encoding and decoding delay is very high. In fact, we can use less complex codes such as concatenation of Hamming and TMR (Triple Modular Redundancy) to produce 90% correct blocks in presence of 10% bit error rate.
In engineering, redundancy is the duplication of critical components of a system
so as to enhance system reliability, typically in the case of a backup or fail-safe.
In many safety-critical systems, e.g., fly-by-wire and hydraulic systems in
aircraft, some parts of the control system should be triplicated. An error in
one component can then be out-voted by the other two. In a triply redundant
system, its three sub components must fail before the system fails. Since, each
one seldom fails and is expected to fail independently, the probability that
all three fail is calculated to be extremely small. Redundancy is also known
as the term majority voting systems (Srihari, 1982)
or voting logic. More generally, there are four major forms of redundancy as
||Hardware redundancy, such as DMR (Dual Modular Redundancy)
||Information redundancy, such as error detection and correction methods
||Time redundancy (Lisnianski et al., 2000),
including transient fault detection methods such as alternate logic
||Software redundancy, such as N-version programming (Goseva-Popstojanova
and Grnarov, 1993)
Redundant elements can be used at the system level or the circuit level. At
the system level, three separate microprocessor boards may independently compute
an answer to a calculation and compare their answers. Any system that produces
a minority result will recalculate. Logic may be added to shut down the board
occurring repeated errors. At the circuit level, a single bit may be replaced
with three bits and separate voting logic for each bit to continuously determine
its result. However, this strategy will increase the area of a chip design by
a factor of 5, so it must be reserved for smaller designs. But it has the secondary
advantage that it is also fail-safe in real time. In the event of a single-bit
failure, the voting logic will continue to produce the correct result without
resorting to a watchdog timer. System-level voting between three separate processor
systems will generally need to use some circuit-level voting logic to perform
the votes between the three processor systems.
Recently, Nepal et al. (2006) introduced a new
redundancy element, the MRF reinforce, which achieves significant immunity to
single-event upsets and noise. Myers and Rauzy (2008)
studied the assessment of the reliability of redundant systems with imperfect
fault coverage. They termed fault coverage as the ability of a system to isolate
and correctly accommodate failures of redundant elements. For highly reliable
systems, such as avionic and space systems, fault coverage is in general imperfect
and has a significant impact on system reliability. They reviewed different
models of imperfect fault coverage and proposed efficient algorithms to assess
them separately. Gonzalez and Mazumder (2000) presented
a survey of circuit implementations of redundant arithmetic algorithms in three
||Conventional binary logic circuits, which encode the multivalued
digits of redundant arithmetic into two or more binary digital signals
||Current-mode multiple-valued logic circuits, which directly represent
multivalued redundant digits using non-binary digital current signals
||Heterostructure and quantum electronic circuits, intended for very compact
designs capable of operating at extremely high speeds
For each of the circuits, the operating principle was described and the main
advantages and disadvantages of the approach were discussed and compared.
A watchdog timer can be employed to perform a hard reset of a system unless some sequence is performed that generally indicates the system is alive, such as a write operation from an onboard processor. During normal operations, software schedules a write to the watchdog timer at regular intervals to prevent the timer from running out. If the radiation causes the processor to operate incorrectly, it is unlikely that the software will work correctly enough to clear the watchdog timer. The watchdog eventually times out and forces a hard reset to the system. This is considered as a last resort to other methods of radiation hardening.
Recently, El-Attar and Fahmy (2007) studied the ability
of different watchdog timer systems to recover the system from failure and a
new improved watchdog timer system design was introduced. They first introduced
standard watchdog timers and windowed watchdog timers and then proposed their
sequenced watchdog timers. A standard watchdog timer in its simplest form is
a monostable timer. When the timer reaches its maximum value it changes its
logical state. The system must reset the timer before it reaches maturity. If
the system fails to reset the timer an action is taken whether to change the
state of an output or to immediately restart the system. In order to solve the
problem of fast watchdog resets, the windowed watchdog timer adopt a new supervisory
system based on two timers instead of one. The first Timer has a timeout of
T1 and the second timer has a timeout of T3. The ClearWDT instruction must be
executed within a time window of (T3-T1) to reset both timers, where T3>T1.
The sequenced watchdog timer is an improved design of the windowed watchdog
timer. It requires minor modifications to the ClearDWT instruction. The ClearDWT
instruction is originally an inherent, which means it does not require and operand
to be executed. The instruction is modified to include an operand. Once the
Opcode is Fetched and decoded, the control unit resets the Windowed watchdog
timer. If a slow or fast resets occur the watchdog immediately resets the system.
If the ClearDWT opcode is executed within the safe window then the operand is
compared to the value of the sequenced timer register. If the value matches,
then the system is operating properly. If the value does not match then a faulty
reset occurred within the safe window of the watchdog timer. The sequenced watchdog
timer then resets the whole system.
It should be noticed that, in addition to above hardening techniques, how to
test the reliability of the integrated circuit is also a very important topic.
Recently, a novel approach for MMIC (Monolithic Microwave Integrated Circuits)
reliability testing based on Weibull distribution (Huang
et al., 2009b) was proposed by the authors of this study. We also
proposed a methodology to predict the GaAs MMICs reliability by combining empirical
and statistical methods based on zero-failure GaAs MMICs life testing data (Huang
et al., 2009a). Besides, we investigated the effect of accelerated
factors on MMICs degradation and make a comparison between the Weibull and lognormal
distributions. The method has been used in the reliability evaluation of GaAs
PROPOSED RADIATION-HARDENING TECHNIQUES FOR MICROWAVE POWER
AMPLIFIER CHIP DESIGN
An amplifier is one of the most common electrical elements in any circuit
system. The requirements for amplification are as varied as the systems where
they are applied. Amplifiers are available in various forms ranging from minuscule
ICs to the largest high-power transmitter amplifiers. An RF power amplifier
(Grebennikov, 2005) is a sort of electronic amplifier
employed to convert a low-power radio-frequency signal into a larger signal
with significant power typically in order to drive the antenna of a transmitter.
It is usually optimized to have high efficiency, high output power compression,
good return loss on the input and output, good gain and optimal heat dissipation.
As a high-power device with large gain, it provides large output signal power
while requiring very small amount of RF power and it is commonly available from
any commercial signal generator. Therefore, the power amplifier is normally
known as the RF source or sometimes the Transmitter. Microwave power amplifiers
can be utilized in testing applications ranging from passive elements such as
antennas to active devices such as limiter diodes or MMIC based power amplifiers.
Furthermore, some other applications include testing requirements where a relatively
large amount of RF power is necessary for overcoming system losses to a radiating
element, or where there is a system requirement to radiate a Device-Under-Test
(DUT) with an intense electromagnetic field.
As a key link for wireless applications, solid-state active power amplifiers have been widely used in satellite communication, radar, electronic warfare, satellite navigation and weapon guidance systems. Due to the harsh requirements and the complexity of the transmission environment in wireless communication, the design of microwave power amplifiers almost becomes one of the most difficult functional units in the front of an RF transceiver system and its linearity, output power and efficiency greatly affect the signal quality, communication distance and communication time of the wireless communication system. Among these systems, because of the adverse external environment, aerospace applications not only impose higher more stringent requirements upon the performance of solid-state active power amplifier but also bring greater challenges for its design.
All the countries in the world (particularly the US and Japan) pay much attention
to solid-state active power amplification chip design and manufacturing technology
and they have developed many new products, such as the entire microwave system
integrated in a single chip with a diameter of few centimeters. This chip has
been substituted for the earlier microwave hardware chassis. This new microwave
system on chip has greatly improved the performance of microwave systems and
promotes the development of communication technology, radar technology and aerospace
technology. In 1996, the Department of Defense of USA announced three national
defense science and technology strategy files, i.e., Joint Warfighting Science
and Technology Plan (JWSTP), Defense Technology Area Plan (DTAP), Basic Research
Plan (BRP), which were modified in 1997. With regard to the microwave and millimeter
wave circuit design, they planned that in the late 1990s to the early 21st century
they would focus on the research and development of millimeter-wave/microwave
monolithic circuits, high- temperature and high-power circuits and multi-module
circuits, whose core is the heterojunction devices and circuits. One of the
prominent properties of microwave component is small volume and weight. The
development of Monolithic Microwave Integrated Circuits (MMICs) also speeds
up the millimeter wave instrument size and weight reduction, higher reliability.
This also brings more rigorous requirements for microwave component thermal
design. Recently, the authors of this study proposed a microwave component thermal
design method based on microstructure heat transfer (Yu
et al., 2010b).
Among various solid-state power amplifier chip production technologies, Heterojunction
Bipolar Transistor (HBT) becomes one of the most popular power chip technologies
due to its high efficiency, large gain, good linearity and high power density.
So far, HBT has been paid widespread attention and has been made great progress.
The HBTs made from a variety of materials continue to emerge and their performance
is unceasingly improved. Currently available and abuilding global networks based
on Low Earth Orbit (LEO) communication satellites have a rising demand for PAs
used in interstellar communication. To achieve high reliability and miniaturization
purposes, it is necessary to solidify the existing Traveling Wave Tubes (TWT)
and the HBT with high power density is suitable for high-power requirements.
At present, the demand of applying HBTs in L ~ C bands is growing and the HBT
has become the best candidate to substitute the former power traveling wave
tubes in L~ C bands due to its high power density, high efficiency and high
linearity characteristics. To improve the linearity, recently, an adaptive linearization
bias technique (Yu et al., 2010a) for microwave
solid-state active power amplifier design was proposed by the authors of this
study. Although, a single-chip radio-frequency integrated circuit cannot output
large power due to current restrictions on the production process and also cannot
provide comparable power level as TWT, it is a feasible solution to combine
them with the circuit or array synthesis technology. For aerospace applications
of solid-state active power amplifier chip, during the chip design, we should
meet the power targets as well as considering the specific characteristics of
aerospace applications in addition to performing complex trade-off among the
key performance such as efficiency, power and reliability.
Radiation is an important reason to cause anomalies or failures of spacecrafts electronic equipments and it is reported that about 40% of the faults come from space radiation. Therefore, radiation hardening technology is the key technology to keep aerospace electronic equipments operating with long life and high reliability and it is the research focus and hot in the field of astronautic electronics. According to the number of particles causing damage, radiation effects can be divided into: the effect caused by Single-Particle Events (SEE) and the cumulative effect such as Total Ionizing Dose (TID) and Displacement Damage (DD).
In the radiation hardening design, our solution is to obtain relatively high resistance to total ionizing dose effects based on derating design and fabrication process line selection. When a large number of cosmic particles enter into the chip substrate, a large amount of ionization will be produced, but the chip performance will remain stable and no deterioration will occur. Secondly, relatively high resistance to burnout caused by the particle radiation can be obtained, i.e., the chip will not be burned out even if the parasitic effect has been inspired by high energy particles or a large disturbance of electrical properties has occurred. Thirdly, the resistance to displacement damage to a certain degree can be obtained, i.e., even if a considerable amount of particle radiation has changed the circuit performance, the chip will still be able to work stably and reliably in the space environment with minor injuries. In allusion to various damage mechanisms, we mainly take following two measures:
When traditional MOS or bipolar elements are injected with high-energy particles, they will interact with the oxide (SiO2) and will be ionized, resulting in a large number of electron-hole pairs, which have two motion trends, i.e., recombination and drift. If there is no external electric field, the recombination trend is stronger. If there is electric field, the electron and hole will move to opposite directions along the electric field, the electron can quickly leave the oxide due to its transfer rate is very high, resulting in the accumulation of holes within the oxide, forming a gate oxide hole capture. The greater the electric field strength is, the higher electron mobility is and the stronger the gate oxide capture is. That is why the TID damage with electronic components is severer than that without electrical components. Gate oxide and interface capture generate the parasitic electric field in the device work area, resulting in the drift of threshold voltage Vth and the propagation delay Tpd, the increase of static current Icc, as well as the reduction of the magnification coefficient of the transistor. When the damage exceeds a certain threshold, the elements will fail to work.
With respect to material selection, we notice that GaAs is a kind of III-V semiconductors with high speed, high frequency, high temperature resistance, low noise and light and so on. Compared with Si, GaAs has many advantages in the physical nature, particularly the high electron mobility (8500 cm/V≥s) and large band gap (1.424 eV), which make GaAs devices not only be able to work at a high temperature but also possess high radiation hardness. Many studies have verified that the GaAs material has good particle detection performance and good resistance to γ-ray induced damage.
In the design of microelectronic devices within the circuit, we adopt high radiation-tolerant devices to improve the radiation resistance of outside circuits. HBT occupies a unique and important position in high-speed, large dynamic range, low harmonic distortion and low phase noise circuits. It uses wide-band emitter and allows high-doped base region, which can achieve high cutoff frequency, high gain, high efficiency and high linearity and high breakdown voltage (10-15 V), while its excellent radiation-tolerant performance is very suitable for space power amplifier applications. Compared with the radiation tolerance of the field-effect transistors MESFET and HEMT, HBT has the following advantages: (1) It has high breakdown voltage. The HBT collector has wide band-gap materials, we can get high breakdown voltage through the design of collector thickness and doping concentration, resulting in large output power. And, the reverse breakdown voltage of collector junction (BC junction) that determines the breakdown voltage of HBT depends on the epitaxial material and it is less susceptible to the process. (2) it has low leakage current. (3) the turn-on voltage of HBT is determined by the intrinsic energy gap of the epitaxial material and has nothing to do with the process and it has good reproducibility.
||Improved tube core structure
(4) it can avoid the back gate effect. In a word, it is because of the high
breakdown voltage of GaAs HBT, its low leakage current and its vertical structure
that it can avoid the back gate effect and it has high radiation hardness to
With regard to the process, we cooperate with foreign fabrication manufacturers to modify the process according to special requirements of radiation hardening. We alleviate the radiation effects by reducing parasitic device parameters as much as possible, e.g., adjusting the doping concentration of GaAs to release TID effects. We also adjust the structural layout to reduce parasitic devices to ease radiation effects and thus slow the TID damage.
In addition, we modify the original tube core model by adding a thermal model
as shown in Fig. 1. We perform the accurate simulation to
investigate the performance change due to the radiation under the actual conditions,
so as to optimize the circuit later and improve the manufacturability and reliability
of the chip.
Radiation-tolerant circuit design:
Radiation-tolerant circuit design covers the entire process of circuit design, analysis and simulation, including failure link analysis, tolerance design, derating design methods that aim to find and predict the weak links in damage, adopt hardening or tolerance design, so as to improve the capacity of the circuit against radiation.
Failure link analysis:
By analyzing the changes in damage parameters and the impact on other parts, we find the key link or weak link, which is the foundation of derating design and tolerance design.
By deliberately reducing the heat and electronic stresses imposed on electrical components, we reduce the radiation failure rate of components. We reasonably design the bias circuit and choose the work point, so as to let the circuit work with a relatively reliable voltage and thus effectively reduce the possibility of burnout due to cosmic particles. We adopt multi-tube parallel structure and preserve appropriate power redundancy to cope with the circuit disturbance caused by cosmic particles.
With regard to the cumulated damage such as TID and displacement damage, their radiation damage behaves the drift of circuit parameters such as the transistor amplification rate and reverse breakdown voltage. If the circuit parameters are set incorrectly, once the parameters drift because of radiation, the system will not work. Thus, based on the parameter drift rule, we perform the tolerance design for device parameters, so as to make the device reliably work with minor injuries. We can add the appropriate feedback and adopt a stable structure and select the appropriate operating point to make the DC operating point be in the secure area, so as to get greater tolerance capability.
Through the radiation hardening design, the result shows that the resistance of our chip to neutron radiation is larger than 1015 n cm-2, while the resistance to ionizing radiation is larger than 107rad and the resistance to transient radiation is larger than 109rad sec-1.
In this study, we overview physical and logical radiation hardening techniques
and propose some effective solutions in our power amplification chip design
to resist the radiation. Typical physical radiation-hardening techniques are
using insulating substrates, utilizing bipolar integrated circuits, adopting
radiation-tolerant SRAM. Typical logical radiation-hardening techniques are
using error correcting memory, utilizing redundant elements, adopting a watchdog
timer. In the radiation hardening design for power amplification chips, our
solution is to obtain relatively high resistance to total ionizing dose effects
based on derating design and fabrication process line selection. We adopt the
GaAs HBT process because of the high breakdown voltage of GaAs HBT, its low
leakage current and its vertical structure and it can avoid the back gate effect
and it has high radiation hardness to space environment. We alleviate the radiation
effects by reducing parasitic device parameters as much as possible and we adjust
the structural layout to reduce parasitic devices to ease radiation effects
and thus slow the TID damage. In radiation-tolerant circuit design, we perform
failure link analysis, tolerance design, derating design methods to improve
the capacity of the circuit against radiation.
This study is supported by the key project granted by the Innovative Foundation of Astronautic Science and Technology of Year 2009 in China and the key project granted by CAST Innovative Foundation of Year 2009 in China.
1: Adell, P.C., R.D. Schrimpf, C.R. Cirba, W.T. Holman, X. Zhu, H.J. Barnaby and O. Mion, 2005. Single event transient effects in a voltage reference. Microelectronics Reliability, 45: 355-359.
2: Albadri, A.M., R.D. Schrimpf, K.F. Galloway and D.G. Walker, 2006. Single event burnout in power diodes: Mechanisms and models. Microelectronics Reliability, 46: 317-325.
3: Arimura, I., 1982. Neutron effects in modern semiconductor devices. J. Nulc. Mater., 108-109: 635-642.
4: Badhwara, G.D., J.U. Patela, F.A. Cucinotta and J.W. Wilsonb, 1995. Measurements of the secondary particle energy spectra in the space shuttle. Radiation Measurements, 24: 129-138.
5: Badila, M., P. Godignon, J. Millan, S. Berberich and G. Brezeanu, 2001. The electron irradiation effects on silicon gate dioxide used for power MOS devices. Microelectronics Reliability, 41: 1015-1018.
6: Barbara, N.V. R.D. Schrimpf and W.J. Kerwin, 1990. Ionizing-radiation-induced degradation in electronic power amplifiers. Proceedings of the Conference Record of the 1990 IEEE Industry Applications Society Annual Meeting, Oct. 7-12, Seattle, WA., pp: 1667-1672
Direct Link |
7: Bauer, F.D., 2004. The super junction bipolar transistor: A new silicon power device concept for ultra low loss switching applications at medium to high voltages. Solid State Electron., 48: 705-714.
8: Baumann, R., T. Hossain, S. Murata and H. Kitagawa, 1995. Boron compounds as a dominant source of alpha particles in semiconductor devices. Proceedings of the 33rd Annual IEEE International Reliability Physics Symposium, April 4-6, Las Vegas, pp: 297-302
Direct Link |
9: Cherng, M., I. Jun and T. Jordan, 2007. Optimum shielding in Jovian radiation environment. Nulc. Instruments Method Phys. Res. Sect. A Accelerators Spectrometers Detectors Assoc. Equip., 580: 633-636.
10: Dodd, P.E., 2005. Physics-based simulation of single-event effects. IEEE Trans. Device Mater. Reliability, 5: 343-357.
Direct Link |
11: El-Attar, A.M. and G. Fahmy, 2007. An improved watchdog timer to enhance imaging system reliability in the presence of soft errors. Proceedings of the 2007 IEEE International Symposium on Signal Processing and Information Technology, Dec. 15-18, Giza, pp: 1100-1104
Direct Link |
12: Gaisser, T.K., 1990. Cosmic Rays and Particle Physics. Cambridge University Press, Cambridge
13: Gonzalez, A.F. and P. Mazumdar, 2000. Redundant arithmetic, algorithms and implementations. Integr. VLSI J., 30: 13-53.
CrossRef | Direct Link |
14: Goseva-Popstojanova, K. and A. Grnarov, 1993. N version programming with majority voting decision: Dependability modeling and evaluation. Microprocess. Microprogramming, 38: 811-818.
Direct Link |
15: Grebennikov, A., 2005. RF and Microwave Power Amplifier Design. The McGraw-Hill Co. Inc., USA
16: Higgins, D.F., K. Lee and L. Marin, 1978. System-generated EMP. IEEE Trans. Electromagnetic Compatibility, 20: 14-22.
Direct Link |
17: He, C., M. Jacome and G. De-Veciana, 2005. Scalable defect mapping and con guration of memory-based nanofabrics. Proceedings of the IEEE International High- Level Design, Validation and Test Workshop, Nov. 30-Dec. 2, Austin, TX, USA., pp: 11-18
Direct Link |
18: He, W., Z.X. Zhang, E.X. Zhang, W.J. Yu, H. Tian and X. Wang, 2008. Practical considerations in the design of SRAM cells on SOI. Microelectronics J., 39: 1829-1833.
19: Holmes-Siedle, A.G. and L. Adams, 2002. Handbook of Radiation Effects. Oxford University Press, England
20: Huang, Z., Y. Fa-Xin, Z. Shuting, L. Hao, W. Pinghui and Z. Yao, 2009. Empirical-statistics analysis for zero-failure GaAs MMICs life testing data. IEICE Trans. Fundam., 92: 2376-2379.
Direct Link |
21: Huang, Z.L., F.X. Yu, S.T. Zhang, Y. Zheng and J.X. Liu, 2009. A novel approach for mmic reliability testing based on weibull distribution. Infom. Technol. J., 8: 1080-1083.
CrossRef | Direct Link |
22: Jeffery, C.M., A. Basagalar and R.J.O. Figueiredo, 2004. Dynamic sparing and error correction techniques for fault tolerance in nanoscale memory structures. Proceedings of the 4th IEEE Conference on Nanotechnology, Aug. 16-19, USA., pp: 168-170
Direct Link |
23: Johansson, A., 1977. Principles and techniques of radiation hardening: Norman J. rudie western periodicals company, North Hollywood, USA. Nulc. Instruments Methods, 140: 408-408.
24: Kern, W. and R.K. Smeltzer, 1986. Borophosphosilicate glasses for integrated circuits. Microelectonics Reliability, 26: 792-792.
25: Kuznetsov, N.V. and R.A. Nymmik, 1996. Single event upsets of spacecraft microelectronics exposed to solar cosmic rays. Radiation Measurements, 26: 959-965.
26: Lantos, P. and N. Fuller, 2003. History of the solar particle event radiation doses on-board aeroplanes using a semi-empirical model and Concorde measurements. Radiatation Prot. Dosimetry, 104: 199-210.
27: Lisnianski, A., G. Levitin and H. Ben-Haim, 2000. Structure optimization of multi-state system with time redundancy. Reliability Eng. Syst. Safety, 67: 103-112.
28: Meulenberg, A., H.L.A. Hung, K.E. Peterson and T.W. Anderson, 1988. Total dose and transient radiation effects on GaAs MMICs. IEEE Trans. Electron. Devices, 35: 2125-2132.
Direct Link |
29: Miller, J., L. Taylor, C. Zeitlin, L. Heilbronn and S. Guetersloh et al., 2009. Lunar soil as shielding against space radiation. Radiation Measurements, 44: 163-167.
30: Myers, A.F. and A. Rauzy, 2008. Assessment of redundant systems with imperfect coverage by means of binary decision diagrams. Reliability Eng. Syst. Safety, 93: 1025-1035.
31: Naber, J., 1995. Digital GaAs Integrated Circuits. Gallium Arsenide IC Applications Handbook. Academic Press, New York, pp: 57-78
32: Nepal, K., R.I. Bahar, J. Mundy, W.R. Patterson and A. Zaslavsky, 2006. MRF reinforcer: A probabilistic element for space redundancy in nanoscale circuits. IEEE Micro, 26: 19-27.
Direct Link |
33: Ou, E. and W. Yang, 2004. Fast error-correcting circuits for fault-tolerant memory. Proceedinds of the 2004 International Workshop on Memory Technology, Design and Testing, August 9-10, USA., pp: 8-12
Direct Link |
34: Pease, R.L., 2003. Total ionizing dose effects in bipolar devices and circuits. IEEE Trans. Nulc. Sci., 50: 539-551.
35: Roig, J., D. Flores, S. Hidalgo, J. Rebollo and J. Millan, 2004. Thin-film silicon-on-sapphire LDMOS structures for RF power amplifier applications. Microelectronics J., 35: 291-297.
36: Schrimpf, R.D., R.A. Weller, M.H. Mendenhall, R.A. Reed and L.W. Massengill, 2007. Physical mechanisms of single-event effects in advanced microelectronics. Nulc. Instruments Methods Phys. Res. Sect. B Beam Interactions Mater. Atoms, 261: 1133-1136.
Direct Link |
37: Simoen, E., A. Mercha, C. Claeys and N. Lukyanchikova, 2007. Low-frequency noise in silicon-on-insulator devices and technologies. Solid State Electronics, 51: 16-37.
38: Snyder, C.W., 1959. The upper boundary of the van allen radiation belts. Nature, 184: 439-440.
Direct Link |
39: Srihari, S.N., 1982. Reliability analysis of biased majority-vote systems. IEEE Trans. Reliability, R-31: 117-118.
Direct Link |
40: Sun, F. and T. Zhang, 2006. Two fault tolerance design approaches for hybrid cmos/nanodevice digital memories. IEEE International workshop on defect and fault tolerant nanoscale architectures (Nanoarch), 2006.
41: Szmidt, J., 1999. Electronic properties of nanocrystalline layers of wide-band-gap materials. Chaos Solitons Fractals, 10: 2099-2152.
Direct Link |
42: Tahoori, M.B., 2005. A mapping algorithm for defect-tolerance of reconfigurable nano-architectures. Proceedings of the 2005 IEEE/ACM International Conference on Computer-Aided Design, Nov. 6-10, USA., pp: 668-672
Direct Link |
43: Torvik, J.T., J.I. Pankove and B.V. Zeghbroeck, 2000. GaN/SiC heterojunction bipolar transistors. Solid State Electronics, 44: 1229-1233.
Direct Link |
44: Van Lint, V.A.J., T.M. Flanagan, R.E. Leadon, J.A. Naber and V.C. Rogers, 1980. Mechanisms of Radiation Effects in Electronic Materials. Wiley, New York
45: Vollrath, J., U. Lederer and T. Hladschik, 2001. Compressed bit fail maps for memory fail pattern classification. J. Electronic Testing, 17: 291-297.
Direct Link |
46: Yu, F.X., H. Chen, Z.L. Huang, H. Luo and Z.M. Lu, 2010. Microwave component thermal design based on microstructure heat transfer. Int. J. Comput. Sci. Eng. Syst., 4: 1-4.
47: Yu, F.X., H. Luo, Z.L. Huang and Z.M. Lu, 2010. Adaptive linearization bias technique for microwave solid-state active power amplifier design. Int. J. Comput. Sci. Eng. Syst., 4: 5-8.
48: Walsh, D.S., P.E. Dodd, M.R. Shaneyfelt and J.R. Schwank, 2001. Investigation of body-tie effects on ion beam induced charge collection in silicon-on-insulator FETs using the Sandia nuclear microprobe. Nulc. Instruments Methods Phys. Res. Sect. B Beam Interactions Mater. Atoms, 181: 305-310.
49: Wang, G., W. Gong and R. Kastner, 2006. Defect tolerant nanocomputing using bloom filters. Proceedings of the 14th Annual IEEE Symposium on Field-Programmable Custom Computing Machines, April 24-26, Napa, CA., pp: 277-278
Direct Link |
50: Ziegler, M. and M. Stan, 2003. CMOS/nano co-design for crossbar-based molecular electronic systems. IEEE Trans. Nanotechnol., 2: 217-230.
Direct Link |