• ETRI Journal
• /
• 제42권4호
• /
• pp.549-561
• /
• 2020

We developed a 0.1-㎛ metamorphic high electron mobility transistor and fabricated a W-band monolithic microwave integrated circuit chipset with our in-house technology to verify the performance and usability of the developed technology. The DC characteristics were a drain current density of 747 mA/mm and a maximum transconductance of 1.354 S/mm; the RF characteristics were a cutoff frequency of 210 GHz and a maximum oscillation frequency of 252 GHz. A frequency multiplier was developed to increase the frequency of the input signal. The fabricated multiplier showed high output values (more than 0 dBm) in the 94 GHz-108 GHz band and achieved excellent spurious suppression. A low-noise amplifier (LNA) with a four-stage single-ended architecture using a common-source stage was also developed. This LNA achieved a gain of 20 dB in a band between 83 GHz and 110 GHz and a noise figure lower than 3.8 dB with a frequency of 94 GHz. A W-band image-rejection mixer (IRM) with an external off-chip coupler was also designed. The IRM provided a conversion gain of 13 dB-17 dB for RF frequencies of 80 GHz-110 GHz and image-rejection ratios of 17 dB-19 dB for RF frequencies of 93 GHz-100 GHz.

• ETRI Journal
• /
• 제42권4호
• /
• pp.491-504
• /
• 2020

We present AB9, a neural processor for inference acceleration. AB9 consists of a systolic tensor core (STC) neural network accelerator designed to accelerate artificial intelligence applications by exploiting the data reuse and parallelism characteristics inherent in neural networks while providing fast access to large on-chip memory. Complementing the hardware is an intuitive and user-friendly development environment that includes a simulator and an implementation flow that provides a high degree of programmability with a short development time. Along with a 40-TFLOP STC that includes 32k arithmetic units and over 36 MB of on-chip SRAM, our baseline implementation of AB9 consists of a 1-GHz quad-core setup with other various industry-standard peripheral intellectual properties. The acceleration performance and power efficiency were evaluated using YOLOv2, and the results show that AB9 has superior performance and power efficiency to that of a general-purpose graphics processing unit implementation. AB9 has been taped out in the TSMC 28-nm process with a chip size of 17 × 23 ㎟. Delivery is expected later this year.

• ETRI Journal
• /
• 제42권4호
• /
• pp.575-584
• /
• 2020

Electromagnetic (EM) waves used to send signals under seawater are normally restricted to low frequencies (f) because of sudden exponential increases of attenuation (𝛼) at higher f. The mathematics of EM wave propagation in seawater demonstrate dependence on relative permeability (𝜇_r), relative permittivity (𝜀_r), conductivity (𝜎), and f of transmission. Estimation of 𝜀_r and 𝜎 based on the W. Ellison interpolation model was performed for averaged real-time data of temperature (T) and salinity (S) from 1955 to 2012 for all oceans with 41 088 latitude/longitude points and 101 depth points up to 5500 m. Estimation of parameters such as real and imaginary parts of 𝜀_r, 𝜀_r', 𝜀_r", 𝜎, loss tangent (tan 𝛿), propagation velocity (V_p), phase constant (𝛽), and α contributes to absorption loss (L_a) for seawater channels carried out by using normal distribution fit in the 3 GHz-40 GHz f range. We also estimated total path loss (L_PL) in seawater for given transmission power P_t and antenna (dipole) gain. MATLAB is the simulation tool used for analysis.

• ETRI Journal
• /
• 제36권2호
• /
• pp.317-320
• /
• 2014

A balanced RF duplexer with low interference in an extremely narrow bandgap is proposed. The Long-Term Evolution band-7 duplexer should be designed to prevent the co-existence problem with the WiFi band, whose fractional bandgap corresponds to only 0.7%. By implementing a hybrid bulk acoustic wave (BAW) structure, the temperature coefficient of frequency (TCF) value of the duplexer is successfully reduced and the suppressed interference for the narrow bandgap is performed. To achieve an RF duplexer with balanced Rx output topology, we also propose a novel balanced BAW Rx topology and RF circuit block. The novel balanced Rx filter is designed with both lattice- and ladder-type configurations to ensure excellent attenuation. The RF circuit block, which is located between the antenna and the Rx filter, is developed to simultaneously function as a balance-tounbalance transformer and a phase shift network. The size of the fabricated duplexer is as small as $2.0mm{\times}1.6mm$. The maximum insertion loss of the duplexer is as low as 2.4 dB in the Tx band, and the minimum attenuation in the WiFi band is as high as 36.8 dB. The TCF value is considerably lowered to $-16.9ppm/^{\circ}C$.

• ETRI Journal
• /
• 제36권3호
• /
• pp.352-360
• /
• 2014

This paper presents a 900 MHz zero-IF RF transceiver for IEEE 802.15.4g Smart Utility Networks OFDM systems. The proposed RF transceiver comprises an RF front end, a Tx baseband analog circuit, an Rx baseband analog circuit, and a ${\Delta}{\Sigma}$ fractional-N frequency synthesizer. In the RF front end, re-use of a matching network reduces the chip size of the RF transceiver. Since a T/Rx switch is implemented only at the input of the low noise amplifier, the driver amplifier can deliver its output power to an antenna without any signal loss; thus, leading to a low dc power consumption. The proposed current-driven passive mixer in Rx and voltage-mode passive mixer in Tx can mitigate the IQ crosstalk problem, while maintaining 50% duty-cycle in local oscillator clocks. The overall Rx-baseband circuits can provide a voltage gain of 70 dB with a 1 dB gain control step. The proposed RF transceiver is implemented in a $0.18{\mu}$ CMOS technology and consumes 37 mA in Tx mode and 38 mA in Rx mode from a 1.8 V supply voltage. The fabricated chip shows a Tx average power of -2 dBm, a sensitivity level of -103 dBm at 100 Kbps with PER < 1%, an Rx input $P_{1dB}$ of -11 dBm, and an Rx input IP3 of -2.3 dBm.

• ETRI Journal
• /
• 제30권3호
• /
• pp.383-393
• /
• 2008

The etching mechanism of $ZrO_2$ thin films and etch selectivity over some materials in both $BCl_3$/Ar and $BCl_3/CHF_3$/Ar plasmas are investigated using a combination of experimental and modeling methods. To obtain the data on plasma composition and fluxes of active species, global (0-dimensional) plasma models are developed with Langmuir probe diagnostics data. In $BCl_3$/Ar plasma, changes in gas mixing ratio result in non-linear changes of both densities and fluxes for Cl, $BCl_2$, and ${BCl_2}^+$. In this work, it is shown that the non-monotonic behavior of the $ZrO_2$ etch rate as a function of the $BCl_3$/Ar mixing ratio could be related to the ion-assisted etch mechanism and the ion-flux-limited etch regime. The addition of up to 33% $CHF_3$ to the $BCl_3$-rich $BCl_3$Ar plasma does not influence the $ZrO_2$ etch rate, but it non-monotonically changes the etch rates of both Si and $SiO_2$. The last effect can probably be associated with the corresponding behavior of the F atom density.

• ETRI Journal
• /
• 제25권5호
• /
• pp.337-344
• /
• 2003

In this paper, we introduce a fully synthesizable 32-bit embedded microprocessor core called the AE32000B. The AE32000B core is based on the extendable instruction set computer architecture, so it has high code density and a low memory access rate. In order to improve the performance of the core, we developed and adopted various design options, including the load extension register instruction (LERI) folding unit, a high performance multiply and accumulate (MAC) unit, various DSP units, and an efficient coprocessor interface. The instructions per cycle count of the Dhrystone 2.1 benchmark for the designed core is about 0.86. We verified the synthesizability and the area and time performances of our design using two CMOS standard cell libraries: a 0.35-${\mu}m$ library and a 0.18-${\mu}m$ library. With the 0.35-${\mu}m$ library, the core can be synthesized with about 47,000 gates and operate at 70 MHz or higher, while it can be synthesized with about 53,000 gates and operate at 120 MHz or higher with the 0.18-${\mu}m$ library.

• ETRI Journal
• /
• 제26권4호
• /
• pp.315-320
• /
• 2004

This paper proposes a 10-${\mu}m$ thick oxide layer structure that can be used as a substrate for RF circuits. The structure has been fabricated using an anodic reaction and complex oxidation, which is a combined process of low-temperature thermal oxidation (500 $^{\circ}C$ for 1 hr at $H_2O/O_2$) and a rapid thermal oxidation (RTO) process (1050 ${\circ}C$, for 1 min). The electrical characteristics of the oxidized porous silicon layer (OPSL) were almost the same as those of standard thermal silicon dioxide. The leakage current density through the OPSL of 10 ${\mu}m$ was about 10 to 50 $nA/cm^2$ in the range of 0 to 50 V. The average value of the breakdown field was about 3.9 MV/cm. From the X-ray photo-electron spectroscopy (XPS) analysis, surface and internal oxide films of OPSL prepared by a complex process were confirmed to be completely oxidized. The role of the RTO process was also important for the densification of the porous silicon layer (PSL) oxidized at a lower temperature. The measured working frequency of the coplanar waveguide (CPW) type short stub on an OPSL prepared by the complex oxidation process was 27.5 GHz, and the return loss was 4.2 dB, similar to that of the CPW-type short stub on an OPSL prepared at a temperature of 1050 $^{\circ}C$ (1 hr at $H_2O/O_2$). Also, the measured working frequency of the CPW-type open stub on an OPSL prepared by the complex oxidation process was 30.5 GHz, and the return was 15 dB at midband, similar to that of the CPW-type open stub on an OPSL prepared at a temperature of $1050^{\circ}C$ (1 hr at $H_2O/O_2$).

• ETRI Journal
• /
• 제21권4호
• /
• pp.52-64
• /
• 1999

QoS monitoring is a kind of real-time systems which allows each level of the system to track the ongoing QoS levels achieved by the lower network layers. For these systems, real-time communications between corresponding transport protocol objects is essential for their correct behavior. When two or more entities are employed to perform a certain task as in the case of communication protocols, the capability to do so is called interoperability and considered as the essential aspect of correctness of communication systems. This paper describes a formal approach on modeling and interoperability test case generation of a real-time QoS monitoring protocol. For this, we specify the behavior of flow monitoring of transport layer QoS protocol, i.e., METS protocol, which is proposed to address QoS from an end-to-end's point of view, based on QoS architecture model which includes ATM net work in lower layers. We use a real-time Input/Output finite State Machine to model the behavior of real-time flow monitoring over time. From the modeled real-time I/OFSM, we generate interoperability test cases to check the correctness of METS protocol's flow monitoring behaviors for two end systems. A new approach to efficient interoperability testing is described and the method of interoperability test cases generation is shown with the example of METS protocol's flow monitoring. The current TTCN is not appropriate for testing real-time and multimedia systems. Because test events in TTCN are for message-based system and not for stream-based systems, the real-time in TTCN can only be approximated. This paper also proposes the notation of real-time Abstract Test Suite by means of real-time extension of TTCN. This approach gives the advantages that only a few syntactical changes are necessary, and TTCN and real-time TTCN are compatible. This formal approach on interoperability testing can be applied to the real-time protocols related to IMT-2000, B-ISDN and real-time systems.

• ETRI Journal
• /
• 제24권5호
• /
• pp.349-359
• /
• 2002

With Ni/Au and Pd/Au metal schemes and low temperature processing, we formed low resistance stable Ohmic contacts to p-type GaN. Our investigation was preceded by conventional cleaning, followed by treatment in boiling $HNO_3$:HCl (1:3). Metallization was by thermally evaporating 30 nm Ni/15 nm Au or 25 nm Pd/15 nm Au. After heat treatment in $O_2$ + $N_2$ at various temperatures, the contacts were subsequently cooled in liquid nitrogen. Cryogenic cooling following heat treatment at $600^{\circ}C$ decreased the specific contact resistance from $9.84{\times}10^{-4}$ ${\Omega}cm^2$ to $2.65{\times}10^{-4}$ ${\Omega}cm^2$ for the Ni/Au contacts, while this increased it from $1.80{\times}10^{-4}$ ${\Omega}cm^2$ to $3.34{\times}10^{-4}$ ${\Omega}cm^2$ for the Pd/Au contacts. The Ni/Au contacts showed slightly higher specific contact resistance than the Pd/Au contacts, although they were more stable than the Pd contacts. X-ray photoelectron spectroscopy depth profiling showed the Ni contacts to be NiO followed by Au at the interface for the Ni/Au contacts, whereas the Pd/Au contacts exhibited a Pd:Au solid solution. The contacts quenched in liquid nitrogen following sintering were much more uniform under atomic force microscopy examination and gave a 3 times lower contact resistance with the Ni/Au design. Current-voltage-temperature analysis revealed that conduction was predominantly by thermionic field emission.