by Victoria Pereira, CEO, Otava Inc
MPD: From a semiconductor perspective, what remains to be done to allow 5G to fulfill its promise?
VP: 5G devices demand smaller, more integrated chipsets to handle the network’s increased data processing and communication demands. Semiconductor manufacturers must refine their processes to produce RF components with improved bandwidth, power efficiency, and linearity. Compact, multi-beam antennas capable of handling substantial amounts of data, beamforming techniques and phased array antennas will enable beam steering and enhance signal coverage in millimeter-wave deployments. Finally, the cost of the technologies, especially at higher frequencies, must come down as 6G begins to be deployed in 3 to 4 years in high volumes.
Satellite-based communication will be a requisite factor in 6G as it provides global coverage, including in remote and underserved areas where it may be uneconomical to deploy terrestrial infrastructure. It can help offload traffic in congested urban areas or during major events, expand IoT networks, provide disaster response, and potentially serve as 5G backhaul. I predict that the 5G FR3 band between 12 and 30 GHz will be the most exciting band. In short, 5G non-terrestrial networks will be an essential piece of the network infrastructure.
Other challenges include heterogeneous integration, in which different semiconductor materials and device types are integrated into a single “chip”. Although it is used in some products today such as processing and memory products, true 3D stacking technology in RF chips is still developing. 3D stacking offers significant performance, integration, and cost reduction advantages for frequencies up to 200 GHz. The 2.5D interposer technology is another innovative technique that allows multiple chips to be integrated into a single package while operating over wide bandwidths with lower insertion losses between chips. These and other emerging innovations will be essential for 5G radios and systems to meet future size, weight, and power requirements.
MPD: Tell us about some of the work Otava is doing.
VP: We have focused on three major product areas that cover frequencies up to 40 GHz: beamformer ICs, digitally tunable band-pass filters, and RF switches. The SiGe BiCMOS beamformer IC remains the widest beamforming chip in the market in terms of frequency range and has eight half-duplex transmit channels and eight receive channels. Each channel has 20 dB of gain control and 360 deg. of phase control. The IC includes temperature sensing, RMS power detection, and drivers to control amplifiers and switches.
The digitally tunable band-pass filter is fabricated with silicon on insulator (SOI) technology and is a single-chip replacement for discrete filter banks. It achieves 32 states of digital control, 20 dB of rejection, and has an instantaneous bandwidth of 1.5 GHz. The filter is controlled with a three-wire serial interface and measures 2.3 x 1.6 mm. Three models are available with coverage between 2.5 and 7.5 GHz, 14 and 24 GHz, and 24 to 40 GHz. Our switches are fabricated using SOI technology and cover DC to 40 GHz with insertion loss of less than 1.5 dB, isolation of at least 35 dB, and a switching speed of 175 ns.
In addition, we have developed a 64-element phased array antenna module that uses 16 of our beamformer ICs and covers 24 to 30 GHz. It has an EIRP of +45 dBm and scans +/-45 deg. in azimuth and +/-15 dB in elevation for each antenna element. It combines with an AMD (Xilinx) Zynq UltraScale+ RFSoC Gen 3 and is suited for 5G infrastructure and backhaul and K- and Ka-band radar and satellite communications systems.
Finally, our millimeter-wave radio card provides upconversion and downconversion of two transmit and two receive signals between 20 and 30 GHz to IF signals sampled directly by integrated RF-DACs and RF-ADCs within the AMD RFSoC. Designers can add their own front end or use an 8×8 dual-polarized antenna array jointly developed by Otava and Avnet for hybrid beamforming applications and massive MIMO radios.
All these products are supported by evaluation boards and MathWorks Simulink models tailored for antenna arrays that capture the circuit performance over operating frequency, VGA gain and phase shifter setting, and input and output power level.
MPD: Why have you chosen silicon-on-insulator (SOI) technology for the company’s tunable filters and switches?
VP: SOI has inherent advantages over bulk silicon for RF applications, so it has become the technology of choice for high-speed circuits and microwave applications. Many of SOI’s benefits are achieved because it has a buried oxide layer, which isolates the device layer from the substrate, which reduces parasitic capacitance. The high-resistivity substrate creates high-performance passives with lower insertion loss. This reduces power dissipation because less gain is required to compensate for losses.
SOI substrate inherently reduces interference and crosstalk and has greater radiation hardness tolerance, which makes it well-suited for space applications. SOI technology is very appealing for products needed in large numbers and several processes are available around the world. In fact, the SOI processes are more appealing outside the US. It is important that we have the best silicon fabs geared for RF front ends in the US.
MPD: The media sometimes characterizes the Department of Defense as needing to catch up technologically to keep pace with increasing threats. Do you agree with this statement?
VP: I will not answer this question directly because of the sensitivity of these systems. However, the Chips Act enacted last year will create multiple paths for collaboration in microelectronics between private companies, large and small, and universities.
New, innovative defense systems will result from collaboration between the government and the private sector, the Army, Air Force, and Navy research laboratories, and many others, which are driving the development of emerging technologies. It remains to be seen how small companies can truly benefit. Small companies can react quicker to technological innovation and prototype builds and have a product focus, but typically small companies such as Otava Inc. do not have the ample resources required to participate in the myriad of Chips Act-related proposals.
Additionally, the process of proposal submission and selection remains slow for small semiconductor companies. This is unfortunate because the start-ups for defense technologies will be the fastest way toward innovative systems. Nevertheless, the combined efforts of academic institutions that are major winners in the Microelectronics Commons hub proposals will provide an exciting path for the U.S. to dominate critical defense technologies with potential dual use applications.
About the Author
Victoria Pereira has more than 30 years of experience leading engineering teams in microwave semiconductor technology and applications and co-founded Otava, Moorestown, NJ, which manufactures RF front-end IC solutions for 5G, satcom, and defense systems. Victoria’s experience spans Hughes Aircraft Radar Systems, IBM Microelectronics, Instye Corp (ITT/L3 Harris), and Lockheed Martin. She has five patents, was awarded the Technology All Stars Award from Women of Color, and an Individual Technology Innovation award from Lockheed Martin. She received her BSEE from the University of Southern California.