Electrical Engineering

Electrical engineering at Caltech emphasizes both devices and systems. Closely allied with computation and neural systems, applied physics, bioengineering, computer science, and control and dynamical systems, it offers students the opportunity for study and research, both theoretical and experimental, in a wide variety of subjects, including wireless systems, photonics, quantum information science, modern optics, biophotonics, MEMS/NEMS, solid-state materials and devices, power electronics, energy systems, control theory, data science, nanoscale systems, signal processing, learning, data compression, and communications.

Substantial experimental laboratory facilities, housed mainly in the Moore Laboratory of Engineering, are associated with each of the research fields described on the following pages.

• Artificial Intelligence, Machine Learning, and Neural Networks (Yaser Abu-Mostafa, Katie Bouman, Jehoshua [Shuki] Bruck, Michelle Effros, Georgia Gkioxari, Pietro Perona, Yang Song) Theoretical and applied research in machine learning. Current theoretical research directions include estimating the information (data and hints) needed to learn a given task, characterizing the computational complexity of computing with neural networks, and modeling how biological networks of neurons create brain function. Current applied research directions include applications applications of AI and Machine Learning in medicine and in computational finance. Examples of active projects include real-time, non-invasive AI diagnostics of the heart and AI-enabled ultrasound detection of mini-clots in the bloodstream (Abu-Mostafa), a new approach for incorporating prior knowledge (hints) in learning from data (Abu-Mostafa), a learning-based approach to jointly optimize sensor and algorithm designs in computational camera pipelines in order to automatically discover new imaging strategies (Bouman), and a Our latest project is a data-driven approach to predicting the spread of COVID-19 in every U.S. county.theoretical investigation of how networks of neurons form, store, and retrieve memories (Effros). Learning tools are also used for more efficient uncertainty quantification in ill-posed inverse problems, to improve our underlying model of a physical system, and to incorporate more sophisticated knowledge into inference methods.
• Bio-Electronics, Bio-Optics, and Medical Imaging (Azita Emami, Ali Hajimiri, Alireza Marandi, Axel Scherer, Yu-Chong Tai, Lihong Wang, Changhuei Yang)
  Electrical engineering has broadly impacted biomedicine. Integrated circuits are used to detect biological matters, such as DNA and proteins, in various (e.g., electrical, magnetic, optical) sensing modes with high sensitivity. Applications include microarrays and point-of-care. MEMS and nanotechnologies are developed for a new generation of micro-implants. Examples include retinal implants, drug delivery pumps, and bio-analyte sensors. Optical technologies are developed for medical imaging. Current areas of interest include optofluidics, wavefront shaping, wide field-of-view imaging, chip-scale microscopy, Fourier ptychographic microscopy, photoacoustic tomography, microwave-induced thermoacoustic tomography, and light-speed compressed ultrafast tomography.
• Biological Information Science (Michelle Effros)
  Electrical engineering provides tools useful for studying information transmission, storage, and understanding.  While the application of these tools is wide spread and well known in technological systems, their application to biological systems is less common but similarly powerful. Current areas of interest include the mathematical theory of information transmission, storage, and understanding in the brain and the investigation of information transmission in DNA.
• Circuits and VLSI (Azita Emami, Ali Hajimiri)
  Analysis, design, simulation, verification, and testing of integrated circuits for various applications, such as high-speed and wireless communications, wireless local-area networks, highly stable frequency sources, distributed integrated circuit design techniques for ultrahigh speed silicon-based circuits, system and circuit design for multi- band systems, single-chip spectrum analyzers, performance limitation of A/D and D/A data converters, and robust circuit- design techniques. Projects also include millimeter-wave silicon-based circuits and arrays, flexible and lightweight arrays, wireless power transfer at distance, silicon-photonics and electronics integration, self-healing circuits, high frequency power generation in CMOS, analysis and design of distributed circuits, multimode reconfigurable systems, as well as modeling of the effect of substrate and supply noise in large integrated circuits and design techniques to minimize their effect, examination of integrated passive structures and their fundamental performance limits, and noise modeling in amplifiers, mixers, and oscillators. More information can be found atchic.caltech.edu and mics.caltech.edu.
• Control and Learning (Venkat Chandrasekaran, John Doyle, Babak Hassibi, Victoria Kostina, Steven Low, Yang Song, Joel Burdick, Mechanical & Civil Engineering, Richard Murray, Control & Dynamical Systems)
  Theoretical research is conducted in all aspects of control, with emphasis on robustness, multivariable and nonlinear systems, optimal control, networked control with information constraints, learning for dynamical systems, online learning and control, control-theoretic perspectives on deep learning, and reinforcement learning Theoretical developments are applied to a wide variety of areas, including internet, autonomous systems, wireless, power systems, cell biology, autonomy, neuroscience, medical physiology, turbulence, wildfire ecology, earthquakes, economics and finance, and foundations of physics.
• Devices (Azita Emami, Ali Hajimiri, Alireza Marandi, Axel Scherer, Yu-Chong Tai, Lihong Wang, Changhuei Yang, Amnon Yariv)
  Research on devices in the Electrical Engineering department deals with a variety of fundamental topics including semiconductor physics, quantum mechanics, electromagnetics, and optics for developing hardware technologies for applications including sensing, communication, computing, and imaging. These devices can range in size from kilometer scale (such as antenna arrays for radio astronomy), to nano scale (such as building blocks of electronic and photonic integrated circuits).
• Electromagnetics, RF, Microwave Circuits, and Antennas (Azita Emami, Andrei Faraon, Ali Hajimiri, Alireza Marandi, Mohammad Mirhosseini, Sander Weinreb)
  Electromagnetics spans most aspects of electrical engineering (and everyday life) from RF, microwave, and mm-wave to infrared and visible optical systems that form the underlying platforms for many of modern marvels. Research involves theoretical study, design, and implementation of devices and systems, such as resonators, radiators, and arrays in microwave and mm-waves, as well as silicon photonics integrated circuits in conjunction with active high speed electronics and photonics integrated circuits. Some examples of such systems include optical phased arrays (OPA) transmitters and receivers, lightweight flexible deployable active phased arrays for communications and power transferm, multimodality quantum computing qubits, and radio astronomy receivers and arrays. This area includes strong theoretical investigations as well as extensive experimental work in realization of these systems in practice.
• Energy and Sustainability (John Doyle, Ali Hajimiri, Babak Hassibi, Steven Low, Guillaume Blanquart (Mechanical & Civil Engineering), Adam Wierman (Computing & Mathematical Sciences)
  Energy research at Caltech encompasses control, optimization, and economics of energy systems, especially future smart grid. The goal is to develop a thorough understanding of the world’s largest and most complex cyber-physical network as well as an intellectual basis for its transformation into a more sustainable, dynamic, and open system. Current research spans optimal power flow problems, convex relaxations, electric vehicle charging, wireless power transfer, space solar power, energy storage, power system dynamics and control, electricity markets, nonconvex pricing, market power, and cascading failure.
• Information and Data Science (Katie Bouman, Venkat Chandrasekaran, Michelle Effros, Babak Hassibi, Victoria Kostina, Yang Song)
  The information and data sciences are concerned with the acquisition, storage, communication, processing, and analysis of data. These intellectual activities have a long history at Caltech both in electrical engineering as well as in allied areas such as applied mathematics. Current research centers around activity at the interface of statistical inference, inverse problems, and machine learning, and has significant synergies with ongoing efforts in information theory and signal processing.
• Information Theory (Yaser Abu-Mostafa, Jehoshua (Shuki) Bruck, Michelle Effros, Babak Hassibi, Victoria Kostina, P.P. Vaidyanathan)
  Theoretical work in a wide range of problems involving transmission, storage and manipulation of information, with strong links to optimization, statistics, control, learning, and wireless communications. Current research foci include bitrate/energy efficiency of computing systems, coding for control, computing with stochastic circuits, error correcting codes for digital storage, coding for delay-sensitive systems, feedback communications, multi-user information theory, parameter estimation with unconventional sampling strategies, random access communications, and random asynchronous computations.
• Learning, Pattern Recognition, and Neural Networks (Yaser Abu-Mostafa, Katie Bouman, Jehoshua [Shuki] Bruck, Michelle Effros, Pietro Perona, Yang Song)
  Theoretical and applied research in machine learning. Current theoretical research directions include estimating the information (data and hints) needed to learn a given task, characterizing the computational complexity of computing with neural networks, and modeling how biological networks of neurons create brain function. Current applied research directions include applications of ML in biology and medicine, in computational imaging, in e-commerce and profiling applications, and in computational finance. Examples of recent projects include a data-driven approach to predicting the spread of COVID-19 in every U.S. county (Abu-Mostafa), an ML approach to medical diagnostics using low-resolution ultrasound (Abu-Mostafa), a learning-based approach to jointly optimize sensor and algorithm designs in computational camera pipelines in order to automatically discover new imaging strategies (Bouman), and a Our latest project is a data-driven approach to predicting the spread of COVID-19 in every U.S. county.theoretical investigation of how networks of neurons form, store, and retrieve memories (Effros). Learning tools are also used for more efficient uncertainty quantification in ill-posed inverse problems, to improve our underlying model of a physical system, and to incorporate more sophisticated knowledge into inference methods.
• MEMS and Micromachining (Alireza Marandi, Mohammad Mirhosseini, Axel Scherer, Yu-Chong Tai)
  We exercise MEMS, Micro- and nanotechnologies to build various sensor and actuator devices. Current research projects focus on bioMEMS and microimplant applications, including integrated biochips, microfluidic chips, neuron chips, blood-count chips, neuroprobes, retinal implants and spinal cord implants, wireless ECG, etc. Hands-on fabrication of these devices is specially emphasized for every student in the laboratory at Caltech.
• Networks and Wireless Communication (Michelle Effros; Babak Hassibi; Victoria Kostina; Adam Wierman, Computing & Mathematical Sciences)
  Research in networks and wireless communication at Caltech spans the spectrum from characterizing theoretical limits of communication system performance to developing algorithms to approximate those limits in practice. Current areas of interest include models and performance analysis for wireless networks, code design and capacity characterization for communication environments characterized by unknown or time-varying channel characteristics, novel modulation schemes, random access communications, remote control over wireless channels, and cloud and edge computing and resource allocation for energy-efficient computing.
• Optimization (Venkat Chandrasekaran, John Doyle, Andrei Faraon, Babak Hassibi, Steven Low)
  Optimization is the science of choosing the best element from a collection subject to some constraints.Current research at Caltech spans the spectrum from theoretical foundations to algorithmic development and eventual deployment in applications. In the context of electrical engineering, optimization methods play a prominent role in signal processing, communications systems design, statistical modeling, control, and machine learning.
• Quantum Science and Engineering (Andrei Faraon, Alireza Marandi, Mohammad Mirhosseini, Lihong Wang)
  Experimental and theoretical research on physical implementations of quantum information processing systems and on quantum imaging and physics. Current areas of research include: quantum computing hardware (memory, logic, interconnects) based on integrated superconducting circuits, photonics, and acoustics components; architectures for scaling quantum networks and modular quantum computing based on quantum transducers; understanding the sources of quantum decoherence in solid-state qubits; quantum imaging and physics. See qse.caltech.edu or http://coilab.caltech.edu/ for further details.
• Signal Processing (Katie Bouman, Venkat Chandrasekaran, Michelle Effros, Georgia Gkioxari, Babak Hassibi, Pietro Perona, Yang Song, P.P. Vaidyanathan)
  Theoretical and computer-oriented work on a wide variety of problems in digital signal processing, with strong links to optimization, statistics, inverse problems and other areas of applied mathematics. Research areas include sparse sensor arrays, sparse signal reconstruction, compressive sensing, phase retrieval, structured signal recovery, high dimensional statistics, array signal processing, multirate digital filters and filter banks, radar signal processing, genomic signal processing, spectrum sensing, graph signal processing, asynchronous random access communication systems, and lossless source coding and channel coding. Other topics include novel machine learning algorithms with applications in computer vision, visual recognition and categorization, and computational imaging with systems that tightly integrate algorithm and sensor design to find hidden signals for scientific discovery and technological innovation.
• Vision (Katie Bouman, Michelle Effros, Georgia Gkioxari, Pietro Perona, Yang Song, Richard Andersen, Computation & Neural Systems, John Allman, Computation & Neural Systems)
  Theory and applications of computer vision. Psychophysics and modeling of the human visual system. Modeling of vision-based decision-making in humans and animals. Current emphasis is on visual object recognition; vision-based human-machine interfaces; perception and modeling of human and animal behavior. Areas of collaboration include statistical machine learning, artificial intelligence, neural networks, computer graphics, neurophysiology, psychology, applied probability, robotics, geometry, and signal processing.