Page 32 - SPECTRUM

What does Machine Learning

have for an EC Engineer?

Dr. Ruchi Gajjar
Assistant Professor

Being an EC Engineer, which are the areas where I can employ machine learn-

ing (ML)/ Deep learning (DL)? This could be the question that would be bug-

ging a lot of students like you, these days. The field of ML and DL is not just
limited to computer/ IT applications like face recognition, object detection,

or prediction of weather and stock market. You can apply this buzz word and
booming technology in applications related to Electronics and Communica-

tion – your own degree of Engineering. For instance, let’s talk about:
• VLSI design, then ML is currently used in Chip Design (e.g., new intercon-

nect fabrics, new combinations of memory and computation, etc.), predicting
places where chip may experience manufacturing defects, load prediction on

CPU, voltage scaling to save energy. For an ASIC design, ML can be applied
for RTL code analysis to detect and correct problems for scan insertion or for

coding guideline violations, Regression analysis in Verification for identifying

test cases, in Synthesis for early detection of issues with floorplanning or
congestion early, before and after the layout.

• Electronics, where ML is used for prediction of successful field-pro-
grammable gate arrays (FPGA) compilation strategies, behavioral modeling

of microelectronic circuits and systems, to predict the Power/performance/
area (PPA) given a register-transfer level description of a circuit, eliminating

the need to undertake the lengthy physical design process.

• Antenna and Wireless Communication, where ML is used for parameter

optimization in antenna design and Wireless Communication offers a wide
scope for ML in areas like channel modeling, signal estimation and detection,

energy efficiency, cognitive radios, wireless sensor networks, vehicular com-
munications, and wireless multimedia communications. To give you a bet-

ter idea, ML is used for resource management like power control, spectrum
management, backhaul management, cache management, and beamformer

design and computation resource management in the MAC layer, networking
and mobility management in the network layer for applications in clustering,

SPECTRUM Pg. 27 ISSUE 1