Swindlehurst Wins ICC Best Paper Award

Swindlehurst and colleagues explore machine learning to improve performance and reduce cost and power usage in next-generation wireless communications systems.

May 27, 2020 - The IEEE International Conference on Communications (ICC) has awarded Samueli School professor Lee Swindlehurst and two colleagues a 2020 Best Paper Award for their work that details using a type of machine learning to improve performance in multi-antenna wireless communications systems.

Swindlehurst, professor of electrical engineering and computer science, collaborated on “SVM-based Channel Estimation and Data Detection for Massive MIMO Systems with One-Bit ADCs” with San Diego State professor Duy Nguyen and Ly Van Nugyen, a doctoral student in the UC Irvine/SDSU Joint Program in Computational Science. The paper describes using a classical method of machine learning, called support vector machines, to target new 5G-and-beyond systems, which employ large arrays of antennas. These arrangements, called massive MIMO – multi-input, multi-output wireless systems – can be costly and use a lot of power, so researchers seek ways to reduce price and power usage without sacrificing performance. “Our paper shows how to use machine learning and signal processing to reduce the performance loss that results when very low resolution (one-bit) sampling hardware, referred to as analog-to-digital converters, are used in a massive MIMO implementation,” Swindlehurst said.

ICC, the flagship conference of the IEEE Communications Society, is schedule to take place June 7-11. It was to have been held in Ireland, but due to coronavirus will be held online instead, much to Swindlehurst’s disappointment. “I was really looking forward to going,” he said.

-Anna Lynn Spitzer