Link Adaptation for Energy Efficiency Maximization of Practical Wireless Communication Systems
Sep 17, 2013
from 11:00 AM to 01:00 PM
|Where||Engr. IV Bldg., Tesla Room 53-125|
|Contact Name||Eren Eraslan|
|Add event to calendar||
Advisor: Prof. Babak Daneshrad
Energy consumption of the modern wireless communication systems is rapidly growing due to the ever-increasing data demand and the advanced solutions employed in order to address this demand, such as MIMO and OFDM. These MIMO systems are power hungry, however, they are capable of changing the transmission parameters such as number of spatial streams, number of transmitter/receiver antennas, modulation, code rate, and transmit power. They can thus choose the best mode out of possibly thousands of modes in order to optimize an objective function. This problem is referred to as the link adaptation problem.
In this work, we focus on the link adaptation for energy efficiency (EE) maximization problem which is defined as choosing the optimal transmission mode to maximize the number of successfully transmitted bits per unit energy consumed by the link. We model the energy consumption and throughput performances of a MIMO-OFDM link and develop a practical link adaptation protocol which senses the channel conditions and changes its transmission mode in real-time. It turns out that the brute force search, which is usually assumed in previous works, is prohibitively complex, especially when there are large numbers of transmit power levels to choose from. We analyze the EE and transmit power relationship, and prove that EE is a single-peaked quasiconcave function of transmit power. This leads us to develop a low-complexity algorithm that finds a near-optimal transmit power and take this dimension out of the search space. We further prune the search space by analyzing the singular value decomposition of the channel and excluding the modes that use higher number of spatial streams than the channel can support. These algorithms and our novel formulations provide simpler computations and limit the search space into a much smaller set, and hence reduce the computational complexity by orders of magnitude without sacrificing the performance.
The result of this work is a highly practical link adaptation protocol for maximizing the energy efficiency of modern wireless communication systems. Simulation results show orders of magnitude gain in energy efficiency of the link. We also implemented the link adaptation protocol on real-time MIMO-OFDM radios and we report on the experimental results. To the best of our knowledge, this is the first reported testbed that is capable of performing energy-efficient fast link adaptation using PHY layer information.
Eren Eraslan received his BS degree from Bilkent University, Ankara, Turkey in 2008, and his MS degree from University of California, Los Angeles (UCLA) in 2010 both in electrical engineering. He is currently a PhD candidate at UCLA where he conducts research on green wireless communications and MIMO systems as a member of the WISR laboratory. In 2012, he was with Broadcom Corp. where he worked on adaptive time-frequency domain channel interpolation techniques for digital TV receiver ICs, and in 2010 he worked at Silvus Technologies Inc. where he developed channel tracking algorithms for mobile MIMO links.