Collaborative Research: Delay and Energy: Design Tradeoffs in-Spectrally Efficient Systems
Grant
Overview
Affiliation
Other
View All
Overview
abstract
Energy consumption has become of increasing importance by the transition in communications to mobile devices with limited battery capacity. At the same time, data transmission delay needs to be carefully controlled due to the need to support the proliferation of time-sensitive services, such as video. Seen from a user perspective, these are the main things that affect the mobile experience: energy consumption (how long does the battery last), delay (how long to wait for content). From a societal perspective, wireless communications contribute significantly to global energy consumption and carbon dioxide emissions. An estimate is that currently 0.5-1% of global carbon dioxide emissions is due directly to wireless communications, comparable to that of air traffic (2%). At the same time, the number of devices that share the wireless medium is increasing dramatically. Yet, the desire for low delay and low energy consumption in congested systems conflict. In order to design future communication systems, it is therefore necessary to understand the fundamental tradeoff between energy and delay in congested systems. The insights from this fundamental analysis will be used to jointly design new communications software and energy efficient communication hardware, which will be tested in a wireless testbed. This hardware can be used as a basis for future wireless communication devices. The project will involve underrepresented minorities, including native Hawaiians, in undergraduate research experiences to encourage the participation of these minorities in the STEM field. The project aims to understand the factors that influence energy and delay such as the time characteristics of data to be transmitted and the characteristics of the channel. To obtain a fundamental understanding of these factors, it is necessary to analyze the information theory underlying delay and energy, taking into account realistic features of the systems, including networking and hardware. The project builds on the recent line of work in information theory on finite block length communications. Basically, if the block length is infinite, as in traditional information theory, the delay is also infinite. Therefore, in order to reach a more fundamental understanding of delay, finite block length analysis is needed. However, block length is not a direct indicator of delay. Therefore, finite block length theory has to be developed specifically for investigating delay and energy. Preliminary results show that delay affects energy consumption much more strongly than one would expect from bandwidth constraints alone. This project studies the tradeoff between delay and energy in four steps. First it develops the basic information theory relating delay and energy. It then applies this to multiuser systems, where spectral efficiency is essential. In the next step, it extends the models with realistic hardware constraints. In the final step, the theory is tested on a hardware testbed with state-of-the-art coding. This award reflects NSF''s statutory mission and has been deemed worthy of support through evaluation using the Foundation''s intellectual merit and broader impacts review criteria.
