PhD thesis: Energy-Efficient Wireless Communications
Download  : http://www.cs.utwente.nl/~smitl/pubs
Author    : Lodewijk T. Smit

Abstract:

Mobile devices have limited battery capacity and operate in a frequently
changing environment. This thesis describes a control system that adapts the
amount of signal processing needed for the wireless communication to the
current reception quality at run-time. The benefits are twofold. In case of a
bad reception, the amount of signal processing is increased so that the
connection still delivers the quality of service that is requested by the
user. In case of a good reception, the amount of signal processing is
decreased so that the energy consumption is reduced. In this way the system is
optimized for the typical case at run-time instead of being optimized for the
worst case at design-time. To be able to adapt to the current environment, the
control system has to be aware of the quality of this environment. The thesis
introduces a bit error rate (BER) estimation algorithm that is able to
estimate the current BER and to predict the BER if the control system plans to
change parameters.
Two concrete examples are worked out for the Universal Mobile
Telecommunications System (UMTS) standard (using Wideband Code Division
Multiple Access (WCDMA) technology) and for Wireless Local Area Network (LAN):
the HIgh PERformance LAN (HIPERLAN/2) standard similar to the IEEE 801.11a
standard. Both LAN standards are based on Orthogonal Frequency Division
Multiplexing (OFDM) technology. It turns out that global optimization over
different components of a receiver delivers more gain (in terms of energy
consumption as well as quality) than local optimization of the individual
parts of a receiver. The thesis shows that adaptation to the current
environment to minimize the energy consumption, while still delivering an
adequate quality of service to the end-user at run-time is useful and
possible.