Research On Energy Efficient Resource Allocation In Frequency Division Multiplexing Systems

Wireless mobile system mainly focuses on how to overcome the unbalancebetween quality-of-service (QoS) and limited resources, such as bandwidth and transmitpower. Spectrum is a non-renewable natural resource that must be utilized as efficientlyas possible. However, with the rapidly rising energy consumption and increasingconcern over global warming, the energy-efficient oriented design for greencommunications has become a hot research direction.Frequency division multiplexing (FDM) systems is nowadays widely used inseveral international standards due to its inherent resistance to multipath propagationand the flexibility in resource allocation. The conventional design on resource allocationin FDM systems can be classified into two major categories:1) maximizing the spectralefficiency (SE) with a maximum transmit power constraint; and2) minimizing thetransmit power for a given data rate. However, both of these two optimization methodsmay not performance well in terms of energy efficiency (EE). This thesis investigateshow to configure the system to operate in the most energy efficient manner whileensuring a certain system SE from different perspectives.This thesis first introduces the background of resource allocation in FDM systemsand lists some conventional technical specifications. By analyzing the relationshipbetween SE and EE, the main content of research is determined.Chapter2describes the weighted-sum rate maximization problem in multiuserorthogonal frequency division multiple access (OFDMA) systems. Accurate channelstate information (CSI) is assumed to be available to the transmitter. The maximum SEis achieved through Lagrangian duality method. The energy efficient resource allocationfor OFDMA systems with low complexity is proposed. In order to reduce thecomputational burden, we divide this optimization problem into two subproblems:subcarrier scheduling and power allocation. In contrast to traditional scheme, weanalyze the relationship between EE and water level rather than SE. The EE is aquasi-concave function of water level. A closed form solution for optimal power allocation with the given subcarrier assignment is derived in each iteration.Chapter3studies the energy efficient resource allocation in both the multiusersingle-input single-output (SISO) and virtual multiple-input multiple-output (MIMO) ofsingle carrier frequency division multiple access (SC-FDMA) systems. A closed formsolution for optimal energy efficient power allocation of each user is derived with thefixed subcarrier scheduling. Simulation results show that the proposed scheme can get abetter balance between SE and EE.Chapter4formulates the optimization model of EE in FDM systems with differentpeak-to-average power ratio (PAPR) reduction methods. In the time domain, thetransmit signal has a large PAPR, which translates to low amplifier efficiency. Both ofthe EE and the effects of system parameters on performance are considered.Finally, we conclude this thesis and suggest some possible areas for futureresearch.