Optimum Design of Spectral Efficient Green Wireless Communications

This dissertation focuses on the optimum design of spectral efficient green wireless communications. Energy efficiency (EE), which is defined as the inverse of average energy required to successfully deliver one information bit from a source to its destination, and spectral efficiency (SE), which is defined as the average data rate per unit bandwidth, are two fundamental performance metrics of wireless communication systems. We study the optimum designs of a wide range of practical wireless communication systems that can either maximize EE, or SE, or achieve a balanced tradeoff between the two metrics. There are three objectives in this dissertation. First, an accurate frame error rate (FER) expression is developed for practical coded wireless communication systems operating in quasi-static Rayleigh fading channels. The new FER expression enables the accurate modeling of EE and SE for various wireless communication systems. Second, the optimum designs of automatic repeat request (ARQ) and hybrid ARQ (HARQ) systems are performed to by using the EE and SE as design metrics. Specifically, a new metric of normalized EE, which is defined as the EE normalized by the SE, is proposed to achieve a balanced tradeoff between the EE and SE. Third, a robust frequency-domain on-off accumulative transmission (OOAT) scheme has been developed to achieve collision-tolerant media access control (CT-MAC) in a wireless network. The proposed frequency domain OOAT scheme can improve the SE and EE by allowing multiple users to transmit simultaneously over the same frequency bands, and the signal collisions at the receiver can be resolved by using signal processing techniques in the physical layer.