Present and future generation wireless systems are designed to support a wide variety of services, including video conferencing, media streaming, Internet access, email transfer, remote file access and web-browsing. Different applications have different traffic characteristics and quality of service (QoS) requirements. These QoS parameters include delay constraint, packet loss rate (PLR) and packet error rate (PER) requirement. For example, depending on the throughput or delay requirements of the services, wireless data services have already been classified as guaranteed or best-effort services. Mobile wireless links are time-varying in nature and error prone. In addition, communication resources, e.g., transmit power and bandwidth, are scarce. Through a cross-layer design approach, these inherent quality parameters of different services can facilitate efficient use of limited communication resources. In particular, the relative delay tolerance of data applications, in conjunction with the bursty activity patterns and time-varying nature of the wireless channels, opens up the possibility of scheduling transmissions to efficiently allocate limited wireless communication resources.

The main contribution of this dissertation is the design of innovative scheduling schemes through cross-layer design approach that jointly considers the time-varying nature of wireless channels as well as the required QoS parameters of different services. This dissertation makes four major contributions, as follows. First, we develop modulation-assisted two-user opportunistic scheduling schemes. The novelty of the proposed schemes is improved delay performance and fairness for the serviced users in wireless networks without degradation of the system spectral efficiency (SE), compared with the classical single best user scheduling (SBS) scheme. Second, we develop an adaptive scheduling protocol for multi-resolution image/video data transmission over time-varying channels. The protocol is optimized for real-time multi-resolution data transmission in order to maintain a given quality of the received multi-resolution data. Third, we devise a novel variance-bounded rate adaptation (VBRA) scheme for delay-constrained communications. We explore an application of the proposed VBRA scheme for media streaming over time-varying channels in order to minimize the transmit power while maintaining a low play-out buffer outage probability. Fourth, we develop link adaptation algorithms for a multicarrier system for generalized delay-constrained communication over time-varying channels.