Orthogonal frequency division multiplexing¡]OFDM¡^systems play an important role in modern wireless communications due to following advantages: bandwidth saving¡Bcombat with frequency selective fading channel and high throughput. The performance of wireless communications is often degraded by fading channel . adaptive modulation and subcarrier allocation are proposed to overcome the degration to meet the quality of servie¡]QoS¡^. Lagrange method and heuristics method, two of the subcarrier allocation technology under multi-user OFDM, can achieve the goal that maximizing bit rate with minimizing transmitted power. However, significantly high complexity of either Lagrange method or heuristics method makes the implementation difficult. Zhang and Letaief proposed a method of making subcarriers detected one by one to reduce the complexity. However, in piratical, an OFDM system accommodates hundred of , or even thousand of subcarriers, so the method can be improved. In this thesis, we propose a subcarrier allocation method. The users that are not satisfied with the QOS requirement are named demander, and the users satisfied with the QOS requirement are named supplier. In the proposed subcarrier allocation method, we evaluate the number of subcarriers that demanders need and remove the subcarriers from supplier to directly compensate demander. Then the system has lower complexity due to less iterations.

This study considers the average system throughput and the average delay performances of subcarrier allocation algorithms in OFDMA systems. The effects of varying the number of users, the number of subcarriers, and the statistical characteristics of incoming packets are investigated on the throughput and delay performances of the algorithms. Moreover, a new subcarrier allocation algorithm with low-order computational complexity, which performs very well almost all cases, is proposed. With the aid of the simulations, the significance of channel v.s. queue state information varying with the statistical characteristic of incoming packets is examined and reached some results which can be very valuable for channel estimation and feedback systems. Finally, the stability issue is considered in OFDMA systems and a new heuristic simulation-based method for obtaining the stability region of an OFDMA subcarrier allocation algorithm is proposed.

Advances in communications technologies entail demands for higher data rates. One of the popular solutions to fulfill this requirement was to allocate additional bandwidth, which unfortunately is not anymore viable due to spectrum scarcity. In addition, spectrum measurements around the globe have revealed the fact that the available spectrum is under-utilized. One of the most remarkable solutions to cope with the under-utilization of radio-frequency (RF) spectrum is the concept of cognitive radio (CR) with spectrum sharing features, also referred to as spectrum sharing systems. In CR systems, the main implementation issue is spectrum sensing because of the uncertainties in propagation channel, hidden primary user (PU) problem, sensing duration and security issues. Hence, the accuracy and reliability of the spectrum sensing information may inherently be suspicious and questionable. Due to the imprecise spectrum sensing information, this dissertation investigates the performance of an orthogonal frequency-division multiplexing (OFDM)-based CR spectrum sharing communication system that assumes random allocation and absence of the PU's channel occupation information, i.e., no spectrum sensing is employed to acquire information about the availability of unused subcarriers or the PU's activity. In addition, no cooperation occurs between the transmitters of the PUs and secondary users (SUs). The main benefit of random subcarrier utilization is to uniformly distribute the amount of SUs' interference among the PUs' subcarriers, which can be termed as interference spreading. The analysis and performance of such a communication set-up provides useful insights and can be utilized as a valid benchmark for performance comparison studies in CR spectrum sharing systems that assume the availability of spectrum sensing information. In the first part this dissertation, due to the lack of information about PUs' activities, the SU randomly allocates the subcarriers of the primary network and collide with the PUs' subcarriers with a certain probability. The average capacity of SU with subcarrier collisions is employed as performance measure to investigate the proposed random allocation scheme for both general and Rayleigh channel fading models. In the presence of multiple SUs, the multiuser diversity gain of SUs is also investigated. To avoid the subcarrier collisions at the SUs due to the random allocation scheme and to obtain the maximum sum rate for SUs based on the available subcarriers, an efficient centralized sequential algorithm based on the opportunistic scheduling and random allocation (utilization) methods is proposed to ensure the orthogonality of assigned subcarriers. In the second part of this dissertation, in addition to the collisions between the SUs and PUs, the inter-cell collisions among the subcarriers of SUs (belonging to different cells) are assumed to occur due to the inherent nature of random access scheme. A stochastic analysis of the number of subcarrier collisions between the SUs' and PU's subcarriers assuming fixed and random number of subcarriers requirements for each user is conducted. The performance of the random scheme in terms of capacity and capacity (rate) loss caused by the subcarrier collisions is investigated by assuming an interference power constraint at PU to protect its operation. Lastly, a theoretical channel fading model, termed hyper fading channel model, that is suitable to the dynamic nature of CR channel is proposed and analyzed. To perform a general analysis, the achievable average capacity of CR spectrum sharing systems over the proposed dynamic fading environments is studied.