Thesis Abstract and Dissertation Abstracts

Cooperative relaying in a cognitive radio network (CRN) has been used to achieve diversity gain and energy eciency through the relay node. This dissertation investigates the Neuro-Fuzzy (NF) based relay selection, spectrum access and fair resource allocation for cognitive cooperative networks. The main idea is to increase the data transmission rate and spectrum utilization efficiency while decreasing the outage probability and complexity of the system.

In a dual-hop heterogeneous network, the primary radio network (PRN) and CRN use the best relay to assist the source communication. This study focuses on a joint radio resource allocation (subcarrier, power)-relay selection problem for an Amplify-and-Forward (AF) cooperative network. A priority based suboptimal resource allocation algorithms for heterogeneous, i.e., real time (RT) and non-real time (NRT), services have been proposed for the PRN. The aim is to maximize the system throughput while satisfying the minimum quality of service (QoS) requirement of RT and NRT users. The simulation results are compared with the fixed as well as dynamic resource allocation algorithms proposed in different researches. The proposed scheme reduces outage probability and increases throughput of the system. A power allocation optimization problem is also formulated to maximize the spectral efficiency of dual-hop CRN under total transmit-peak and average interference power constrains. Numerical results show that relay-assisted CRN (RCRN) provides significant performance improvement over single-hop CRN (SCRN).

NF selection algorithm for optimal relaying in cooperative networks is proposed. The aim is to select the best relay in order to minimize the system outage probability and bit-error-rate (BER). The proposed algorithm takes into consideration instantaneous signal-to-noise ratio (SNR), link delay (propagation and queuing delay) and energy saving due to cooperative diversity. The performance of the proposed NF based relay selection algorithm is compared with blind search, informed search, Fuzzy-based search and selection AF with power allocation algorithms. The outage probability and BER expressions have been derived. Simulation results depict close consensus between analytical results and simulations. NF selection algorithm provides better performance improvement over conventional algorithms. The complexity analysis shows that the proposed algorithm requires less time to select the best relay.

We investigate opportunistic spectrum access and spectrum hand-o (HO) for RCRN using intelligent NF based approach. The main aim is to increase the average throughput of the RCRN while reducing the number of HO decisions. Without the coordination among SUs, NF based algorithm enables each SU to determine the most appropriate channel for its data transmission. SUs maintain their transmission power below the interference temperature through the power control and do not interfere the normal operation of a PRN. If QoS of the SU is not maintained, HO decision is performed. The proposed algorithm is compared with existing algorithms. Performance evaluations reveal that the outage probability is reduced and the system's capacity is increased.

Dissertation Images:
	Shamim (click to enlarge)