Abstract:
Non-Orthogonal Multiple Access (NOMA) is considered to be one of the possible enabling solutions for achieving the high spectrum efficiency required for 5G commu- nication. NOMA allows multiple users to share time and frequency resources in the same spatial layer via power domain or code domain multiplexing. Most of the exist- ing resource allocation works in NOMA systems are based on capacity maximization criterion. Capacity maximization criterion based resource allocation optimization framework results on full utilization of power budget and consequently, incurs huge energy consumption. Unlike to prior works on resource allocation in NOMA sys- tems, the goal of this work is to study Quality of Service (QoS) constrained transmit power and bits allocation based on power minimization criterion. QoS is measured in terms of Signal to Interference plus Noise Ratio (SINR)/ Bit Error Rate (BER) and minimum bits requirement. Specifically this research work considers a down- link Non-orthogonal Multiple Access (NOMA) system where Base Station (BS) can communicate its single cognitive user through multiple channels as well as multiple users can be communicated through a single channel. Under such system model, in this work, two proposed optimization frameworks introduce two separate approaches to minimize the transmit power and maximize bits without degrading the QoS. The former framework (referred as two-stage optimization) splits the objective function into two sequential stages. In the first stage, optimal transmit power to maintain a certain SINR is determined; using that result, the optimal bits allocation strategy is calculated in the second stage. The later framework, which is referred as joint optimization, both the transmit power and the bits are optimized concurrently by
xviii
xix
employing multi-objective optimization framework. Simulation results show that, BS requires to transmit with more power in a channel associated with more noise power; whereas, allocation of bits is higher in a channel with higher SINR and vice versa. It is also observed that, the two-stage optimization framework adapts only one vari- able in each stage; power in the first stage, bits in the second stage. This causes a lower degree of freedom. On the other hand, the joint optimization framework adapts two variables simultaneously, results higher degrees of freedom. The higher degrees of freedom make the joint optimization framework more powerful in satisfying QoS requirement of users.
