Research Of Resource Allocation Scheme For Massive MIMO In Cognitive Radio Ad Hoc Networks

In recent years, people have carried out a wide range of researches in CRAHNs(Cognitive Radio Ad Hoc Networks) for improving spectral efficiency, speeding up the communication rate, increasing the reliability of communication and decreasing the interference to primary users. Massive MIMO has great advantage in enhancing spectrum and energy efficiency, thus Massive MIMO is utilized in CRAHNs to improve the network communication capability and decrease energy consumption.This paper mainly focuses on the resource allocation problem based on Massive MIMO in CRAHNs.(1) In the case of perfect and imperfect CSI(Channel State Information), this paper does some derivations about the lower capacity bounds of Massive MIMO uplink systems when they adopt MRC, ZF and MMSE detector. This paper also conducts some simulation to prove the lower capacity bounds are very close to the ideal ones. Besides, based on the lower capacity bounds, a resource allocation model is proposed which takes into account the maximum power of each node.(2) In the circumstance that the cluster head utilizes MRC and ZF detector, energy-efficient resource allocation algorithms(i.e. UMRA and UZRA) are proposed. UMRA is modeled as a lower energy.efficient optimization which takes the circuit power consumption, minimum required rate and maximum tolerable interference level into consideration. Under the he minimum required rate and maximum power constraint, UZMA sets up a non.convex optimization which takes lower energy.efficient as objective function. According to fractional programming, the resulting energy.efficient optimization in the fractional form is transformed into subtractive form. Then convex optimization is exploited to obtain the best energy efficiency(bit/Hz/J). Simulation results show that the proposed algorithms are excellent efficiency.(3) Based on the capacity of single MN OFDM system, a resource allocation is proposed which exploits water-filling algorithm(WFA) to allocate power. WFA and mean power allocation(MPA) have been simulated, and it shows that WFA performs better. For multi-SNs, an energy efficiency resource allocation algorithm(DMRA) is proposed for MF Precoding. In consideration of channel interruption, maximum power and minimum required rate, DMRA is modeled as a non.convex optimization based on energy efficiency. The energy-efficient optimization is transformed into subtractive form. Then equivalent data rate is adopted in the new optimization to satisfy channel outage probability. At last, the new optimization is transformed into standard convex problem, and dual.decomposition is used to solve it. DMRA cooperatively adjusts subcarrier power, data rate and the number of antennas for optimal performance. The simulation results indicate that DMRA can converge to energy-efficient optimization in a small number of iterations.