Scheduling algorithms for LAA and SWIPT networks

Abstract

In the fifth generation (5G) networks, several objectives such as increasing data rate, decreased latency, and enabling massive connectivity should be addressed for a better quality of service (QoS) in accordance with various wireless applications. To satisfy these objectives, radio resource management (RRM) schemes have been significantly considered since the mobile data traffic and wireless devices are exponentially increased. In wireless mobile communication systems, the available radio resources consist of spectrum, energy/power, time, space and so on. These resources can be divided into two categories: direct recourses which can directly affect wireless communication systems and indirect resources which do not directly affect systems. The spectrum, energy/power, and time resources could belong to the direct resources. As the indirect resources, there are diversity and degree of freedom (DoF) gain resulted from the randomness of fading channel environment. In this dissertation, author handles the decentralized scheduling algorithm for long term evolution - licensed-assisted access (LTE-LAA) networks and the centralized scheduling algorithm for simultaneous wireless information and power transfer (SWIPT) - enabled Internet of Thing (IoT) networks to improve the performance.

Firstly, this dissertation investigates the efficient usage of spectrum resources in the unlicensed spectrum where the wireless local area network (WLAN) (e.g., Wi-Fi) and LTE-LAA networks are coexisted under non-saturated traffic condition. In the unlicensed spectrum, the LAA networks perform the decentralized scheduling algorithm called listen-before-talk (LBT) protocol to access the channel with guaranteeing the performance of incumbent Wi-Fi networks. To evaluate coexistence scenarios of the Wi-Fi and LAA networks, an analytical framework is proposed under the non-saturated traffic condition. The optimal contention window (CW) sizes can be obtained according to the traffic densities of Wi-Fi and cellular networks. In addition, the LBT schemes are proposed based on various channel sensing mechanisms and compare the performance in terms of energy-efficiency. From the results, the some guidelines in terms of the adjustment of CW size and the LBT operations according to the traffic density can be provided to cellular operators who want to operate the cellular systems in the unlicensed spectrum.

Secondly, for energy management, author studies SWIPT-enabled IoT networks. In multi-cell single input single output (SISO) environment, a novel weighted proportional fairness (weighted-PF) scheduling algorithm is proposed for SWIPT networks. In the SWIPT-enabled networks, the fairness issues can be occurred among users with respect to harvested energy as well as achievable data rate. To solve the above problem, the fairness measure of harvested energy is defined. Then, the proposed weighted-PF scheduler shows the performance improvement with respect to the fairness of achievable rate without significant degradation of other performance such as achievable data rate and fairness performance for harvested energy. Therefore the proposed scheduler can be regarded as a proper scheduling method for IoT or massive connectivity networks.

Through the proposed the centralized / decentralized scheduling algorithms in this dissertation, various performance can be improved for 5G or beyond 5G wireless communication systems. As future works, licensed spectrum access (LSA) where the licensed spectrum can be shared and multi-cell multi-user multiple antenna systems in SWIPT networks will be considered.