Localization-based device-to-device discovery

As spectrum scarcity becomes more serious, device-to-device (D2D) communication attracts great attention. It allows devices in proximity to communicate directly, improving network capacity and resource utilization over the cellular network. In D2D communications, an important process is to discover neighbours and set up the D2D link. D2D applications benefit from D2D discovery as an end-to-end link maintenance or as a transmission relay when the direct path is obstructed. A novel energy efficiency D2D link setup scheme is proposed in cellular networks by employing localization information that can be easily obtained using new radio technologies. In the principle of the proposed D2D discovery schemes, to set up a D2D link, the obtained location estimates are used to evaluate whether two user devices can be a D2D pair and to control the transmission power between the two devices, considering the accuracy of the estimated distance between them. A power allocation algorithm was proposed for the scheme to overcome localization inaccuracy and improve the D2D link setup success rate while delivering significant energy efficiency. Simulation results show that the scheme eliminates the neighbour discovery step, resulting in faster setup of a D2D communication link, and significantly improves energy efficiency and resource block utilization while maintaining a high success probability for D2D discovery and communication setup. Frequent reuse is a useful way to increase spectral efficiency in cellular communications. Resource utilisation would significantly increase if frequency reuse could be applied to D2D discovery. Therefore, based on localization information in this thesis, a frequency reuse scheme is proposed by allowing D2D pairs to use the same frequency for D2D discovery to increase resource utilization. When frequency reuse is permitted in the D2D discovery, intra-cell interference will affect the discovery Abstract 5 accuracy and become the main challenge of D2D discovery. Therefore, the thesis considers an intelligent power allocation scheme at the base station (BS) to overcome intra-cell interference caused by frequency reuse. In the proposed power allocation algorithm, the impact of localization inaccuracy is also addressed when allocating the same frequency to different D2D pairs, with the objective of minimizing energy consumption under a given outage probability or the lowest overall outage probability. The algorithm could be extended to overcome inaccuracies of different localization technologies. The lowest overall outage probability for both D2D pairs using the same frequency under the power allocation algorithm was determined and studied. When the transmit power increases, the lowest overall outage probability is limited by the received interference. The lowest overall outage probability depends on the ratio between intra-pair and inter-pair distances from D2D pairs. The proposed algorithm could allocate the most efficient power with the two different outage probabilities of the two D2D pairs while achieving the lowest overall outage probability and highest energy efficiency under intra-cell interference.