Research On Indoor Millimeter Wave D2D Communication Resource Optimization Algorithms

The specific characteristics of millimeter wave(mmWave)technology can meet the requirements of large amount of data communications among smart devices for indoor scenario.Reducing interference in enclosed areas while maintaining high throughput remains a challenge.(1)First,a random model is proposed to capture the characteristics of high-density enclosed indoor areas for the transmission performance of different locations.Due to the reflections of the smooth surface of the indoor scenario,the probability that blocked by obstacles will appear fluctuation.An indoor reflection model by considering the first-order reflections of the walls and ceilings is built.At the same time,considering that the human body is the main blockage in addition to the conventional blockage,the dissertation also combines the multi-LOS ball with the self-blocking model.Thus,the characteristics of conventional blockage and human can be easily caputured.The self-blocked model is used to characterize the effects of the human body,while the multi-ball line of sight(LOS)model is used to describe the traditional blockage.Therefore a complete indoor blockage model can be completed.Combined with the reflection model and the blockage model,this dissertation can characterize the indoor scene more comprehensively.(2)Then,the stochastic geometry is used to calculate the threshold range to divide the strong interferer and weak interferer.A closed approximation expression of the coverage probability is dereived to demonstrate the accuracy of the model.The results show that the proposed model can accurately describe the reflection blockage in indoor conditions.(3)Finally,this dissertation represents resource allocation as an optimization problem in different location scenarios after modeling.The goal of this optimization is to achieve maximum throughput by minimizing interference.In order to achieve this goal,an improved Powell Multi-Vertex Algorithm(PMVC)for indoor center locations scene is proposed,so that more communication flows can be placed evenly in different time slots.Numerical experiments show that the improved algorithm can achieve higher system throughput than the traditional greedy algorithm and the original Powell Vertex Coloring(PVC)scheme.On the basis of completing the indoor central location resource scheduling,this dissertation proposes a combined temperature random algorithm(CTRA)to arrange the reasonable allocation of resources with reference to the transmission characteristics of the corner position.The CTRA algorithm is divided into two phases: partitioning and matching.At the time of partitioning,the edge weights are calculated using the transmission distance(including the first-order reflection)as a metric.And use the meric to characterize the interference level between the two transmission links in order to minimize interference.In the matching phase,the definition function compares the performance of the system under a specific temperature factor,so that the CTRA algorithm will approximate the global optimum with a higher probability.This algorithm improves the average throughput and is significantly better than other traditional algorithms.