Research On Resource Allocation Algorithm In Visible Light Communication Nerwork

The increasing number of mobile devices and various mobile services is saturating the currently used RF spectrum. Visible light communication(VLC) technology focuses on the unregulated and unused optical spectrum for wireless communication. In addition to its ability to combine illumination and communication, the VLC technology can achieve high data rates. An additional benefit of the different spectrum employed is that VLC does not cause interference to traditional RF communication, thus giving the possibility of hybrid networks that combine VLC with RF.However, a large number of access users also bring pressure on the transmission of upper core network, such as backhaul communication. In heavy load cells, the resource allocation of cells and the access problems of cell edge users are particularly important. When the network is in heavy load,the system can improve the network performance by allocating reasonably the wireless resources and balancing the load. Load balancing is an effective way to increase the network throughput and the network flexibility.Therefore, fair and efficient load balancing can be a challenge in a VLC network.This thesis aims to solve the load balancing mechanism of resource management in visible light communication. The load balancing strategy and the resource allocation algorithm of VLC network are proposed to improve user fairness, promote system throughput and reduce user handover cost. And the validity of the proposed algorithm is verified by software simulation platform. The main contributions of this thesis are as follows:First, according to the match theory, the thesis proposes a mobility-aware load balancing scheme in hybrid VLC-LTE networks, which dynamically allocates the users to their corresponding APs. Applying matching theory, the original allocation problem is formulated as a college admission problem model (CAPM). Considering the performance metrics of throughput, handover overheads, and fairness, the preferences of CAPM combines the channel condition, users' mobility, and the local fairness index.Meanwhile, this paper proves our proposed preferences are responsive, and therefore, the set of stable matching is not empty. The outcome of the matching procedure is an optimal stable matching from the users' point of view. The system-level simulation results show that the proposed matching scheme significantly improves the system performance in terms of system throughput, handover overheads and network fairness.Second, taking resource allocation into account, we propose a load balancing strategy for visible heterogeneous networks. The strategy is illustrated as follows. According to the system limitation of bandwidth and power, we distribute bandwidth and frequency resources to maximize system throughput. An exhaustive search iterative algorithm is developed for users to select the optimal access network and corresponding bandwidth and power resources. The resource allocation task is decoupled into two sub-problems: 1) Using extreme value method allocates frequency resources for users according to concavity of the objective function. 2) Considering the users' Quality of Service (QoS) requirements, channel utilization and users'fairness, an improved the water-filling algorithm is utilized. Simulation results show that the proposed algorithms and architecture for user model with strong compatibility, and better guarantee the user fairness while improving the system throughput. Therefore, the proposed algorithm achieves the purpose of design.