提出一种跨层设计机制,目的是最大化网络下行链路吞吐量,同时保证业务公平性.网络架构为多小区、多用户、多业务的Macro-Femto异构网络架构.物理层传输技术为多输入多输出正交频分复用技术.所建立的最优化问题在多项式时间内是不可解的.结合工程实际,提出一种易于实现的次优算法.此外,针对异质业务环境,提出一种公平性观点,并改进传统的Jain's公平指数.修正的公平指数能够有效地评估未来多业务无线网络各种机制的公平性.系统级仿真结果表明:在异构网络中,本文提出机制的吞吐量性能和公平性能均优于简单跨层设计SCLD(Simple Cross-Layer Design)机制.本文提出的机制和SCLD机制的业务公平性能均优于比例公平机制的公平性能.比例公平机制仅考虑速率公平,没有考虑业务公平.%A new cross-layer scheme was developed,aiming to maximize the whole-network downlink throughput and guarantee service fairness. The networks were the multi-cell, multi-user and multi-service Macro-Femto heterogeneous networks (HetNets). The transmission technology was multiple-input multiple-output orthogonal frequency division multiplexing (MIMO-OFDM) . Optimization was non-deterministic polynomial-time hard (NP-hard). Thus, from the practical engineering perspective, a feasible suboptimal algorithm was put forward. Additionally,a perspective of fairness for the heterogeneous service environment was set, and the traditional Jain's fairness index was improved. The improved index is more efficient in evaluating fairness of different kinds of schemes in future multi-service wireless networks. System level simulation results show that the proposed scheme is superior to the simple cross-layer design (SCLD) scheme in terms of average sector throughput and fairness. The proposed and SCLD schemes both provide better fairness performance than the proportional fair scheme which focuses on rate fairness rather than on both rate and quality of service (QoS) fairness.
展开▼
