蜂窝网络流量卸载中基于网络容量最优化的竞争机制设计

发布时间：2018-04-19 23:24

本文选题：竞争机制 + 网络最优化　；参考：《南京大学》2017年硕士论文

【摘要】：为了应对蜂窝网络服务提供商(CSP)面临的严重的蜂窝网络流量不足的问题,将部分蜂窝网络流量通过已有的代替无线网络,例如微蜂窝和WiFi网络,是一种很有前景的方案。现有的大多数蜂窝网络流量卸载方案都假设CSP预先知道需要卸载的流量信息,从而具体分配对应的流量。然而,在实际应用场景中,伴随着移动设备流量的爆发性增长,需要卸载的蜂窝流量往往会超过区域中可卸载无线网络资源的总和。因此,在一定的预算限制下,为CSP提供最优化的流量卸载容量是非常关键的。本篇论文考虑了 CSP不依赖于具体的需要卸载流量的需求的情况,并针对性地提出了 2种高效的移动流量卸载竞争拍卖机制,CAMEO-min和CAMEO-ws。这2种机制均被证明了满足无预算赤字(non-budget-deficit),独立理性(individually rational)和诚实性(incentive-compatible),并且保证了最坏情况下CSP的容量收益不低于其在理想拍卖模型下(即CSP拥有第三方资源拥有者的真实价格与资源信息时)收益的一定比例。最后,本文进行了大量的实验,实验结果显示CAMEO-min和CAMEO-ws机制均保证了较好且稳定的CSP容量收益,特别是当全局报价主导参数(global bidder dominance)或者区域主导参数(region dominance)较大的情况下。
[Abstract]:In order to deal with the serious problem of insufficient cellular network traffic faced by cellular network service provider (CSP), it is a promising solution to use some cellular network traffic to replace the existing wireless networks, such as microcellular and WiFi networks. Most of the existing traffic unload schemes in cellular networks assume that CSP knows the traffic information to be unloaded in advance, so that the corresponding traffic is allocated concretely. However, in practical applications, with the explosive growth of mobile device traffic, the amount of cellular traffic that needs to be unloaded is often greater than the sum of unloadable wireless network resources in the region. Therefore, under certain budget constraints, it is critical to provide optimal flow unload capacity for CSP. In this paper, we consider that CSP does not depend on the specific need to unload traffic, and propose two efficient mobile traffic unload competitive auction mechanisms, CAMEO-min and CAMEO-ws. Both mechanisms have been proved to satisfy non-budget deficit, independent rational and honest incentive-compatible, and to ensure that in the worst case, CSP's capacity gains are not lower than those in the ideal auction model (that is, CSP has third-party resource ownership). The true price of the person and the resource information when) a certain proportion of the benefit. Finally, a large number of experiments are carried out in this paper. The experimental results show that both the CAMEO-min and CAMEO-ws mechanisms guarantee a good and stable CSP capacity return, especially when the global bidder dominance (global bidder dominance) or the region dominance (a region dominance) is larger.
【学位授予单位】：南京大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN929.5

【相似文献】