We consider a general multi-connectivity framework, intended for ultra-reliable low-latency communications (URLLC) services, and propose a novel, preallocation-based combinatorial auction approach for the efficient allocation of channels. We compare the performance of the proposed method with several other state-of-the-art and alternative channel-allocation algorithms. The two proposed performance metrics are the capacity-based and the utility-based context. In the first case, every unit of additional capacity is regarded as beneficial for any tenant, independent of the already allocated quantity, and the main measure is the total throughput of the system. In the second case, we assume a minimal and maximal required capacity value for each tenant, and consider the implied utility values accordingly. In addition to the total system performance, we also analyze fairness and computational requirements in both contexts. We conclude that at the cost of higher but still plausible computational time, the fairness-enhanced version of the proposed preallocation based combinatorial auction algorithm outperforms every other considered method when one considers total system performance and fairness simultaneously, and performs especially well in the utility context. Therefore, the proposed algorithm may be regarded as candidate scheme for URLLC channel allocation problems, where minimal and maximal capacity requirements have to be considered.
翻译:我们考虑一个通用的多连通性框架,旨在为极端可靠的低时空通信服务(URLLC)设计一个通用的多连通性框架,并为高效分配渠道提出一个新的、基于预先分配的组合拍卖办法;我们将拟议方法的性能与其他几种最先进的和替代的频道分配算法进行比较;两种拟议的性能衡量标准都是基于能力和基于实用的计算环境;在第一个情况下,每个新增能力单位被视为对任何租户都有利,独立于已经分配的数量,而主要措施是系统的总吞吐量;在第二个情况下,我们为每个租户设想一个最低和最高所需的能力价值,并相应地考虑隐含的效用价值;除了整个系统性能之外,我们还分析这两种情况下的公平和计算要求;我们的结论是,以更高但仍然有说服力的计算时间为代价,拟议的基于配置前的组合拍卖算法的公平性版本,在考虑系统总体业绩和公平性时,超越了所有其他考虑的方法;在使用能力方面,我们设想的最低限度的频道和最大能力分配办法,因此,在认为最低水平的磁带上,拟议的磁带是最低的磁带。