基于无速率编码的无线并发传输调度机制研究

Rateless coding-based concurrent medium access allows multiple nodes to transmit simultaneously, which utilizes wireless spectrum resources more effectively and hence significantly boosts the capacity of wireless networks. However, current works on scheduling mechanism of concurrent medium access are limited to feasibility test and small scale validation experiments, and so far there is rare effort on investigating the two key problems of concurrent scheduling. The first problem is the degree of concurrence, that is, the number of nodes that transmit simultaneously. The second problem is to decide the exact set of nodes to be involved in a concurrent medium access. Moreover, current works do not take into account requirements of upper layer applications. Given this situation, in this proposal we first plan to study the degree of concurrence for rateless coding-based concurrent medium access, specifically, we want to derive a model to quantitatively understand the tradeoff between the number of concurrent nodes and the error of concurrent decoding. Through this model, we can compute the optimal degree of concurrence under any given channel condition. Secondly, we plan to formulate a utility optimization model that integrates correlation among channel coefficients of concurrent nodes, and then we calculate the exact set of concurrent nodes that achieves the maximum network utility. At last, based on the two works above, we plan to take into account delay demand of upper layer real-time service, and propose a novel concurrent transmission scheduling method.The method will provide a framework to carry controllable adjustment to the resulting delay performance in a per-flow level so as to satisfy delay demand of real-time service. Meanwhile, the method will guarantee the maximum network utility unchanged.

基于无速率编码的并发传输允许多个节点同时传输数据，可更有效的利用无线频谱资源，提高无线网络容量。然而，当前基于无速率编码的并发传输的调度机制研究仅局限于对并发传输可能性的探讨和小规模的并发调度验证性试验，对并发调度的核心问题并发度与并发节点组合选择缺乏系统深入的研究，也没有考虑到上层应用的延迟需求。鉴于这种情况，本课题拟首先开展基于无速率编码的并发传输的并发度研究，建立并发传输节点数量与解码误差间的制衡关系模型，以计算给定网络条件下的最优并发度。其次，研究建立引入信道相关系数的并发传输网络效用优化模型，并根据该模型计算能够达到最大网络效用的并发节点组合。进一步地，本课题将考虑上层实时业务的需求，提出面向实时业务延迟需求的创新的并发调度方法，在保持网络效用最大化的同时，对实时业务流的延迟进行可控性调整，实现实时业务的延迟保障。

当前基于无速率编码的并发传输调度机制研究仅局限于对并发传输可能性的探讨和小规模的并发调度验证性试验，对并发调度的核心问题并发度与并发节点组合选择缺乏系统深入的研究，也没有考虑到其在未来新型网络环境下面临的困难和挑战。鉴于这种情况，本课题从并发传输的并发度模型、并发节点的选择和调度、新型网络中的无速率编码等三方面展开深入研究：1）建立了能够根据网络环境计算最优并发接点数的并发度模型，为提出新的并发传输调度方法奠定了基础；2）提出了基于双边异构移动群智感知市场的用户协同激励机制以及可扩展的、分布式的MU-MIMO网络并发用户选择算法，在保证用户公平性前提下充分利用并发传输的潜在性能增益；3）深入研究了无速率编码在无人机网络等未来新型网络中面临的挑战，提出能够消除错误码元干扰的无速率编码方法，大幅度提升信息传输效率。本课题共发表国内外论文19篇，申请发明专利4项，制订中国通信行业标准2项，超额完成了预期目标。