标题: | 多约束赢者全拿类神经网路及其平行分级排程应用 Multiple-Constraint Winner-Take-All Neural Networks and Their Applications in Parallel Prioritized Scheduling |
作者: | 柯柏宇 Ke, Bo-Yu 田伯隆 Tien, Po-Lung 电信工程研究所 |
关键字: | 赢者全拿;类神经网路;平行分级排程;Winner-Take-All;neural network;parallel prioritized scheduling |
公开日期: | 2013 |
摘要: | 赢者全拿(WTA)及k-赢者全拿(k-WTA)类神经网路被广泛应用于解决各种基于竞争的问题,由于其高效率的平行计算特性,赢者全拿类神经网路特别适合用于即时应用。然而,现有的赢者全拿模型皆仅将所有的神经元置于单一共同的竞争群体里面,这严重地限制了其应用范围,于是在此论文中我们提出了一类新的名为k-多限制赢者全拿(k-MCWTA)的类神经网路模型,其考虑神经元分布于多个互相影响的竞争群体中。 我们提出的第一个模型称为排名Hopfield类神经网路(RHNN),其为着名的Hopfield类神经网路(HNN)的延伸,专门用来解决1-MCWTA问题。RHNN是由排名神经元所组成,其允许较高排名的神经元可以停用在之前迭代中已启用的较低排名的神经元;但是将神经元排名无可避免地会导致无法收敛的问题,于是我们提出了两个定理用以提供RHNN收敛到最佳解的充分条件。然而,另一个问题是RHNN类似于HNN及其他基于能量函数的模型,若以同步的方式更新神经元时,将会产生状态震荡而无法收敛;于是,在非同步更新神经元的限制下,若RHNN要实现虚拟的平行运算则必须采用连续时间模型,因此必须以类比电路来实现;然而,如今的系统开发平台大多适用于数位设备,故非同步更新神经元的要求变得不切实际。 我们于是提出了另外一个名为离散时间同步排名类神经网路(DSRN)的1-MCWTA模型。DSRN同样是由排名神经元所组成,但其能够在完全平行(即,同步)离散时间的方式下运作,因此可以在数位系统中实现。我们透过一个定理证明DSRN会收敛到最佳解,此定理同时提供了收敛延迟的理论上限。最后,为了应付更一般的问题,即k-MCWTA问题,我们引入了一个迭代平行分群演算法(IPGA),IPGA可以建模为一个离散时间二进制值的双层递回式类神经网路,其中每一个神经元皆可以在完全平行的方式下运作。我们提供了一个定理,用以给出收敛时间的理论上限: ,其中 是神经元的总数;然而在此论文中将会展示在大多数实际情况中,收敛时间会比理论值更为减少许多。 为了展示所提出模型的优越性,我们运用我们上述所提出的类神经网路模型来解决两个具体的排程问题:用于WDM交换系统的优先性封包排程(1-MCWTA问题)及用于资料中心网路的优先性流量调度(k-MCWTA问题)。模拟结果显示在一个系统时间槽内的计算时间中,RHNN排程器可以达到接近100%的吞吐量及多层次优先性排程;此外,通过基于CUDA的模拟,DSRN排程器以大约 的收敛延迟实现接近最佳的吞吐量和优先性排程,其中 是交换机埠数;模拟结果还表明IPGA调度器在接近不变的收敛时间下,实现QoS差异化及高达30%的数据中心网路节能效果。 Winner-Take-All (WTA) and k-Winners-Take-All (k-WTA) neural networks have been widely used to solve various competition-based problems. Thanks to highly efficient paral-lel-computation nature, the class of WTA neural networks is especially suitable for real-time practices. However, the existing WTA models simply consider all neurons are within a common competition group, which severely limits the applications of WTAs. In this thesis, we propose a new class of WTA neural networks, called k-multiple-constraints-winners-take-all (k-MCWTA) models, which take consideration of neurons contained in multiple and joint competition groups. The first model, called Ranked Hopfield Neural Networks (RHNN), is an extension from well-known Hopfield neural network (HNN) and is dedicated for resolving 1-MCWTA problems. Structured with ranked neurons, the RHNN allows higher-rank neurons to disable lower-rank neurons that have been enabled during previous iterations. Ranking the neurons unfortunately gives rise to a convergence problem. We present two theorems that give the sufficient conditions for the RHNN to converge to the optimal solution. This RHNN, howev-er, is similar to HNN and other energy-function-based methods, which give rise to a state os-cillation problem when the neuron updates are in a synchronous manner. Under the con-straint of asynchronous neuron update, for the RHNN to accomplish virtually parallel opera-tion, the neural model must be in continuous-time and hence be implemented in an analogi-cal circuit. However, as the system-design platforms are widely available with digital devic-es, the requirement of asynchronous neuron update becomes impractical. We thus propose another 1-MCWTA model, called Discrete-time Synchronous Ranked Neural-network (DSRN). DSRN is also structured with ranked neurons, but is capable of op-erating in a fully parallel (i.e., synchronous) discrete-time manner, and thus can be imple-mented in digital systems. We delineate via a theorem that DSRN will converge to the opti-mal solution. The theorem also provides a theoretical upper bound of the convergence la-tency. To cope with more general problem, i.e., k-MCWTA problem, we then introduce an Iterative Parallel Grouping Algorithm (IPGA). IPGA can be modeled as a discrete-time bi-nary-value two-layer recurrent neural network, in which each neuron is capable of operating in a fully parallel manner. We provide a theorem that gives a theoretical upper bound of the convergence time, , where is the total number of neurons. As will be shown in the thesis, the convergence time can be largely reduced in most real cases. To demonstrate the superiority of proposed models, we apply our methods to resolve two specific scheduling problems: prioritized packet scheduling (1-MCWTA problem) for WDM switching system and prioritized flow scheduling (k-MCWTA problem) for data cen-ter networks. Simulation results show that, with the computation time within one system slot time, the RHNN scheduler achieves near 100% throughput and multi-level prioritized sched-uling. Moreover, via CUDA-based simulations, the DSRN scheduler achieves near-optimal throughput and prioritized scheduling, with nearly convergence latency, where is the switch port count. The simulation results also show that the IPGA scheduler achieve QoS differentiation and up to 30% energy saving for data center network, with near-ly constant convergence time. |
URI: | http://140.113.39.130/cdrfb3/record/nctu/#GT079813802 http://hdl.handle.net/11536/74222 |
显示于类别: | Thesis |