No abstract

Summary With the development of cloud computing and big data, the internal communication business in data center has increased dramatically, and then the traffic in data center has also significantly increased. The bandwidth of data center is difficult to meet the bandwidth requirements of those intensive applications, and data center is facing a risk of network congestion. Under the background of the development of network intelligence, software‐defined network (SDN) should demonstrate its intelligence as a future network architecture. In this paper, we introduce reinforcement learning into the SDN data center to implement congestion control based on flow. We improve the Q‐learning and Sarsa algorithms and propose two methods of congestion control based on the algorithms. Test results show that these two congestion control methods can control congestion effectively. And Sarsa method has a better performance of link utilization. The average link utilization of the Sarsa method is 2.4% higher than the Q‐learning method and is 4.48% higher than the on‐demand method.

A 1K-bit phase change random access memory (PCRAM) with improved periphery circuits for better reliable operations has been successfully developed in 130 nm CMOS technology. A flexible write driver is proposed to provide a novel continuous step-down pulses by studying programming strategies while a reliable read circuit is designed by investigating the special transition characteristics of PCRAM, leading to an effective write operation and a non-destructive read operation without any additional changes of the storage states. In addition, a large sense margin has been achieved and the read results corresponding well with the write operations, which demonstrate the influences of technology variations have been considerably decreased with the proposed periphery circuits.

No abstract

Kelvin–Helmholtz instability (KHI) is considered important in transporting energy and mass at the magnetopause of Earth and other planets. However, the ion kinetic effect influences the generation and evolution of KHI, as the spatial length of the magnetopause may be smaller than the Larmor radius of the ion; this influence is not yet fully understood. In this investigation, laboratory experiments were designed to study the excitation of KHI at the ion kinetic scale. The ion kinetic scale was modeled by controlling the ratio of the Larmor radius and the electric scale length ρ i / L E > 1, and the KHI was excited at the spatial scale of LE by a controllable sheared E × B flow. It was found that the ion kinetic effect on KHI growth manifests as the ion Larmor radius reaches the shear length scale, and the KHI is suppressed as the ion Larmor radius increases. Incorporating a theoretical analysis by substituting our experimental parameters, the suppression of the KHI was attributed to the fact that the KHI linear growth rate decreases with the ratio change of the ion Larmor radius because the relative orientations of the ion diamagnetic drift velocity ( V d) and the shear flow velocity ( V 0) are opposite. Our experimental conditions ( V d / V 0 < 0) are similar to the dusk-side conditions of the magnetospheres of Earth and Mercury under northward interplanetary magnetic fields; therefore, this result can be extended to understand the evolution of KHI in the planetary boundary layer.