System aspects of the next-generation data-center networks based on 200G per lambda IMDD links

Author(s):  
Jinlong Wei ◽  
Stefano Calabrò ◽  
Talha Rahman ◽  
Nebojsa Stojanovic
Author(s):  
Jirasak Ponchua ◽  
Suchada Sitjongsataporn

The increasing demands within and between the data centers used for data traffic has required. Efficient links are important to data center applications for supporting the unlimited demand. Transmission capacity of single-mode fiber (SMF) is limited by fiber nonlinearity which prevents the increasing transmission power and finite amplifier bandwidth. Single-mode multi-core fibers (SM-MCFs) that are expected to overcome the current limitation of optical communication capacity. However, the inter-core crosstalk still has an effect on SM-MCF, which can limit the transmission of the inter-data center. In this paper, the design of four-core uncoupled multicore fiber is discussed for next generation inter-data center networks in order to support the unlimited use of data traffic in the future. The objective of this paper is to determine the appropriate range of core radius and core pitch, which are taken into consideration to reduce the inter-core crosstalk inside the optical fiber. These parameters can be able to improve various constraints to achieve the best multi-core fibers design. From the simulation concerned with the inter-core crosstalk, the experiment results show that the range of core pitch is at 47.5 μm to 50 μm and the range of core radius starts from 4.5 μm to 5.5 μm, that can achieve with crosstalk lower than – 30 dB/100 km for the future inter-data center networks.


2016 ◽  
Vol E99.B (11) ◽  
pp. 2361-2372 ◽  
Author(s):  
Chang RUAN ◽  
Jianxin WANG ◽  
Jiawei HUANG ◽  
Wanchun JIANG

Author(s):  
Jiawei Huang ◽  
Shiqi Wang ◽  
Shuping Li ◽  
Shaojun Zou ◽  
Jinbin Hu ◽  
...  

AbstractModern data center networks typically adopt multi-rooted tree topologies such leaf-spine and fat-tree to provide high bisection bandwidth. Load balancing is critical to achieve low latency and high throughput. Although the per-packet schemes such as Random Packet Spraying (RPS) can achieve high network utilization and near-optimal tail latency in symmetric topologies, they are prone to cause significant packet reordering and degrade the network performance. Moreover, some coding-based schemes are proposed to alleviate the problem of packet reordering and loss. Unfortunately, these schemes ignore the traffic characteristics of data center network and cannot achieve good network performance. In this paper, we propose a Heterogeneous Traffic-aware Partition Coding named HTPC to eliminate the impact of packet reordering and improve the performance of short and long flows. HTPC smoothly adjusts the number of redundant packets based on the multi-path congestion information and the traffic characteristics so that the tailing probability of short flows and the timeout probability of long flows can be reduced. Through a series of large-scale NS2 simulations, we demonstrate that HTPC reduces average flow completion time by up to 60% compared with the state-of-the-art mechanisms.


Sign in / Sign up

Export Citation Format

Share Document