Asset Publisher

Time-sensitive networks in redundant low latency communications (H + T Net). TEC2017-84011-R

Period:
from 2018 to 2020
Financing entity:
Ministry of Economíy, Industry and Competitiveness
Description:

Information technologies have also broken into industrial areas dominated, until recently, by scarce, specialized and disconnected communication systems, with their usual disadvantages: slow, isolated, etc. In contrast, current networks and technologies bring improvements: on the one hand, using common and standardized equipment reduces costs; on the other hand, the latest information technologies allow greater speed and more services. However, these advantages are accompanied by several drawbacks: patented and specialized networks resolved requirements such as low latency, redundancy and synchronism in a satisfactory manner; On the other hand, general networks are unable to cope with them. In this context, it has been necessary to adapt and complete the traditional channels, mainly the binomial Ethernet + TCP / IP, to the demands of the smart distribution networks (smart grids) or the industry. Two main strategies guide this process of adaptation: one sponsored by the IEC and the other by the IEEE.

The IEC has opted to modify, minimally, layer 2 of the OSI (Open Systems Interconnection) standard, in order to offer redundancy and low latency. This process has led to the HSR (High-availability Seamless Redundancy) protocol, aimed at networks in electrical substations and, secondarily, industrial networks. On the other hand, the IEEE has decided to evolve from the audio and video transport networks, providing them with advanced mechanisms in layers 2 and 3 of OSI, called time sensitive networks or TSN (Time Sensitive Networking). These changes are deeper and seek to modify all existing Ethernet networks, providing them with different levels of service quality - constant latency, high availability, determinism and synchronization - in order to allow them to coexist from a domestic network to a deterministic industrial network.

These two currents seem doomed to understand each other in the short term, and it is in this apparent confrontation where the present project finds its place. We intend, based on the redundant technologies sponsored by the IEC, and used in the most demanding environments in terms of reliability, to incorporate promising features of the TSN networks, promoted by the IEEE. The first -IEC- provide protection against failures, synchronism and can minimize latency. The second ones -IEEE- guarantee the maximum availability of bandwidth, determinism and mechanisms to resynchronize the system and stabilize the latency. By integrating both technologies, all these advantages will converge in a single polyvalent industrial network.

In addition to this objective, we aspire to advance in the physical layer towards a higher speed, since, although it is not currently required, as the price of 10 Gbps (10G) networks decreases, the demand for industrial technology at such speed. For this reason, we want to incorporate some of the possibilities that are generated in the project to 10G networks. It should be mentioned that it is not only a change in the physical layer, but that the substantial increase in speed will lead to important challenges in the area of information search in memory and data exchange between different ports.