IL276480B2 - Synchronization in distributed communication systems - Google Patents

Description

Synchronization in Distributed Communication Systems TECHNICAL FIELD The present disclosure relates generally to the field of distributed computing. More specifically, it relates to the operation of a distributed router. GLOSSARY ASIC - Application-Specific Integrated Circuit; eNB - Evolved Node B; L2 - Layer 2; OOB - Out of band; PCI – Peripheral Component Interconnect; PTP - Precision time protocol; RAN - Radio Access Network; SDH – Synchronous Digital Hierarchy; SoC - System on a Chip; Sync-E - Synchronous Ethernet; TOD - Time of Day; NCM– Network Cluster Management; Control Plane – A logical layer that encompasses all applications related to all the functions and processes that determine which path to use, e.g. the management of the data plane. This definition encompasses among others but not limited to, configuration engines, routing stacks, routing protocols, spanning tree, ldp (Identity Provider), and user-facing services; PCIe – PCI express; the term refers to a high-speed serial expansion bus standard. It is for example the common motherboard interface for various connections such as Ethernet hardware connections. PCIe includes high maximum system bus throughput, low I/O pin count and small physical footprint, good performance scaling for bus devices, a more detailed error detection and reporting mechanism (Advanced Error Reporting, AER), native hot-swap functionality and hardware support for I/O virtualization; and White Box – a commodity, being an open or industry-standard compliant hardware for switches and/or routers within the forwarding plane. White boxes provide users with the foundational hardware elements of a network. BACKGROUND Cellular systems have always required strict synchronization. Initial implementations relied on transmission systems (such as SDH) as the clock source for the synchronization process. But as SDH networks in RAN architectures are being replaced by packetized networks, SDH systems are no longer available. On the other hand, the advances being made in radio technology and the trend of using disaggregated radio system comprising eNB units, create the need for more precise clock distribution systems that can be implemented in such disaggregated radio systems. The industry standard technologies for clock distribution are based on two main protocols: a) The Precision Time Protocol (PTP) which is a protocol used to synchronize clocks throughout a computer network. On a local area network, it achieves clock accuracy in the sub-microsecond range, making it suitable for measurement and control systems. This approach provides the option to distribute Phase and Time of day (TOD) information. The PTP selects a master source of time for an IEEE 15domain and for each network segment in the domain. Clocks determine the offset between themselves and their master. In order to accurately synchronize to their master, clocks must individually determine the network transit time of the Sync messages. The transit time is determined indirectly by measuring round-trip time from each clock to its master and the clocks initiate an exchange with their master designed to measure the transit time. b) Synchronous Ethernet (SyncE) is an ITU-T standard for computer networking that facilitates the transfer of clock signals over the Ethernet physical layer. Such a signal can then be made traceable to an external clock, and provides the option to distribute phase information via a dedicated channel on the ethernet physical layer. Yet, systems such as 5G cellular systems require synchronization precision at the order of nanoseconds (ns). In turn, to reach such a high precision level on a packetized network, clock correction must be affected at every network node belonging to the system. Synchronization packet processing in prior art solutions is done at the forwarding ASIC level. Devices referred to as "system on a chip" (SoC) devices work very well when dealing with the problem of synchronization. However, a single chip is limited in the amount of traffic it can forward, hence multi-chip systems are built. The current approach to building a multi- chip system is using a chassis cage that encompasses several line cards, where each line card includes a small number of forwarding chips, interconnected with dedicated fabric cards. As every chip needs to update the associated clock, all chips comprised in such a chassis cage device, need to have a very good synchronization level between themselves. The chassis cages are proprietary devices manufactured by using a specialized design. The intra chassis synchronization is achieved by implementing dedicated synchronization lanes, that have been designed as part of the chassis cage during its design. Disaggregated and distributed systems are systems whose components are located on different networked nodes, which communicate and coordinate their actions by forwarding messages to one another. Implementing this concept relies on using commodity hardware, such as white box network devices and commercial off the shelf servers. While the white box network devices are capable of time synchronization when operating as a stand-alone device, when they are assembled to form a distributed cluster, synchronizing the internal clock between the various cluster components becomes a significant challenge. Clusters are usually deployed to improve performance and availability over that of a single device (e.g. a computer), while typically being much more cost-effective than single devices of comparable speed or availability. However, cluster computing technique poses a number of challenges. Two of these challenges stand out: the first being application complexity and the second - cluster element synchronization. The application complexity stems from the distributed nature of cluster computing. For instance, the solution architecture must be one that is able to address the question of how can the network elements be used when the task at hand is being divided therebetween, while ensuring that from the customer application's side, it would still appear that it communicates with a single logical unit. Element synchronization on the other hand relates to the internal cohesiveness of the system. Every datum unit shared between elements must be synchronized to ensure coherence of the cluster-wide behavior. Therefore, a solution is needed, one that will ensure that: a. The cluster components are synchronized to a level needed to meet cellular networks requirements; b. The solution should rely on standard components that are available as off the shelf components by commodity hardware vendors. c. The system must appear to the outside world as a single node in term of clock mechanisms.

Claims (6)

1.CLAIMS 1. A distributed routing system for use in a communication network, wherein said distributed routing system includes at least one cluster comprising a plurality of cluster elements and characterized in that the cluster elements used for forwarding communication traffic from among the plurality of cluster elements are synchronized there-between to a single clock and then synchronized to an external communication element.

2. The distributed routing system of claim 1, wherein all the cluster elements that are used for forward communication traffic, are configured to implement IEEE 1588 standard and/or Synchronous Ethernet (Sync-E).

3. The distributed routing system of claim 2, wherein all the cluster elements that are used for forward communication traffic, are configured to be synchronized by using their out of band network as an intra cluster synchronization network.

4. The distributed routing system of claim 3, wherein in case the out of band network comprises a plurality of L2 devices, said plurality of L2 devices are synchronized there-between.

5. The distributed routing system of claim 1, further comprising a dedicated timing device associated with the cluster elements that are used for forward communication traffic, wherein said cluster elements that are used for forward communication traffic are directly connected to an out of band management network.

6. The distributed routing system of claim 1, wherein the native management ports of the cluster elements that are used for forward communication traffic are characterized in that they do not support needed features for affecting a synchronization. For the Applicant, By: