The present invention relates to data communication technologies and in particular to a method and system for data transmission.
With the development of communication technologies, transmission networks have evolved quickly from the era of Synchronous Digital Hierarchy (SDH) to the era of packet-based technologies. For packet-based technologies, a forwarding table is typically maintained at each network element, e.g. Ethernet, Multiple Protocol Label Switching (MPLS). A forwarding table for Ethernet is indexed by a Destination Address (DA) or by DA+VLAN-ID (Virtual Local Area Network Identification) in the case that a VLAN ID is present. Due to the discreteness of Media Access Control Layer (MAC) addresses, a hash algorithm is required to create and search the forwarding table. Whether the hash algorithm is superior or inferior depends upon selecting of a hash formula and adjusting of the size of the forwarding table. The algorithm is of a constant complexity in most cases, and of a relatively high complexity in the case of a failure in hitting, and may search throughout the forwarding table in a worst case. The label space of MPLS is 1M, and practically, not all labels will be used for a network element, so that in a system with limited memory, the hash algorithm can be equally used to create and search the forwarding table. The size of a hash table determines the efficiency of the hash algorithm, but adds to the overhead of the system. Such a mechanism has been used so far. Centralized data switching based upon a forwarding table for packets has direct influence upon the efficiency of data forwarding in a communication network and the complexity of system processing, which limits increasing of the switching capacity of a network element and raising of the throughput of a port.
In view of the above, an important approach for improving the efficiency of packet switching is to seek an appropriate method which can reduce the complexity of a searching algorithm for data packet forwarding to a constant value while not occupying too many system resources.
At present, there has been provided in the prior art a method and device for direct switching of frames containing destination information, as disclosed in publication No. CN 20050027881, which presents the following disclosure:
However, the inventors have studied this solution and found that there are still the following drawbacks:
Embodiments of the present invention provide a method and system for data transmission to achieve fast switching of data frames in a large network.
An embodiment of the invention provides a method for data transmission including:
adding a Routing Information Header (RIH) in each transmitted data packet, the RIH containing routing information of the data packet in a network;
reading, by a network element, routing information required for a present network node from the RIH upon receipt of a data packet with the RIH; and
transmitting the data packet in accordance with the read routing information.
Another embodiment of the invention provides a system for data transmission including an existing network and a data relaying network device, wherein a data encapsulation protocol supporting an RIH, for the data relaying network device, is of a service layer of an original data encapsulation protocol in the existing network, and the both protocols constitute a client/server relationship; at an outer layer of an original data packet, the RIH is encapsulated as being a forwarding criterion for each node along a link; and an original encapsulation format, including various signaling and protocols, is transmitted transparently through the data relaying network device supporting the RIH.
Still another embodiment of the invention provides a system for data transmission including an existing network and a data relaying network device, wherein a data encapsulation protocol supporting an RIH, for the data relaying network device, can be of the same layer as an original data encapsulation protocol, and both protocols constitute a peer-to-peer relationship. The RIH is inserted into an original data packet encapsulation. An original encapsulation format, including various signaling and protocols, is terminated on the data relaying network device at a border of a network; and intercommunication between networks is enabled through a conversion between two mechanisms under a control of management-plane or end-to-end upper-layer protocols or signaling.
The data relaying device constitutes by itself a management domain in which topology discovery and routing selection protocols and various control-signaling supporting the Routing Information Table (RIT) are operated.
Obviously from the embodiments of the invention, in a data transmission network, the number of nodes through which a service passes will be limited under a strict management and control. For a typical telecommunication network, the service goes through an access node at an entrance to the network of a service provider, then goes through a service convergence point, and reaches a backbone point at which the service is subject to remote transmission or local service switching. Then, the service goes to another end of the service in a downlink in an opposite way. Due to this procedure, the RIT can be 10 table items in size, and the RIH can be of 8˜24 bytes. Such an overhead is totally acceptable in comparison with an Ethernet data packet allowance for a maximum size of 1.5K or an IP/MPLS packet allowance for a larger size. If a network with such an architecture including routing information is only a partial network through which an end-to-end service goes, the RIH may gain a higher relative bandwidth efficiency, and the fast switching for data transmission is achieved.
The invention will become more readily apparent from the Detailed Description of the Invention, which proceeds with reference to the drawings, in which:
Embodiments of the present invention present a new architecture and an implementation mechanism thereof in a communication network for data transmission, adding in each transmitted data packet a Routing Information Header (RIH) containing routing information of the data packet in the network, so that fast switching for data transmission can be achieved.
The invention is further described as follows with reference to the drawings and the embodiments. Referring to
In step S11, an RIH is added in each transmitted data packet, wherein the RIH contains routing information of the data packet in a network.
In step S12, upon receipt of a data packet with the RIH, a network element reads routing information required for the present network node from the RIH.
In step S13, the data packet is transmitted in accordance with the read routing information.
In other words, the invention as shown with reference to
Further referring to
Referring to
Upon receipt of a data packet which could be an Ethernet frame, an MPLS packet, an Asynchronous Transfer Mode (ATM) information element, or a frame-relayed segment, etc. with the RIH, the network element reads information on an egress port, which can, for example, be determined in accordance with a protocol or signaling, or be configured manually, that is required for the present network node from the RIH. The data packet is then transmitted through the read egress port. In this procedure, a series of processing steps are performed on the RIH, including a Frame Check Sequence (FCS), where after updating the RIT, the network element regenerates and puts a BIP8 of the RIH into the check field. Then, it is determined in accordance with a system requirement whether to check the entire data packet with respect to the FCS or the like. The FCS field is updated upon completion of the check.
Finally, the data packet is forwarded to the egress port of the present network element in accordance with a read egress port number.
Referring to
The network element of
For a normal data packet with an RIH, the network element reads the information on the egress port of the present network node from the RIT of the RIH, and transmits the data packet. Also, the HC field of the RIH is decremented, and the read information of the present network node is removed from the RIH. Here, after the information on the egress port of the present network node has been read, the data packet is transmitted to the egress port. Alternatively, the data packet can be transmitted to the egress port after a check of the entire data packet.
Further referring to
The RIH is structured with a variable length, which is an integer multiple of 4 bytes. Upon an initial configuration and when a table item of the RIH is removed for each hop, full-zero bytes can be filled at the front of the table item for a complement, where the number of the full-zero bytes to be filled can be calculated from a length of the removed table item and the number of bytes filled initially.
In a data transmission network, the number of nodes through which a service goes will be limited under a strict management and control. For a typical telecommunication network, the service goes through an access node at an entrance to the network of a service provider, then goes through a service convergence point, and reaches a backbone point at which the service is subject to remote transmission or local service switching. Then, the service goes to another end of the service in a downlink in an opposite way. Due to this procedure, the RIT can be 10 table items in size, and the number of bytes occupied by the RIH can be 8˜24 bytes. Such overhead is totally acceptable in comparison with an Ethernet data packet allowance for a maximum size of 1.5K or an IP/MPLS packet allowance for a larger size. In a practical application, if a network with an architecture including routing information according to the invention is only a partial network through which an end-to-end service goes, then the RIH can gain a higher relative bandwidth efficiency.
A device provided with the above functions can be called a Data Repeater (DR), which can identify a service in accordance with an original service identifier, and establish a protective relationship with a backup service.
Further referring to
According to the third method, a network topology is reported to a control plane through a topology discovery protocol operating over a network, a routing protocol is operated over the control plane based on the network topology, and information of the egress ports across the connection in the respective network nodes is obtained, thus obtaining and issuing an RIH to a data plane; or within the control plane, a discovery of a topology of the entire network is performed through a topology discovery protocol, a routing analysis of an accessing service is performed in accordance with a routing protocol, an RIT and further an RIH are generated with respect to the discovered network topology, and the RIH is added into data packets of the accessing service for transmission.
For a network level of multi-layer convergence, an uplink port can be appointed by a network operator, and thus the implementation of the topology discovery and the routing selection protocol can be simplified.
Due to the special mechanism requirement of the transmission system according to the embodiments of the invention, existing networks can not support such processing of the mechanism, and thus devices supporting the RIH architecture need to be networked separately. Because the RIH brings an incremental change to original data packets, a network constituting of such devices can be compatible with the existing networks.
Further referring to
Further referring to
The above descriptions are merely illustrative of the preferred embodiments of the invention, and are not intended to limit the invention, a scope of which shall be defined only by the accompanying claims. Any modification, improvement or equivalent substitution made without departing from the spirit and scope of the invention shall be encompassed in the claimed scope of the invention.
Number | Date | Country | Kind |
---|---|---|---|
200610004080.3 | Jan 2006 | CN | national |
This application is a Continuation of PCT/CN2007/000297 filed Jan. 26, 2007, and claims the benefit of Chinese Patent Application No. 200610004080.3 filed on Jan. 27, 2006, both of which are incorporated by reference in their entireties herein. The PCT application published in Chinese as WO2007/085198 A1.
Number | Date | Country | |
---|---|---|---|
Parent | PCT/CN2007/000297 | Jan 2007 | US |
Child | 12042485 | US |