RU2598322C1 - Method of routing data packets between multiple network switching devices - Google Patents

Abstract

FIELD: information technology.

SUBSTANCE: invention relates to construction. Method of routing data packets between multiple network devices intended for switching data packets, after checking the data message a first network device compares values of priorities of the current and second network devices and for the network device with lower priority a response is blocked by means of the first network device port conjugated with said network device, value of the priority of the network device is a value, reverse the distance from the network device to the host-device, expressed in amount of intermediate network devices, located between the network device and host-device, and calculating distances values for each network device is carried out using a remote vector algorithm.

EFFECT: technical result is ensuring minimum delay when transmitting data and uniform distribution of load between network switching devices.

1 cl, 1 dwg

Description

The invention relates to the field of construction of signal switching networks (data transmission), characterized by switching networks, for example, local area networks (LAN).

The present invention is aimed at solving the problem of reducing the delay in the distribution of data packets over the network by optimizing the path of data through the network from the source (host device) to the recipient.

In existing computer networks, consisting of a set of switching devices and designed to transmit data packets, there is a problem of the appearance of redundant connections between switching devices. As a result, there is a significant increase in the data transfer route, the load on the network elements increases significantly, reliability decreases, and the delay in transmitting data packets increases.

Partially indicated problem can be overcome by constructing routes for passing data packets (configuration of connections between switching devices) close to optimal, and storing these routes in the memory of switching devices. However, with an increase in the number of connections, failures in creating new connections are possible due to the limited memory capacity of network switching devices.

Also known are ways to optimize the network structure by analyzing the time-limit for transmitting messages on ports of network switching devices and excluding communications from the connection for which the specified parameter will exceed a predetermined value (timeout exception). The main disadvantage of these methods is local optimization of the network structure and, as a result, the time it takes for data to propagate from a number of sources to a number of recipients exceeds the attainable minimum values.

This is due to the fact that the propagation time of data packets over the network is determined by the timeout value, while the lengths of their paths are not optimized. In addition, local overloads of individual network switching devices are possible.

There is a known method of packet routing and its path-oriented system (Path oriented routing system and method for packet switching networks), US Pat. No. 4,736,363 A, [1]), which does not have common features with the invention, but has a similar functional destination.

In this method, the optimal use of a network with redundant connections is provided, allowing data to be transmitted along the shortest routes and at the same time prevent the appearance of loops (“loop means spurious closure of the output of a network device to its input”). The method under consideration involves storing a route for each incoming connection, and with a large number of connections, it is possible to create new connections due to the limited memory capacity of network switching devices.

There is another method of dynamic routing according to patent US 7969915 B2 (((Technical enhancements to STP (IEEE 802.ID) implementation)), [2]), including: checking by means of the first network device designed to provide packet switching, the first message in the root port of the first network device, which includes the identifier of the current network device connected to the root port and a priority value of the current root network device; after checking the first data message by the first network device and checking the second data message in the second port of the first network device, which includes identifying and calculating the priority of the second network device; after checking the second data message, if the value of the timer (prescription) of messages of the first network device is less than the value remaining before the set timeout limit, then the response is blocked by the second port of the first network device to the second data message from the second network device, and, moreover, the first and second network devices are configured in accordance with the protocol to avoid the formation of loops, the protocol to avoid the formation of loops is the protocol connecting of the tree (Spanning Tree Protocol, STP).

The similarity of most of the features of this method is selected as a prototype.

The prototype has the following disadvantages. The prototype excludes the formation of loops, however, a significant increase in the length of the data transmission route is possible, as a result of which the load on the network elements increases significantly, reliability decreases and the delay in data transmission increases. The prototype proposed a solution to reduce the delay in reconnecting if the network topology changes, however, the length of the data message transmission route is not optimal; the time of data distribution from a number of sources to a number of recipients exceeds the achieved minimum values due to local optimization of the network structure.

These disadvantages are overcome in the proposed method for dynamically routing data packets between multiple network switching devices, which consists in the fact that in the method containing the features of the prototype: the first network switching device checks the first data message in the root port of the first network switching device, the first data message includes identifier of the current network switching device connected to the root port of the first network switching device, and prioritization of the current network switching device, after checking the first data message by the first network switching device, the specified device checks the second data message in the second port of the first network switching device, which includes the identification and calculation of the priority value of the second network switching device connected to the second port of the first network switching devices, new features are included: the first network switching device after Verifying the data message, the priority values of the current and second network switching devices are compared, and for a network switching device with a lower priority, the response is blocked by the port of the first network switching device connected to the specified network switching device, and the priority value of the network switching device is the reciprocal of the distance from network switching device to the host device, expressed as the number of precise Network switching devices located between the network switching device and a host device, and calculating the distance values for each of the network switching devices is made via remotely-vector algorithm (Distance Vector Algorithm, DVA).

Therefore, the proposed method meets the criterion of "novelty."

Comparison with other technical solutions shows that the proposed method has the features that allow to achieve global optimization of the structure of the data packet network, and thereby ensure a minimum delay in data transfer, as well as uniform load distribution between network switching devices.

The invention is illustrated by the following graphic materials:

FIG. 1 is a diagram illustrating the proposed Method for packet routing data packets between multiple network switching devices.

The method of routing data packets between multiple network switching devices (Fig. 1), including many identical ports 1, one of which is conditionally root, and host device 2 is connected to the root port of one of the network devices, is that at the first stage the first network switching device 3 checks the first data message in its own root port.

The first data message includes the identifier of the current network switching device 4 connected to the root port of the first network switching device 3, and the priority value of the current network switching device 4.

After checking the first data message by the first network switching device 3, it checks the second data message in its second port, which includes identification (addresses) and calculating the priority value of the second network switching device 5 connected to the second port of the first network switching device 3.

At the second stage, after checking the data message, the first network switching device 3 compares the priority values of the current 4 and second 5 network switching devices, as a result of which the response is blocked for the network switching device with the lowest priority by the port of the first network switching device 3 connected to it.

The priority value of the network switching device is calculated as the reciprocal of the distance from the network switching device to the host device 1.

The distance from the network switching device to the host device is numerically equal to the number of intermediate network switching devices located between this network switching device and the host device.

The calculation of distances for each of the network switching devices is performed using the well-known DVA algorithm.

The essence of the DVA algorithm described, for example, in [3], is that each network switching device periodically broadcasts over the network to other network switching devices a vector whose components are the distances from the current network switching device to other network switching devices that have connection with him. Packets of routing protocols are used, among other things, for the announcement by the current network switching device to other network switching devices of information about the network configuration known to it.

A network switching device, having adopted a distance vector from the other network switching device connected to it, to network switching devices known to it, increases the components of the distance vector by the distance from itself to another network switching device. In addition, the network switching device supplements the distance vector with information about the network switching devices connected to it, obtained directly through the interface ports or from other network switching devices. Further, the network switching device transmits the updated value of the distance vector connected to subsequent subsequent network switching devices.

As a result, each network switching device receives information through other network switching devices connected to it via ports about all the switching devices available in the composite network and about the distances to them.

Further, the network switching device selects from several alternative routes to each other network switching device the route that has the smallest distance value.

As a result, global optimization of the structure of the data packet transmission network is achieved, and due to this, a minimum delay in data transmission and a uniform load distribution between network switching devices are ensured.

Information Sources Used

1. Patent US 4736363 A dated 10/11/1985 "Path oriented routing system and method for packet switching networks)).

2. Patent US 7969915 B2 dated July 24, 2009. "Technical enhancements to STP (IEEE 802.1D) implementation.

3. http://www.protocols.ru. RIP version 2 protocol.

Claims (1)

A method of routing data packets between multiple network devices for switching data packets, each of which network devices includes many identical ports, one of which is conditionally the root, a host device is connected to the root port of one of the network devices, including: checking first by the network device of the first data message in the root port of the first network device, the first data message includes the identifier of the current network device connected to the root port of the first network device, and the priority value of the current network device, after checking the first data message by the first network device, the specified device checks the second data message in the second port of the first network device, which includes the identification and calculation of the priority value of the second network device connected to the second port of the first network device, characterized in that the first network device after checking the data message is made with setting the priority values of the current and second network devices and for a network device with a lower priority, the response is blocked by the port of the first network device connected to the specified network device, and the priority value of the network device is the reciprocal of the distance from the network device to the host device, expressed in the number of intermediate network devices located between the network device and the host device, and calculating the distance values for each of the network devices Properties are performed using the Distance Vector Algorithm (DVA).

Images