RU2647697C1 - Method for distribution of resources between agents in the heterogeneous episodic computer network - Google Patents

Abstract

FIELD: information technology.

SUBSTANCE: invention refers to the resources distribution means in the computer network. Method for allocating resources between agents in a heterogeneous, episodic computer network comprises the steps of providing, by means of inter-machine information exchange, for interaction between the computing devices, the secure data store is created at each of the mentioned computing devices; the guaranteed storage of incoming messages in the secure data store is ensured; the guaranteed processing of the message body is ensured; agents are connected to the heterogeneous episodic computer network, these agents represent the computing devices and provide for interaction between the computing devices; after creating the decentralized database, which stores information about the current state of resources, the agents form and submit applications that indicate which resource the agents are ready to offer and at what cost. Message routing table is stored in the decentralized database.

EFFECT: technical result is the reduction of computing and information costs for the distribution of heterogeneous resources in the heterogeneous computer network through the creation of the decentralized database.

1 cl, 2 dwg

Description

The invention relates to computer technology and can be used to create and improve heterogeneous episodic computer networks.

A heterogeneous computer network is understood to mean a network connecting personal computers and other devices with various operating systems or data transfer protocols (E. Tanenbaum, D. Weatherlo, “Computer Networks”, 5th edition).

A data transfer protocol is understood to mean a set of rules and a sequence of actions that allows connecting or exchanging data between two or more devices connected to the network (E. Tanenbaum, D. Weatherlo, “Computer Networks”, 5th edition).

The term episodic network is understood to mean peer-to-peer self-configuring networks, which are based on router nodes, differ in that the topology of such networks is changing. Each device in such a network moves independently of any other in absolutely any direction (Filippov AN “Properties and characteristics of Ad Nose networks.” The international scientific journal “Young Scientist”. No. 11 (115) June - 1 2016).

Agents are various electronic computing facilities, including stationary, GRID, mobile devices, etc., having various communication interfaces. The agent determines which resources or resource packages it needs.

As resources, computational (processor time), semantic (necessary data files), network (Internet access), and arbitrary services (information about installed applications) can be used.

Computing costs are understood as processor time (the time spent by the computer processor to process the task), the amount of RAM (the volatile part of the computer memory system, in which the computer executable code (programs) is stored during computer operation, as well as input, output and intermediate data, processed by the processor), the amount of memory occupied on the hard drive, network bandwidth (metric characteristic, showing the ratio of the limit of the number of passing units (inform units, objects, volume) per unit time through a channel, system, unit) [GOST 28195 89].

Information costs are understood as the totality of data organized to effectively obtain reliable information [GOST 7.0 99].

A known method, including selecting a routing path for a data packet from a source node to a destination node in a mobile episodic wireless local area network (United States Patent No. 9036509, H04L 12/16, 05/19/2015), which stores a data packet if the selected path is identified as a dead end , and establishes a connection with the source node. This method also includes transmitting a data packet to the source node if the first distance between the source node and the destination node is less than the second distance between the source node and the destination node. More specific associations include sending a request to obtain information about the location of the destination node, the available paths to it, and also determines the optimal of these paths.

The disadvantage of this analogue is the high computational and informational costs of resource allocation in a heterogeneous episodic computer network.

There is a method of adapting the parameters of resource allocation to achieve one or more quality indicators with improved accuracy (patent for the invention of the Russian Federation No. 2368104, H04W 28/18, 09/20/2009), which represents new dimensions of information based on the so-called mutual information (VI), preferably at the block level. Information measurements based on the VI of the previous transmission, prediction of the subsequent transmission channel and one or more quality requirements are used to determine the quantity and type of resources, for example, time, frequency and power resources, which should be used for subsequent transmission. The resource allocation may comprise, for example, power distribution and / or link adaptation. A preferred embodiment is joint link adaptation and power distribution.

The disadvantage of this analogue is also the high computational and informational costs of resource allocation in a heterogeneous episodic computer network.

The closest in technical essence and functions performed analogue (prototype) to the claimed method is a method of guaranteed bringing information in a heterogeneous computer network (patent for the invention of the Russian Federation No. 2361366, H04L 12/00, 07/10/2009). In this method, system-wide software is installed on each computing installation to obtain a single secure backbone of information exchange between all computing installations of the aforementioned heterogeneous computer network; provide on each computing installation unified unified means of inter-machine information exchange, realizing the physical connectivity of the mentioned computing installations within the framework of the aforementioned unified secure highway of information exchange; by means of said means of inter-machine information exchange, uniform interaction exists both between computing facilities in a local computer network of an automation object included in said heterogeneous computer network and between global automation objects within the framework of said heterogeneous computer network; reconfigure in each operating system all system interfaces for exchanging data with the corresponding data exchange mechanisms of the aforementioned system-wide software; using the said means of inter-machine information exchange, provide interaction between the end-applications of the computing facilities within the framework of the aforementioned heterogeneous computing network by automatically creating a queue of outgoing and incoming messages for each of the said end-applications; in accordance with a predetermined procedure, the body of each message to be exchanged along the aforementioned single secure information exchange trunk is randomized to protect the contents of this message from unauthorized access; encapsulate, in accordance with a predetermined procedure, a randomized body of each outgoing message to be exchanged along the aforementioned unified secure information exchange backbone to ensure global routing within the framework of the aforementioned heterogeneous computer network and process the specified message on the receiving node; provide guaranteed message delivery within the aforementioned heterogeneous computer network by returning to the sending node a receipt for the delivery of the outgoing message to the receiving node; create on each of the mentioned computing installation a secure data warehouse, which is randomized according to the same principle as the message body; storing queues of outgoing and incoming messages in said secure data warehouse of each said computing installation both to the address of said computing installation or a software application running on it and to another computing installation of said inhomogeneous computing network; provide guaranteed storage of incoming and transit messages in said secure data warehouse until the completion of the guaranteed delivery process; storing in said secure data warehouse of each said computing installation a message routing table within the said heterogeneous computing network, as well as the identifier of each final software application operating in the context of said system-wide software; provide guaranteed processing of the message body by indicating the number of the final application handler with which the message body should be processed, and upon receipt of this processing, a processing confirmation is sent to the sending node; systematically check the credentials of each official authorized to work with the information resources of the said heterogeneous computer network, and stop the session of a particular official if at least one parameter from the corresponding access profile stored in the secure data warehouse of the said computing installation does not match; provide guaranteed identification of outgoing, and (or) incoming, or transit messages through the use of a built-in mechanism for identifying officials in a computer network based on an electronic digital signature, thereby achieving the legal significance of information exchange.

The disadvantage of the prototype is that when using this method in a heterogeneous episodic computer network, there are high computational and information costs for resource allocation.

The objective of the invention is to provide a method for distributing resources between agents in a heterogeneous episodic computer network, which allows to reduce the computational and information costs of distributing heterogeneous resources in a heterogeneous computer network by creating a decentralized database of the current state of resources.

This problem is solved by the fact that in the method of distributing resources between agents in a heterogeneous episodic computer network, they provide interaction between computing facilities by means of inter-machine information exchange, create a secure data warehouse on each mentioned computing installation; provide guaranteed storage of incoming messages in a secure data warehouse; provide guaranteed processing of the message body. Additionally, they carry out the following actions: agents representing various electronic computers, including stationary, GRID, mobile devices, etc., are connected to a heterogeneous computer network. When taking into account the cost of resource allocation and heterogeneity of agent connections in a heterogeneous episodic computer network, a double combinatorial auction of resource redistribution in an episodic network is used. After the interaction between the computing facilities is established, a decentralized database is created, which is updated automatically, agents form and submit applications stating which agents are ready to offer the resource and at what cost and what resource the agents can acquire and at what cost. They provide guaranteed storage of incoming messages, save a message routing table, and agent identifiers in a decentralized database, while the incoming message is an application for participation in the auction, provides guaranteed processing of the message body, compares message contents and exchanges resources between agents.

Thus, by creating a secure data warehouse with automatic updating that stores information about applications and the current state of resources, as well as the use of a double combinatorial auction mechanism, the computational and information costs of resource allocation during the interaction of heterogeneous agents and heterogeneous resources in heterogeneous computer networks are reduced.

The analysis of the prior art made it possible to establish that there are no analogues that are characterized by a combination of features that are identical to all the features of the claimed method of distributing resources between agents in a heterogeneous episodic computer network. Therefore, the claimed invention meets the condition of patentability "novelty."

Search results for known solutions in this and related fields of technology in order to identify features that match the distinctive features of the prototype of the claimed method showed that they do not follow explicitly from the prior art. The prior art also did not reveal the popularity of the impact provided by the essential features of the claimed invention transformations on the solution of the problem of the invention. Therefore, the claimed invention meets the condition of patentability "inventive step".

"Industrial applicability" of the method is due to the presence of a developed modern elemental base, on the basis of which devices can be made that implement this method with the achievement of the destination specified in the invention.

The claimed method is illustrated by drawings, which show:

FIG. 1 is a block diagram of a resource allocation in a heterogeneous episodic computer network according to the proposed method;

FIG. 2 is a graph of the dependence of the number of expended resources on the number of agents in a heterogeneous episodic computer network.

Implementation of the claimed method is as follows (Fig. 1).

In block 102, heterogeneous agents appear on the network after connecting to the appropriate interface, allowing interaction to transfer resource packets to each other. Agents can be various electronic computers, including stationary, GRID, mobile devices, etc., having various communication interfaces. An important point is precisely the heterogeneity of both the agents and the communication channels between them. As the interfaces for agents, IR technology, bluetooth and Wi-Fi adapters can be used. In addition, wired network interfaces can be used: USB 2.0, USB 3.0, mini-USB, FireWire, i.Link (IEEE-1394), HDTV, RJ45, SCART, HDMI. (VG Olifer, H.A. Olifer, “Computer Networks.” 3rd edition, part II, chapter 10, p. 325).

The agent determines which resources or resource packages it needs. As resources, computational (processor time), semantic (necessary data files), network (Internet access), and arbitrary services (information about installed applications) can be used. Thus, the resource is heterogeneous. In addition, a distinctive feature is the ability to transfer resource packets.

In block 103, the interaction between the computing units is carried out by installing uniform unified means of inter-machine information exchange that implement the physical connectivity of the computing units within a heterogeneous computer network.

In block 104, each agent creates a copy of the decentralized database, updated automatically, similar to the formation of routing protocol databases in IP networks (“Routing in IP networks. Principles, protocols, settings”, Heather Osterloch), which stores information about applications and the current state of resources.

In block 105, the agent generates an application storing information about the request of several necessary resources. The application is generated in accordance with the transmission control protocol TCP / IP. The application is an information package. The TCP / IP stack protocol solves the problem of organizing the interaction between any arbitrary pair of nodes in a large composite network (V. G. Olifer, N. A. Olifer, “Computer Networks.” 3rd edition, part IV, p. 563). The protocol data stack is independent of the physical data transmission medium, which ensures completely transparent interaction between wired and wireless networks. The data in the information packet is stored in accordance with the structure of the TCP header in a designated field (E. Tanenbaum, D. Weatheroll, “Computer Networks”, 5th edition, chap. 6.5.4, p. 592).

At block 106, the agents submit bids for the auction. An auction is submitted by creating an appropriate entry in the table of a decentralized database, automatically updated by each agent of the network (“Databases” edited by Professor A. D. Homonenko, St. Petersburg, CORONA print, 2000, part 1, p. 10 )

In block 107, incoming requests to the agent’s address are stored, as well as automatic updating of the corresponding entries in the decentralized database tables of other agents (“Routing in IP networks. Principles, protocols, settings”, Heather Osterloh).

In block 108, the identifiers of agents operating within a given heterogeneous computer network are stored in the corresponding database table (“Routing in IP networks. Principles, protocols, configuration”, Heather Osterloch).

In block 109, packet processing with the application is processed. To process a package with an application stored in a database, the DBMS provides two types of languages for programs and users: a data description language, a high-level non-procedural language of the declared type, designed to describe the logical data structure, and a data manipulation language, a set of constructs that provide basic operations for work with data: input, modification and selection of data on request. The language combining the properties of both types of languages is SQL (Structured Query Language) ("Databases" edited by Prof. A. D. Homonenko, St. Petersburg, CORONA print, 2000, part 1, p. 11).

In block 110, a comparison is made of the required resources indicated in the customer’s application with existing data from the implementing agents. Comparison of applications is carried out by applying a query language using the SQL sample. Examples of SQL queries are join queries, server queries, control and subordinate queries. The request to the server transfers commands through the ODBC to the SQL server and allows you to directly work with tables on the server instead of joining them. The result of the query is the search for the required resources and loading the records into the table (“Databases” edited by Professor A. D. Homonenko, St. Petersburg, CORONA print, 2000, part 3, p. 269).

In block 111, if a resource package that meets the criteria of the request is detected, then a resource exchange takes place (block 8), if such a resource package is not found, then the search for matching continues with other agents (block 5).

In block 112, an exchange of resources is carried out if the application processing node finds resources that satisfy the request for an information package according to the submitted application (block 3). Resource exchange is carried out in accordance with the transmission control protocol TCP / IP (VG Olifer, NA Olifer, “Computer Networks.” 3rd edition, part IV, p. 563).

The degree of achievement of the technical result was determined using simulation models of a prototype system and a system that implements the claimed method of distributing resources between agents in a heterogeneous episodic computer network.

The simulation was carried out on the basis of the GPSS modeling language in the GPSS World software environment ("System simulation. GPSS World tools. St. Petersburg, BHV-Petersburg, 2004, Ch. 5, p. 158).

The initial data for the experiment were the following data:

1) the number of heterogeneous resources provided n = 90;

2) the number of agents in a mobile episodic wireless local area network k = 400;

3) loading a node for processing applications from agents using homogeneous resources and homogeneous connections between agents A = 0.7, where A = λ⋅τ, while λ is the intensity of receipt of applications from agents, and τ is the average time for processing applications;

4) loading the application processing node from agents using the proposed method for distributing heterogeneous resources and heterogeneous relationships between agents A = 0.4, where A = λ⋅τ, while λ is the rate of receipt of applications from agents, and τ is the average search processing time applications;

5) the ratio between the costs of homogeneous and heterogeneous resources in the network is γ = 2.75.

A comparison of the experimental results shown in FIG. 2 shows that the application of the proposed method reduces the computational and information costs of the distribution of heterogeneous resources in a heterogeneous computer network in comparison with the prototype method.

Claims (1)

A method for distributing resources between agents in a heterogeneous episodic computing network, which consists in providing interaction between computing facilities by means of inter-machine information exchange, creating a secure data warehouse on each computing installation; provide guaranteed storage of incoming messages in a secure data warehouse; provide guaranteed processing of the message body; characterized in that agents representing the computing facilities are connected to the heterogeneous episodic computing network, and then they provide interaction between the computing installations; after they create a decentralized database that stores information about the current state of resources, agents form and submit applications that indicate which resource agents are willing to offer and at what cost and what resource agents can purchase and at what cost, and then provide guaranteed storage incoming messages are stored in a decentralized database a message routing table, as well as agent identifiers, while the incoming message is an application for participation in the auction; provide guaranteed processing of the message body, compare message contents and exchange resources between agents.

Images