JP2002527994A - Self-predictive communication network - Google Patents

Abstract

A self-predictive communication network and method are provided, that is, a computer network configured to simultaneously act as a communication medium and a simulator. The computer network (100) includes at least one originating unit (102) configured to generate a message, at least one terminating unit (103) configured to receive a message, It has at least one intermediate device (101) coupled to both the originating and terminating units for passing messages therebetween. The intermediate device is configured to respond to the message to provide future status information of at least a portion of the computer network.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

TECHNICAL FIELD OF THE INVENTION

The present invention relates generally to communication networks, and more particularly, to communication networks that can simultaneously model and predict their own behavior by acting as a communication medium and simulator.

[0002]

[Prior art]

Massive global communication architectures have emerged as a result of networking millions of computers via the Internet. Wired, wireless and satellite technologies connect billions of different processors, referred to herein as network users. The goal of network management is to allocate network resources to meet specific operational objectives, for example, to provide the best possible quality of service for all user flows in the network. .
The challenges of managing these large numbers of networks, as well as the challenges of designing and implementing new network architectures, are rapidly compounding.

Given the future use of the network, the problem of allocating network resources is greatly aided. One way to manage a complex network is to use a simulator to model the behavior of the network and predict future performance of the network based on the modeled behavior. The network resources are configured using the models and predictions created by these simulators. That is,
Allocate components such as processors and routers to avoid bottlenecks and optimize network performance.

[0004] To that end, a number of schemes have been developed to simulate the behavior of computer networks. For example, the field of parallel discrete event simulation (PDES) has grown over the last 15 years to address the need to simulate complex networks. Various tools based on PDES have been developed to simulate and predict the behavior of the network in various conditions so that an optimal network design and configuration can be realized. The means for performing this simulation typically comprises a plurality of dedicated parallel distributed processors programmed to execute a discrete event simulation algorithm.

[0005] However, these processors are typically dedicated to simulation tasks and are generally not configured to simultaneously execute network user applications. Typically, the simulators themselves are not as large or complex as the actual networks they are trying to simulate. For that reason, such simulators have drawbacks that limit their usefulness in modeling the performance of "real" networks. For example, scaling solutions often arise when applying solutions and predictions based on simulator models to much more complex real-world cases. Current network modeling tools and simulation schemes cannot be easily scaled to systems such as those having the complexity of the Internet.

[0006]

[Problems to be solved by the invention]

Accordingly, a need exists for an improved method and apparatus for modeling and predicting the behavior of a communication network and managing the network based on the prediction.

[0007]

[Means for Solving the Problems]

The present invention provides a self-predictive communication network and method, i.e., a computer network configured to simultaneously act as a communication medium and simulator. The computer network is coupled to at least one originating unit configured to generate the message, at least one terminating unit configured to receive the message, and to both the originating unit and the terminating unit. , At least one intermediate device between which messages are passed. The intermediate device is configured to respond to the message to provide information on a future state of at least a portion of the computer network.

[0008]

BEST MODE FOR CARRYING OUT THE INVENTION

FIG. 1 is a sketch showing the overall structure of a typical computer network 100. The network 100 includes one or more originating units 102, 1
One or more terminating units 103 and one or more intermediate devices 10
One. The originating unit 102 and the terminating unit 103 are shown separately in FIG. 1 for explanation and for clarity. However, it should be appreciated that the units of the network are generally configured to operate as both components of the originating and terminating portions, and can send and receive messages over the network. In this specification, the word "configured"
It means "including appropriate parts composed of software parts and hardware parts, which are programmed and assembled to be operable". Typically, the sending and receiving units 102 and 103 consist of various processors and computers, including personal computers, mainframe computers, minicomputers and laptop computers. Typical intermediate devices include routers, switches, repeaters, hubs, and the like.

The units 102 and 103 of the network and the intermediate device 101 are connected to one another via a network 104. A network is defined as two or more computers connected to exchange messages and share data and system resources. In one embodiment of the present invention, network 10
4 indicates the Internet. The Internet is available in several countries, such as the United States,
A wide area network (WAN) that connects multiple (eg, thousands) networks in Canada, Europe and Asia. Generally, the Internet
To enable global communication between computers on a network selected from the group consisting of, but not limited to, governmental, educational and industrial networks. Other embodiments of the present invention include those in which network 104 comprises a local area network, interconnected private or private network lines, a wide area network, and combinations thereof.

In the case shown in FIG. 1, communication is achieved by relaying messages between the units 102 and 103 via the network 104. In one embodiment of the present invention, a "packet switched" architecture is used to relay messages. The term "packet exchange" refers to a type of network in which messages are composed of relatively small units of data called packets, and are delivered through the network based on the destination address contained in each packet. To tell. By dividing the communication into packets, many users in the network can share the same data path.

Typically, a packet has a field containing delivery information and a field containing data representing at least a portion of a message. Packets are delivered between the originating and terminating parts according to a predetermined communication protocol. Examples of common communication protocols include Transmission Control Protocol / Internet Protocol (TCP /
IP) and Open System Interconnection (OSI) protocols.

FIG. 1 also shows a conventional simulator 105. The simulator 105
The program is programmed to simulate the network 100 using a conventionally known simulation technique. One example of such a technique is parallel discrete event simulation. Simulator 105 is shown away from network 100 to show that in a typical conventional system, the simulation is performed independently of network operation. Simulator 105 simulates based on knowledge of network 100 provided by an operator or data gathered by a program designed to gather network data.

In the present invention, at least one transmitting unit 102 is configured as shown in FIG. In one embodiment of the present invention, at least one originating unit 102 is a computer, for example, a personal computer connected to the Internet. In another embodiment of the present invention, at least one originating unit 102 comprises:
A dedicated processor connected to the local area network. As those skilled in the art will appreciate, the originating unit 102 can be selected from a group consisting of a wide variety of different processors and computers in different types of networks. In the embodiment shown in FIG. 2, the transmitting unit 102 includes a processor 402, a memory 403,
And a terminal input / output (I / O) device 4 coupled to the display and keyboard 406
It has ordinary computer parts like 05. A data bus 407 interconnects the components so that data and commands can be passed between them. A control line unit (LCU) 420 interconnects the originating unit 102 with the network 104 (also shown in FIG. 1).

However, unlike conventional computer components, the processor 402 of the originating unit 102 includes a drive unit 416 and a virtual message generator 415. The drive unit 416 is a processor configured to execute a “drive program”. As used herein, a "driving program" is a program that monitors the network user's applications as well as the resources these applications consume during their execution. The driving program is executed by the predetermined processor 4 of the transmitting unit 102.
Gather statistics pertaining to the user application and associated processes running at 02. Examples of this statistic include, but are not limited to, message flow,
There are message type, scale, time, etc. The driving program uses these statistics to plan for future processes to be run on the processor 402, including future message traffic and resource utilization information.

[0015] Many commonly used prediction methods of network traffic are suitable for use with the present invention. For example, a suitable scheme is IEEE / ACM INFOC
OM, 1998. And C.I. J., "Wavelet Domain Modeling of VBR Video Traffic," and Proc. IEEE / ACM INFOCO
M., 1998, C.I. Food and C.I. In his paper entitled "Proactive Network Failure Detection." Examples of commercially available drive programs suitable for use in the present invention include CA92037, Tory Pines CtLa
COMN available from CACI Products Company of Jolla, 333N
There is ETIII (registered trademark).

The virtual message generator 415 is a processor programmed to execute the driving process algorithm shown in Table 1 (TABLE 1). The virtual message generator 415 communicates with the drive unit 416 according to the drive process algorithm. The virtual message generator 415 generates a virtual message based on the information collected from the driving unit 416.

As used herein, the term “virtual message” refers to a message that includes data and destination information relating to a logical process. A logical process is a process that represents and emulates a physical process that is expected to be executed at a certain point in the future at the present point in time. A “physical process” is an execution task defined by a program code. Examples of physical processes include updating a table field in a database, incrementing a counter, and calculating a beam table for beam angles in a beam forming communication system.

Logic processes are known in the field of simulation as local virtual time (LVT; loca).
l virtual time) expression of its own intrinsic time known as (n
otion) is typically managed. Although a logical process now mimics a physical process, the processor must know the time at which the corresponding physical process is expected to be run. This time is the local virtual time (LVT)
It is. Thus, a logical process has physical processes and extra data structures and instructions to manage message order and correct operation as the system executes prior to real time.

As will be appreciated by those skilled in the field of simulation, logic processes may "roll back" due to causality violations or "crazy" messages. A description of rollback as a causal avoidance method is described in Jefferson and Sovietral's dissertation published in The Society for Computer Simulation Multi-Conference in 1985 in San Diego, California. Fast Parallel Simulations Used ". This paper teaches a scheme in which each unit of the simulated network is allowed to execute before real time at its own risk. When a message arrives that prevents an event from running, a rollback is performed to release the simulation that was performed. Rolling back a unit is not always difficult. However, the simulated event that has to be rolled back greatly complicates this task due to the fact that the message may have been sent to other units. Jefferson et al. Disclose using "anti-messages" to do the rollback. Anti-message is similar to the original message, but the original behavior is
A different message is that a certain action of "release" is performed. Rollback schemes are sometimes used in discrete event simulations. This scheme was typically not used in communication networks, but is suitable for use in the network of the present invention.

Since each unit and device shown in network 100 operates according to its own expressive time, ie, LVT, the latest virtual time, ie, the future in which each unit of the entire network has finished imitating. Determining the first point toward the destination is important for network management. The latest virtual time at which all units in the entire network have already imitated is known as the global virtual time (GVT). The term "real time" message (RT) refers to a message that has been generated and has the current time, or "wall clock" time stamped as time. RT messages are typically generated by network components during the course of executing a network user's application program.

As outlined in Table 1, virtual message generator 415 communicates with drive unit 416 to generate virtual message 400 by performing the following steps.
First, at (2), the virtual message generator 415 determines that the GVT has not exceeded the look-ahead time. That is, this is because the GVT is close enough to the real time (t) (within Λ) and the virtual messages it generates are executed close enough to the real time, making network performance unacceptable. Inspect that it does not deteriorate as much. In other words, Λ is the “look-ahead” time that limits how far in the future the simulation will run. In one embodiment of the present invention, Λ is 20
It is set to 0 seconds. However, as will be appreciated by those skilled in the art, Λ can vary greatly and involves a number of situations. In some cases, it may be desirable to run the simulations several days in the future, and in other cases, it may be desirable to run the simulations several weeks or more in the future. In contrast, some very dynamic networks will only run simulations in the next few milliseconds.

If the GVT of the network 100 is within Λ of the real time, the virtual message generator 415 queries the driving unit 416 for the message content M at a desired time in the future (LVT + Δ). The message content M is information corresponding to a certain aspect of the expected message traffic, for example, the flow rate. Next, the reception time RT of the virtual message is set so as to correspond to the desired future time (LVT + Δ). Finally, send the newly created virtual message.

[0023]

[Table 1]

FIG. 3 shows an intermediate device 101 of the present invention. The intermediate device 101 is selected from a group of network devices and components including a router, a switch, a hub, a repeater, and the like. Regardless of the type of device selected, the intermediate device 101 includes a processor 202 and a memory 203 connected together via a data bus 207.
It is typical to have ordinary parts such as The memory 203 has stored therein programs, data and messages for transmission and reception.

However, in contrast to conventional components, the intermediate device 101 of the present invention also has a network management unit 204 and a logical processor 208. Processor 2
02 real-time processor 209 acts as a typical network component processor, ie, it responds to real-time messages communicated over network 104, while logical processor 208 executes logical processes in local virtual time. It is configured to be. Further, the logical processor 208 includes the intermediate unit 1
01 is programmed to execute the logical process algorithm shown in Table 3 in response to the received virtual message. In the description of this specification, the real-time processor 207 and the virtual processor 208 are separately illustrated and described.
However, those skilled in the art will appreciate that a single processor can be programmed to perform the operations of both logical processor 208 and real-time processor 207.

Further, in contrast to conventional intermediate devices, memory 203 of intermediate device 101 includes real-time memory 210 and virtual memory 212. The real-time memory 210 stores the intermediate device 10 when the physical process is executed by the processor 202.
The current state information relating to 1 is stored. The virtual memory 212 includes a processor 202
Stores future state information about the intermediate device 101 when the logical process is executed. As used herein, the term "state" refers to a set of parameters that characterize the operating state of a unit at a given time, usually when performing a process. The tabulated set of parameters is stored in memory in the form of a "state table".

In simulation applications, as well as in this specification, “computation
The term "" refers to a set of actions that change the "state" of a unit (called "event computation"). Until the state is changed by the subsequent event calculation, the previous state is stored in the memory. If there is a rollback, the previous state is restored by copying the saved values.

In one embodiment of the present invention, real-time memory 210 and virtual memory 212 are utilized to determine the accuracy of an event calculation simulation. For this reason, it is possible to obtain an indication of how well the logical process simulates the corresponding physical process. After the logical process is executed by the processor 202 of the intermediate unit 101, the state of the unit 101 is stored in the virtual memory 212. Thereafter, when the processor 202 of the intermediate unit 101 executes the corresponding physical process, the state of the unit 101 is stored in the real-time memory 210. The two states are subsequently compared to determine how well they fit. If the states are the same, the simulation was accurate.

In one embodiment of the present invention, intermediate device 101 further includes a network management unit (NMU) 204. The NMU 204 has a management information base (MIB) in the memory of the intermediate device 101. The MIB is a virtual information storage device for accessing a managed object. As will be appreciated by those skilled in the art,
An "object" is a data representation that describes a portion of a unit. The displayed unit may be physical, for example, the intermediate unit 101, or may be a target of software. An object identifier, which is an administratively assigned name, gives each object type a unique identification name. The type of target, combined with the target case, uniquely identifies the target instance. For human convenience, the descriptor (des
Use a textual string called a criptor to indicate the type of object.

MIBs are commonly used to manage network objects. But,
In contrast to typical MIB object types, the NMU 204 MIB has object types that relate to the future state of the network 100. Further, the M of the NMU 204
The IB collects information from virtual message packets when delivered through the intermediate device 101.

In one embodiment of the present invention, at least one active message packet 500 including code portion 510 is deployed to gather information about the network from NMU 204 MIB. Code portion 510 indicates whether there is information that the MIB is collecting from other message packets that have previously passed through intermediate device 101, M
It is programmed to query the IB. In response to this query, active message packet 500 receives information from the MIB and passes the information to the active message packet.
It is embedded in the packet 500. Thus, as the packet 500 passes through the intermediate device 101, the content of the active message packet 500 changes. In the format described above, code portion 510 is a simple moving network management protocol (S
NMP) acts as a client, and the MIB acts as an SNMP agent. As will be appreciated by those skilled in the art, the active message 500 of the present invention is not limited to use in simulation applications. The active message 500 can be expanded to query a typical MIB for information relating to the current state of the network 100, thus allowing the network 100 to have mobile SNMP client capabilities.

In one embodiment of the present invention, the active message 500 is an MI message adapted to the present invention.
Expanded to query B. This MIB comprises information gathered from virtual messages relating to the future state of the network 100. An example of an MIB suitable for use in such an application is described in TABLE4 shown in Tables 4 to 9.

As shown in Table 4 to Table 9 (TABLE 4), the MIB of the present invention has three new types of managed objects, namely, in one embodiment of the present invention, the MIB itself (AvnmpMIB).
), A packet (Avnmp packet) and a logical process (logical process). These objects are the well-known Simple Network Management Protocol (SNMP)
The type of object itself represents one embodiment of the present invention.

As shown in FIG. 3, a virtual message is provided to logical processor 208. Table 2 shows the structural elements of the logical processor 208 according to the present invention, including: That is, a reception queue (QR) that stores messages in order of reception time (TR), a transmission queue (QS) that stores messages in order of transmission time (TS), an LVT table, and the current state of logical and physical processes are stored. Current status table,
State Queue (SQ), which periodically saves the state, the value of the sliding look-ahead window (A), and an acceptable value representing the allowable deviation between real time and LVT.

[0035]

[Table 2]

Further, the logical processor 208 is programmed to execute the steps of the algorithm shown in Table 3 (TABLE 3). As shown in Table 3, the logical processor 20
8 sets the time by executing steps 1 and 2 first, ie, logical processor 208 sets LVT to zero. Next, the logic processor 208 sets the real time “t” according to the wall clock time. In the embodiment of the invention shown in Table 2, the wall clock time is obtained from the Global Positioning System (GPS). Other means of determining wall clock time are commercially available and will be readily apparent to those skilled in the art. Step 3 repeats the first two steps indefinitely, unless a message in the receive queue (QR) is available.

If there are messages in the receive queue, step 4 determines the minimum receive time (TR)
Read the message with. Step 5 updates the current state based on the value of the message read in step 4. In step 6, it is determined whether the message is real or if it is out of tolerance. If so, start a rollback. Step 7 checks if a message has occurred in the past for the local virtual time. If so, start a rollback. If the message is real and the prediction is within tolerance, step 8 invokes any irreversible action that must be taken to perform the process. Step 9 checks whether the logic process has not exceeded its look ahead. If not, the logic process is performed based on the steps described below.

Step 10 updates the status queue based on the incoming message and the current local virtual time. Step 11 sets the time to local virtual time. Step 12 generates a new virtual output message, if necessary, based on the current input message and the local virtual time. Step 14 saves a copy of the output message in the send queue for future rollback cases. Step 15 sends a new virtual message.

[0039]

[Table 3]

[0040]

[Table 4]

[0041]

[Table 5]

[0042]

[Table 6]

[0043]

[Table 7]

[0044]

[Table 8]

[0045]

[Table 9]

FIG. 4 shows a message packet format 40 according to one embodiment of the present invention.
Indicates 0. In one embodiment of the present invention, the message packet format 40
0 is used for both virtual and real messages. In another embodiment, the message packet 400 is used for a virtual message and has a simplified format, i.e., one that does not include the fields used to record virtual time.
Used for reality messages. Field 401 is a typical header field that contains information about the source and destination of the message. Field 402 indicates a transmission time (TS). TS is the LVT of the transmission process when the message 400 is transmitted. Field 403 indicates a reception time (TR). TR is
The time at which the receiving process of the terminating unit is scheduled to receive the message 400. Field 404 identifies message 400 as either a message or an "anti-message." Anti-messages are messages that cause the receiving process to roll back due to out-of-tolerance or causal conditions. Such an error may occur, for example, if the TS of message 400 is earlier than the LVT of the receiving process, as will now be described. Field 405 indicates the actual content of the message, M, and 406 indicates the actual or clock time at which the message was sent. In one embodiment of the present invention, the clock time is determined by the Global Positioning System (GPS).

In one embodiment of the present invention, at least one of the messages described above is an “active” message as shown in FIG. An "active" message is a message that includes, in addition to at least one field shown in the packet format of FIG. In one embodiment of the present invention, the executable code portion 510 of the active packet 500 is a generic code constructed in the GAVA programming language, and when the packet 500 flows to its terminating unit 103, the processor 202 of the intermediate device 101 Performed by As the packet 500 passes through the network, the executable code portion 510 is sent to the network management unit 2 of the intermediate device 401 through which it passes.
From 04, information about the network is collected. Therefore, the packet 500
It provides information for managing the network 100 and allows the network 100 to configure itself to handle future traffic conditions.

FIG. 6 illustrates the interaction of the software portions of the present invention according to one embodiment of the present invention. When the power is first turned on, the network 100 transmits the real message RT.
Start to occur. The driving program 302 begins to predict future events, and the driving process 303 generates a virtual message based on the expected event. Logic process 307 receives both virtual and real messages. Logic process 307 examines each incoming message and determines whether it is virtual or real by examining the RT field of the message. If the message is a virtual message, logic process 307 looks up the transmission time (TS) of the message. The message transmission time is the LVT of the transmission process when the message was sent.

The logical process 307 stores the TS field of the virtual message in the logical process 307.
LVT. If the message has not arrived in the past with respect to the LVT of the logical process 307, the message is received in the reception queue (Q
Enter R). Next, logic process 307 retrieves the next message from the receive queue, updates its LVT, and processes the message. If an outgoing message is generated, a copy of the message is stored in the transmission queue, the RT is set, and the transmission time (ST) is set to the current LVT. Thereafter, the outgoing message is sent to the destination logic process 308. If the virtual message 400 arrives out of order, i.e., arrives in the past TS with respect to the LVT of the logical process 307, the logical process 307 may use the logical process 307 to restore the last stored state prior to the time TS. Have to roll.

In one embodiment of the present invention, the virtual message is sent to the logical process 30
8, the data is stored in the memory of the intermediate device 101, for example, a cache memory. The virtual message is indexed and the intermediate device 101
Are stored according to a key configured with the source-destination address of each message as it passes through. As subsequent messages with a key (source-destination address) that matches the stored key pass through the intermediate device 101, the new message is compared to the stored message. If the stored message has a larger time stamp value (RT), a causal violation may occur when the newer message reaches its destination process. In that case, a rollback is instructed.

As a result, early detection of a potential rollback condition for the network 100 is obtained, and the network 100 can appropriately respond and reduce the result of the rollback. For example, if many such rollback situations occur in the same aisle, the destination process may be slow. Alternatively, the destination process may move to a new destination processor to mitigate the temporal impact of the causal violation. After the comparison, older messages are replaced with newer messages in the queue.

While only certain preferred features of the invention have been illustrated and described in the drawings, various modifications will occur to those skilled in the art. Therefore, it is to be understood that the appended claims are intended to cover all such modifications that fall within the scope of the invention.

[Brief description of the drawings]

FIG. 1 is a sketch showing an overall configuration of a computer network.

FIG. 2 is a block diagram of a transmitting unit according to the present invention.

FIG. 3 is a block diagram of an intermediate device according to the present invention.

FIG. 4 is a diagram showing a message packet structure according to an embodiment of the present invention.

FIG. 5 illustrates an active message packet structure according to an embodiment of the present invention.

FIG. 6 is a block diagram showing a format of a software component of the present invention.

──────────────────────────────────────────────────続 き Continued on the front page (31) Priority claim number 09 / 415,431 (32) Priority date October 8, 1999 (Oct. 8, 1999) (33) Priority claim country United States (US) ( 81) Designated countries EP (AT, BE, CH, CY, DE, DK, ES, FI, FR, GB, GR, IE, IT, LU, MC, NL, PT, SE), EA (AM, AZ, BY, KG, KZ, MD, RU, TJ, TM), IN, JP

Claims (24)

[Claims] 1. In a computer network, at least one originating unit configured to generate a message, at least one terminating unit configured to receive a message, the at least one originating unit and the At least one intermediate device coupled to at least one terminating unit for passing the message from the at least one originating unit to the at least one terminating unit;
The at least one intermediate device is configured to provide future state information of at least a portion of the computer network in response to the message. 2. The computer network of claim 1, wherein said messages include both real and virtual messages, and wherein said originating unit includes a virtual message generator. 3. The computer network according to claim 1, wherein said message is an active message. 4. The computer network according to claim 3, wherein said active message comprises a code portion, said code portion being executed on said intermediate unit. 5. The computer network of claim 4, wherein the content of the active message is changed based on execution of the code portion of the active message. 6. The transmitting unit further includes a driving unit for predicting a state of a network, and the virtual message generator is coupled to the driving unit for generating a virtual message based on the predicted state of the network. The computer network of claim 2, wherein 7. The computer network according to claim 6, wherein said drive unit has a network traffic prediction program. 8. The computer network of claim 1, wherein said intermediate device is selected from the group consisting of a router, a switch, a hub, and a repeater. 9. The computer according to claim 1, wherein the transmitting unit, the receiving unit, and the intermediate device are connected to each other via the Internet.
network. 10. The computer network of claim 2, wherein the intermediate device further comprises a logical processor that executes a logical process in response to the virtual message. 11. The computer network of claim 8, wherein said intermediate device further comprises a virtual memory for storing a state of said intermediate device corresponding to an executed logical process. 12. The computer network of claim 8, wherein said intermediate device further comprises a network management unit (NMU) for collecting and storing information based on the content of said virtual message. 13. The computer network of claim 12, wherein said network management unit includes a management information base (MIB) configured to include packets as objects to be managed. 14. The computer network according to claim 13, wherein the network management unit includes an MIB configured to include a logical process as an object to be managed. 15. An at least one originating unit, comprising: a driving unit for predicting use of the originating unit; and a virtual message generator configured to generate a virtual message based on the predicted use. An originating unit, at least one intermediate unit, the intermediate unit executing a logical process in response to the virtual message, and a future state of the network based on the virtual message. A network management unit for providing relevant information. 16. The computer network of claim 15, wherein said message includes a code portion and a data portion. 17. An originating unit including means for predicting usage of the originating unit, and means for generating a first virtual message based on the predicted usage, and a logic responsive to the first virtual message. An incoming unit including means for executing a process; an intermediate unit including means for executing a logical process in response to the first virtual message and generating a second virtual message based on the executed logical process; A computer network comprising: 18. A method for determining the accuracy of an event calculation simulation, comprising: executing a logical process on a unit including a processor; and determining a state of the unit obtained by executing the logical process in the executing step. Creating a first state table including state variable values to be represented; storing the first state table in a memory; and executing a physical process corresponding to the logical process in the unit including a processor. Creating a second state table including a state variable value representing the state of the unit obtained from execution of the physical process; and comparing the first and second state tables with the state variable Determining the difference between the values. 19. A method for predicting the future state of a communication network having a plurality of interconnected units transmitting and receiving messages and at least one intermediate device passing said messages therebetween. Generating, by the first unit, a virtual message based on a plan for future use of the network; and executing a logical process on the intermediate device in response to the virtual message, the logical process comprising: Mimics the future use of the network,
Storing the state of a second unit associated with the logical process in a memory, wherein the stored state represents a future state of the second unit. Said method. 20. The method of claim 19, comprising monitoring the first unit's use of the network and planning future use of the first unit's network based on the monitored use. Method. 21. A method for managing a network, comprising: gathering information related to the state of the network from messages traveling through the network and storing the information in a MIB; and storing the information in the MIB. Querying the MIB for information that exists, and managing the network based on the results of the query. 22. The method of claim 21, wherein the MIB comprises information relating to a future state of the network. 23. The method of claim 22, wherein said query is performed by executing code contained in a code portion of an active message packet. 24. The method of claim 23, wherein said active message packet is a virtual message packet.