TW319944B - - Google Patents

Description

319944 A7 B7 V. Description of invention (1) Priority according to 35 11.5 _ (:. §119〇) & 37 (:.?. 11_§1.78 (&) (1) This non-provisional patent application is based on previous Provisional U.S. Patent Application 60 / 001,169 (Internal No. 27946-00094) " Effective Path Capacity Configuration Method for Broadband ATM Networks with Generic Message Types " The application was filed on July 14, 1995, and the applicants were Wlodek Holender and Szabolcs Malomsoky, and the application was also transferred to the assignee of the present invention. Comparison of related applications: This non-provisional US patent application contains The subject matter is related to the following application: Non-provisional U.S. Patent Application 08 / 514,480 (internal number 27946-00094) under review "System and method for configuring the optimal virtual path capacity using broadband message traffic", which was in 1995 The application was filed on August 1, 2011; and the non-provisional US patent application 08/5 13,723 (internal number 27946-00095) " System and method for performing adaptive path selection on a virtual path broadband network The case was filed on August 11, 1995. This application hereby cites these non-provisional U.S. patent applications and their disclosures for reference. BACKGROUND OF THE INVENTION Printed by the Employee Consumer Cooperative of the Central Department of Economics, Ministry of Economic Affairs 31 1. Technical Field of the Invention The present invention relates to a system and method for efficiently configuring the path capacity of a telecommunications network, and more particularly to configuring a logical link and a virtual network defined on a physical network with limited capacity The technology of the path capacity. 2 _Relevant technical description The use of the exchange equipment known as the municipal telecommunications bureau (l〇cal exchange) The Ministry of Economic Affairs Central Standards Bureau employee consumer cooperative printed ^ 19944 A7-~~ --________ B7__ V. Invention Explanation— (2) " '一' ~ 一 ~-to facilitate the interconnection of telephone equipment or other communication devices ^ located in a local geographic area. The trunk exchanges are connected to the municipal electricity L bureaus. Located in Each telephone / data equipment in geographically separated areas and connected to different municipal telecommunication bureaus is linked to a telecommunications network by a complex set of municipal telecommunication bureaus and The telecommunications bureaus communicate with each other. The telecommunications network therefore contains multiple interconnected network components, such as municipal telecommunication bureaus, trunk telecommunication bureaus, mobile wireless telecommunication bureaus, long-distance telecommunication bureaus, and combinations of these components in each At the network level, messages transmitted from one network component, such as a telecommunications office, to another network component can take various paths through different telecommunications offices. When performing efficient network message flow management on various communication facilities within a network, it is necessary to enable a sufficient number of circuits to handle each purpose without expanding the degree of congestion on the last selected message transmission path Local information flow requirements. It also requires that the network congestion on all the last selected paths be as uniform as possible, and that the unused capacity in each path actually provided is minimized to ensure efficient use of resources. In addition, the telephone company operating the network has a limited budget, so it must maximize the available resources of each network. In the past, the management of message flow in a communication network included the following procedures: • Investigate the message flow pattern in the network; and change the configuration of circuits and paths in order to increase the efficiency of message flow processing. In addition, when high call density is expected in a specific location or a specific area, or when high call density is expected for some local events in the area, more paths and circuits can be added to the network. The traditional network management system can change the internal location of the network ---------- -5 -_______ This paper is suitable for use by the Chinese National Standard (CNS) 84 (2 i 〇χ 297 mm) ( Please also read the notes on the back " fill in this page), —shang. 、 1Τ 31S944 A7 B7 V. Description of the invention (3) " ~~~ " Choose the relative distribution of the message traffic load between each path, so that In the case of not excessively reducing the quality of service (Q〇Sh々 for short), the use of the network is increased as efficiently as possible. However, traditional message flow network management systems and procedures are generally not easy to handle individual message flow requirements The number of available circuits and paths in the network is increased, but the network is reconfigured. The method is to reconfigure the capacity of each path and circuit in the network at a sound level of abstraction in order to maximize the network. In addition to the inherent need to maximize the efficiency of managing the physical network resources of telecommunications systems, the advancement of information and communication technologies in recent years has generated new business opportunities and new challenges for management. Providers of telecommunications services are constantly encountering new customer demands. Providing ordinary voice communications through telecommunications networks is no longer enough. Current users want to not only transmit language signals, but also transmit data in real time through the packet exchange network , Audio, video, and multimedia signals. Since Asynchronous Transfer Mode (ATM) technology can provide enhanced capabilities for broadband telecommunications facilities, the importance of ATM technology is increasing day by day. One of the main features of printed ATM technology is its flexibility in using network resources. One possible way to use this flexibility is to divide the physical network resources completely or partially into a number of logically defined resources. For example, if Splitting a physical network into multiple virtual networks greatly simplifies the operation, maintenance, and management of the physical network. With this procedure, it is possible to analyze and solve the message transmission of each call on an entire physical network The routing problem is simplified to a much simpler routing problem on each virtual network As a whole, the complexity of each individual virtual network is smaller than that of the entire physical network, so the _____ _- 6 -_— standard is applicable to China National Standards (CNS) Α4 specifications (210X297 mm ) '-319944 A7 B7 V. Description of the invention (4 The solution to the routing problem of message transmission by the Consumer Standardization Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs is relatively simple. Because there are different message types and different levels of service in the dedicated line network , Or constantly changing information flow requirements exist, so there is also a need to separate physical resources. The service provider does not provide an independent physical ancestral path for each client application and can establish several virtual networks, and these virtual The network is defined at the upper layer of a single shared ATM physical infrastructure. This new configuration flexibility in the virtual network structure requires efficient path capacity configuration tools, methods, and algorithms. Because the nature of telecommunications services to be provided in the future is difficult to predict, any path capacity allocation method used to manage the configuration of virtual network resources must be able to handle all types of broadband messages. If the configuration of the virtual network must be modified frequently in order to adapt to changes in the provided message flow patterns, the network reconfiguration path capacity and the computational efficiency of the fe-control system must be quite high. The algorithm selected to perform network reconfiguration path capacity and network configuration reconfiguration should perform its calculations in a period much shorter than the duration of each virtual network. Many existing path capacity allocation technologies do not meet the requirements of path capacity allocation technologies that can model general message traffic distribution. Most of the commonly used path capacity allocation methods cannot handle general message traffic models because of the limitations of these methods due to the use of Erlang blocking measures. The method and system of the present invention overcome these shortcomings. SUMMARY OF THE INVENTION Therefore, an object of the present invention is to allow users to configure the path capacity of a network using a general message flow distribution model. For another purpose of the present invention, please read the notes on the back of this page for binding -7- This paper wave scale is applicable to the Chinese National Standard (CNS) A4g (210X297mm) A7 B7 319944 5. The description of the invention lies in the use of the least computing resources The calculation of path capacity allocation is performed. A further object of the present invention is to implement a path capacity allocation algorithm with lower computational complexity and therefore faster speed. Another object of the present invention is that the calculation of the reconfiguration path capacity can be performed in the shortest possible time period. The present invention (again-aims to improve the efficiency of path capacity allocation and virtual path configuration procedures of a virtual network. If there is a physical network including a plurality of physical links, each physical link has a predetermined transmission capacity Then, the system and method of the present invention show a variety of path capacity allocation techniques that can support a general message traffic model. This kind of road allocation is considered as a problem of load balancing on various physical links. Virtual path allocation problem The best solution is equivalent to the selection of the allocated capacity on various virtual paths, where as much as possible to make the blocking of each virtual path on various links. In one aspect, the present invention includes-an efficient configuration of a telecommunications A method of path capacity of a network, wherein the telecommunications network has a plurality of physical interesting paths for connecting a plurality of telecommunications offices or nodes. The plurality of physical links are related to one or more virtual paths. Each virtual path is The telecommunications network provides a switchable individual connection between the offices or nodes of the telecommunications network. Provision is provided for each virtual path Information flow, and set a transmission capacity limit for each physical link of the telecommunications network. Establish the provided message flow and other calculation parameters on the telecommunications network using an entropy-blocking measure The relationship model, and according to the transmission capacity limits of various physical links, the capacity is allocated to the plurality of virtual paths'. Therefore, within a preselected error limit, the various virtual _______________- 8-This paper standard is applicable China National Standards (CNS) A4 specification (210X297 mm) Printed by the Central Business Bureau of the Ministry of Economics, the Ministry of Economic Affairs, the Consumer Cooperative, 3, 99, 4 A7 --------- B7 5. Description of the invention i: 6) The blocking probability on the path should be as uniform as possible. BRIEF DESCRIPTION OF THE DRAWINGS With reference to the detailed description of the preferred embodiments below and the accompanying drawings, it should be possible to have a more complete understanding of the method and system of the present invention. FIG. 1 is a block diagram illustrating a telecommunications network. The capacity allocation of virtual paths can be affected in the telecommunications network; FIG. 2 is a block diagram illustrating the structure of an ATM cell packet; FIG. 3 is a block diagram of several interconnected virtual paths and virtual channels in an ATM network; Figure 4 is a block diagram of the handover and exchange of virtual paths and virtual channels in an ATM network; Figure 5 is a schematic diagram of the CCITT B-ISDN reference model, showing the various service levels and standard layers supported; Figure 6 shows An example AT Μ network providing virtual leased line (virtual leased line; abbreviated as vll) service is shown; Figure 7 shows a multi-layer SDH transmission network containing several ATM handover devices t printed by the Beigong Consumer Cooperative, Central Bureau of Standards, Ministry of Economic Affairs 8 shows the topological relationship between each virtual path and the physical link; FIG. 9 is an exemplary push-down algorithm for configuring the capacity of each virtual path defined on a physical network Flow chart of steps; Fig. 10. Characteristic curve diagram of the Entropy Rate Function related to the present invention; ____- 9- This paper scale is applicable to the Chinese National Standard (CNS) A4 specification (210X297 published ) 3'ίΗ44 ______ V. Description of invention ί: Α7 Β7 The Ministry of Economic Affairs Central Standards Bureau Beigong Consumer Cooperatives printed 11, circle 11 is a characteristic curve with the average message volume ratio function shown in Figure 10; Figure 1 2 Flow chart of each step of the pseudo-path capacity allocation algorithm of the average message volume ratio function related to the present invention; FIG. 13-Flow chart of successive steps of the adaptive path capacity allocation algorithm related to the present invention; Exemplary average message volume ratio function t for different virtual paths. FIG. 15 shows the calculation steps of the adaptive path capacity allocation algorithm related to the present invention; FIG. 16 shows an example physical network with six nodes; 17 shows the connection topology of ten virtual paths defined on the six-node network shown in Figure 丨 6; Figure 18 does not show a three-layer abstract model to display each physical path, The topological relationship between the pseudo link and the virtual path; Figure 19: A flow chart of the conversion procedure to transform the virtual network configuration path capacity problem into an equal virtual path capacity configuration; Figure 20 shows a An example virtual network that is linked and placed on the six-node network shown in Figure 16; and Figure 21 shows five different subnets placed on the physical network of the six nodes shown in Figure 16 road. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Congestion Control in a Traditional Network First, please refer to FIG. 1 ', which shows a schematic diagram of a traditional public telecommunications network. The virtual parameters of mobile parameters are related to Min. -This paper music scale is applicable to the Chinese National Standard (CNS) A4Ai grid (210X297 mm) 319944 A7 _________B7 5. Description of invention (8), the public telecommunications network includes a plurality of municipal telecommunications bureaus (21_26), each municipal telecommunications The bureau has a number of local users connected to it, and these local users are represented by telephone equipment (27). Two of the municipal telecommunications bureaus (21) and (24) represent remote users with multiplexing levels (28) and (29) connected to them. These remote users with multiplexing levels also have local users connected to them (27). The network shown in Figure 1 is connected to multiple trunk telecommunications offices (31 to 34). These trunk telecommunications offices are mainly used to interconnect the municipal telecommunications offices and provide paths between various parts of the network. The illustrated trunk telecommunications office (3 1} is connected to a mobile telephone exchange (35). The mobile telephone exchange (35) includes a pair of illustrated base stations (36) and (37) to serve a plurality of Mobile radiotelephone users (38). In addition, other telecommunication services such as databases and smart networks can be connected to the telecom offices shown in the figure. In this network, the telecom offices (21- j5) shows a plurality of communication paths (30), each communication path may include a plurality of communication circuits including cables, optical fiber links, or radio links, etc., for each telecommunications office within the network Voice and / or data communication is carried between. Printed by the Central Bureau of Standards ®C Industrial and Consumer Cooperative of the Ministry of Economic Affairs. The network shown in Figure 1 includes a network control system (40), which is via a communication link (41) (Represented by a dotted line) Connect the network control system (4〇) to each telecommunication office (2 1-3 5) in the network to transmit control signals to and receive from each telecommunication office Message data. The network control system (4 〇) issues commands to dynamically access the network Reconfigure each communication path within each message transmission path and control the warning system of each telecommunication bureau on the network to fine-tune the reduction of congestion in the network. AT Μ System Concept ___- 11-This paper size is common China National Standard (CNS) Α4 specification (210X297 mm) B7 V. Invention description (9 Guang ^-Many changes are taking place in the currently implemented public telecommunications transmission network. A major goal of public telecommunications network operators That is, a single type of technology handles all types of telecommunications services in a common infrastructure. One such technology is asynchronous transmission mode (ATM) technology. ATM is currently being implemented in an attempt to meet these needs in a way that generates A carrier telecommunications network 丨 The telecommunications network has considerable “quot bandwidth bandwidth” (bandwidth granularity), which can cope with the connection of extremely high frequency bandwidth. The term bandwidth splitting power is a characteristic of the network , Where the bandwidth required for beer calls can be freely changed throughout the duration of the call.

When using AT Μ technology in a public telecommunications network, it can provide the following capabilities: shared service related switching and transmission, greater bandwidth division capability, support for variable bit transmission rate services, and support for multimedia services . Because of these characteristics, the International Telegraph and Telephone Consultative Committee (CCITT) has selected ATM as the core technology of Broadband ISDN (Broadband ISDN; B-ISDN). Although the ATM Ministry of Economic Affairs Central Standards Bureau's Beigong Consumer Cooperative has some shortcomings, including low-speed synchronization service transmission delays, increased complexity in the network, and the need to add some new performance parameters (such as unit packet discard and congestion ), But CCITT still makes this choice, and the present invention will address these shortcomings, which will be described in detail below. An ATM network can be implemented using a plesiochronous digital hierarchy (pdh), or synchronous digital hierarchy (SDH), or both. In addition, the ATM and STM (synchronous _____- 12- __—_) can be dealt with in accordance with the Chinese National Standard (CNS) 84 specifications (210XW7mm) A7 5. Invention description (1〇) t-side fer mode step When multiple messages are communicated between the transmission modules, you should read the ruthenium. Note that Ια fills in this page, and the resulting performance is reduced, you can use ATM as a way to carry the message. Network 2 is not (ATM cell packet structure is the core part of ATM technology. -Atm cell packet has a fixed length of 53 bytes, these 53 bytes are divided into a 5-byte start code, and a 48-byte information column Bit (also called payload). The structure of the start code of the ATM unit packet is-a number of bits, its main function-that is to help the ATM unit packet from the beginning

The source point is routed to the destination point via one or more switching nodes. The information stored in each ATM cell packet is kept to a small amount in order to reduce the capacity of internal buffers in each switching node and limit the waiting delay in these buffers. AT M is operated in connection-oriented mode. From a modeling point of view, the above method is very important, because this method can use the existing circuit parent to replace the mathematical model to optimize the allocation and control network resources. The Ministry of Economy Central Standards Bureau staff consumption The main function of the ATM unit packet printing start code printed by the cooperative is to identify the virtual connection. The routing information in the ATM unit packet is included in the following two fields: a virtual path identifier (V PI) used to determine which virtual path the ATM unit packet belongs to, and a A virtual channel identifier (virtual channel identifier; VCI for short) that determines which virtual channel in the virtual path the unit packet belongs to. A virtual channel is a dynamically assignable end-to-end connection. The best transmission link can transmit hundreds of millions of bits per second, however the virtual channel may only fill thousands of bits per second. Therefore, it can be transmitted in a single--13 format. The Chinese National Standard (CNS) A4 specification (210X29 * 7mm) is applicable. A7 319944 _____B7 5. Invention Description (11) A large number of simultaneous virtual channels are supported on the link. On the other hand, a virtual path is a semi-permanent connection between endpoints. Each virtual path can transmit a large number of simultaneously connected virtual channels. Because a large number of virtual channels are processed and exchanged together in a single unit, the total processing requirement of a virtual path is less than a virtual circuit, so each _ (virtual) circuit has faster processing, and can be much higher The efficiency of using network resources. Network management of virtual paths is relatively simple and efficient. As shown in Figure 2, the AT-M unit packet start code on the user-net interface (UNI) and the ATM unit on the network-node interface (NNI) The starting code printed by the Beigong Consumer Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs is slightly different. UNI includes general flow control (GFC), and is used to ensure fair and efficient use of the available capacity between a terminal and the network. A payload type indicator (PTI) field is used to indicate whether the ATM unit packet contains user information or special network information such as for maintenance purposes. A cell packet discarding priority (CLP) field is referred to as a two-level priority encoding, and this CLP field is used when the cell packet must be discarded due to network conditions. The start code error control (header error control; HEC for short) field contains a check and protection of the start code information. When using ATM unit package, the information transmission rate can be adjusted according to the actual service requirements. The number of unit packets per unit time can be increased to the upper limit of the transmission bit rate of the physical medium used to carry data according to the required capacity. In addition to the data unit package, there is a single book for sending and maintenance. The Zhang scale is suitable for China) 8 4 inspections (2 similar to 297 directors)--3 ^ 8944 A7 B7 5. Description of invention (12 Printed by the Employee Consumer Cooperative of the Central Ministry of Economic Affairs = packet, and idle unit packet. The sending unit packet is used between the end users in the network or between nodes in the network. The main crime of unit seal I is to establish a mutual service (such as a connection). The maintenance unit package provides supervision of the ATM layer, and the idle unit package is used to transfer the capacity to the transmission media. 3. The figure shows a block diagram of the parent exchange and the parent connection of each virtual channel and virtual channel in an ATM link. From the perspective of a switch designer, " VP exchange " is only using identification The upper part of the code shelf (that is, the shorter field (VPI) exchanges an ATM unit packet. Conversely, in "vp / vc exchange," the entire identification code field (VPI and VCI) is used._ VP / VC path contains multiple interconnected VP / VC lengths. Switching can be performed at the VP or VC level Handover. The virtual path identification code (vPI) and the virtual channel identification code (VCI) define a two-layer processing and routing structure within the ATM circuit. From a network architecture perspective, a virtual path (V p) is— The individual connection of the group is also a type of "highway" in the path comparison table of an atm network. An important task of network management is to allocate the correct amount of transmission capacity to each such highway (also That is, a virtual path) 'in order to optimize the performance of the network. This optimization is the goal of bandwidth management or virtual path capacity configuration technology, and is also a subject of the present invention that will be detailed in the following Then please refer to Figure 4, which shows the concept of virtual path and virtual channel handover and exchange. The virtual path identification code (V PI) and virtual channel identification code (VCI) value is only valid for a specific link. During the handover / exchange, the new V PI / VCI value is assigned to the unit package, and the physical connection is chun and η. First: Idle, read the back s. 'Matters, then I set -15. The paper size is applicable to Chinese national standards CNS) A4 specification (210X297 mm) A7 ----: -----_ B7 V. Description of the invention (13) The combination of VPI / VCI value I enables the identification of the ATM unit packet. Then, as shown in Table 丨Execute with the help of the conversion table—exemplify the routing of the ATM unit packet. The table 1 is as follows: --- 1 From where the port connects VPI VCI ---- ^ To where the port VPI VCI A 1 -----_ 丨 · ~ _ —:-—C 10 _ A 2-D 6 B 5 3 —- C 7 4 B 5 2 D 3 5 Table 1 An ATM simple packet is a basic multiplexing unit within an ATM transmission system A unit packet or information unit contains its own connection and routing information. This feature allows direct multiplexing and demultiplexing of each service channel, where each channel can carry a different bit rate. Use the information contained in the start code in the virtual path identification code (V PI) and virtual channel identification code (v c 丨) fields to identify each atm unit packet and set the transmission path of each atm unit packet. As mentioned above, a virtual path (VP) is a set of multiplexed circuits between two terminals (such as a switching system, LAN (LAN) gateway, or dedicated network gateway). A vp provides a direct logical link between the terminals of each virtual path, and the specific virtual path is identified by the V PI value. Also as mentioned above, the virtual path concept used in the AT M technology can handle multiple virtual channels in a unitary manner (VC). Can have the same service as ------------- 16ji_________ This paper standard is applicable to China National Standard (CNS) A # specifications (2 丨 〇xqiong 7mm) A7 B7 5. Description of the invention (Virtual + I pseudo-channels with common characteristics such as 14-port f (QoS) are combined into a group that can be transmitted, processed, and managed in a single unit by embedded mail A tm m,. This flexible combination simplifies The operation and maintenance of the AT Μ system. Both the virtual path and the virtual channel can be used for semi-permanent road control in the ATM network. One operation support system is set up. I 2 heart X connection or Create and release paths along various tools along a path. Sending messages between a user and the network, or sending messages within the network, you can use the established connection to make each virtual channel available for use The exchange of time. An important feature of ATM technology is related to its communication protocol architecture, and the establishment of the communication protocol is based on the so-called "core and edge" principle. Terminals on the ATM network " edge " Perform retransmission, flow control, and delay equalization in the The communication protocol function related to the type of information. This universal method retains an efficient and non-specific service "core", network, and the network contains only simple unit packet transmission and switching functions. At this core ATM node There is no error check on the information field and no flow control. Just read the information of the unit packet, then use hec to correct the unit error that can affect the address, and then exchange the unit packet towards the destination of the unit packet Printed on the edge of the network using the ATM adjustment layer (ATM adapUti〇n 丨 Acr ^: AAL for short) printed on the edge of the network by the Central Sample Bureau of the Ministry of Economic Affairs. The CCITT reference model for B_IS DN service considers: AAL contains the functions of specific services. As shown in Figure 5, there are three layers in the ATM standard. The first layer is the physical layer that defines the physical interface and the frame communication protocol. The second The Atm layer has nothing to do with the selected real media, and defines the unit packet structure, and provides multiplexing. 17- This paper music standard is applicable to the Chinese National Standard (CNS) A4 specification (210X297 Mm) A7 B7 5. Description of the invention (; 15 and demultiplexing, and VPI / VCI conversion to control the flow of each unit in the logical network. The third layer is AAL, AAL provides between the service and the AT Μ layer Important adjustments, which can allow atm transmission of non-specific services. AAL is implemented to map the service format to the packet information field of the ATM unit. Exemplary functions provided by AAL include variable-length packet description, sequence number, and time Pulse recovery, and performance monitoring. Deploying ATM in a telecommunications network can use ATM technology on the client side to support high-speed data communication within and between users' regional networks. In addition, atm can be used as an infrastructure resource shared by all services in a client network including voice and video communications, data transmission, and multimedia applications. The example service generated after adding ATM nodes to the public telecommunications network is to provide virtual private line (VLL) services. The VLL service system is based on the concept of virtual paths that can directly adjust the line capacity according to user needs, and it is easy to change the line valley volume without modifying the interface structure. A large number of logical connections can be provided to users via the user-network interface (UNI). In addition, the service quality adjusted according to the needs of users can also be provided to users in order to meet the needs of users. Therefore, a variety of service levels, service quality levels, and performance parameters can be selected. For example, voice services require lower transmission delays but can tolerate higher bit errors; on the other hand, data communications are more tolerant of network delays ’but are very sensitive to bit errors. Therefore, the service provider and the user can agree on the service quality level of a specific service in a contractual manner, and perform audits manually or automatically to ensure compliance. As shown in Figure 6, the figure shows an example virtual frequency implemented in a TM network -18 This paper standard is applicable to the Chinese National Standard (CNS) A4 specification (210X297 mm) Please. Read the back *, i matter Binding cooperation of employees of the Central Standardization Bureau of the Ministry of Economic Affairs. Du Printed S1S944 Μ Ι ______ _ Β7 1 V. Description of invention (16) Road-type VLL service. The network terminals (a-E) are coupled to the ATM handover node (611-614) via the process enhancement nodes (601-605), respectively. The AT M network includes a plurality of A T M handover nodes (611-614). These handover nodes can provide path setting at the virtual path and virtual channel level. The process enhancement node (610-1 605) is located at the edge of a τ μ network, so that the network will not be affected by potential overload. This function ensures that any connection will not meet the conditions agreed upon when the anti-connection was established. Add services to one or more handover nodes (611-6M) to implement additional services. In the network shown in Fig. 6, an example of a non-virtual path is shown by a wavy line (621) between terminals (c) and (D). The dashed line (63 丨) shows the first virtual connection between the terminals (A) and (B) and the dashed line (632) between the terminals (C) and (E) shows the second virtual connection. In addition to the virtual private line network shown in Figure 6, as long as several servers are connected to each ATM node in the network, other services such as SMDS / CBDS and frame transmission can be easily added as needed. In residential areas, ATM technology can also be used to provide enhanced new entertainment services such as on-demand video (〇n_demand vide〇) to end users. The flexibility of the Α τ μ network allows it to support a variety of services, such as long-distance education, squatting at home, and video game services. Printed by the Beigong Consumer Cooperative of the Central Standards Bureau of the Ministry of Economy. Figure 7 shows an ατμ network placed on the SDH layered transmission network. These two layers include the client network layer (701), a terrestrial transmission network layer (702), a regional transmission network layer (703), and a national transmission network layer (704). A plurality of ATM enterprise network nodes (Chuan · 714) control data flows from each client's thin computer (715) and each LAN (716) into a plurality of add_dr〇p multiplexer 1 ——_ I y-This paper standard is applicable to China National Sample Falcon (CNS) A ^^ ~ f210x29 ^ En> A7 B7 5. Some multiplexers in the description (17 (721), these add_dr〇p multiplexers (721) serve the local transmission network (705 ) The SDB interconnection nodes (722) in the). Each local interconnection node (7 2 2) is coupled to the regional transmission network via the regional interconnection nodes (7 3 丨), and the add-drop multiplexer (732) Two regional interconnection nodes (731) are coupled, and within the local transmission network layer (702), there are one ATM ATM proximity node (723) and a multiplexer (721) with heart-to-heart multiplexers (721) and The SDH ring serves each interconnected node (722), and is used for user proximity at a full s T M-1 capacity (丨 55 million bits per second π, which is the standard proximity rate of B-ISDN services). . The use of remote multiplexers and other near-end nodes that provide the final local loop connection can also be performed on this ring network such as ordinary old telephone services (Plain Old Te Lephone Service · 'Abbreviation Ρ 〇T s) and other existing message transmission. Each ATM proximity node (723) is shared in order to use different services from one location, and each ATM proximity node (723) can include the use of different νρ / Voice and data of vc. In these ATM proximity nodes (723), ατμ messages are concentrated 'in order to efficiently use the transmission capacity. The size of the ATM proximity node printed by the Employee Consumer Cooperative of the Central Ministry of Economic Affairs of the Ministry of Economy can be based on The capacity varies, from-small multiplexer to-large handover equipment. In the regional transmission network material (703), ATM handover equipment (733) is used to send messages between local areas. Transmission in the country In the network layer (704), as shown in Figure 7, ATM is not obvious. In the -ATM stacked network, as shown in Figure 7, it is easy to add services such as frame transmission and SMDS / CBDS. With the addition of appropriate software and hardware, the B-ISDN function can also be added to the proximity nodes and regional nodes. As also shown in Figure 7, it can be implemented such as operating according to the CCi_mn standard -20- The private paper standard is applicable to China Standard (CNS) A4% iS -(210 ^ 297¾ B7 V. Description of invention (18 of a network management system (750), in order to provide the necessary network management functions to the SDH and ATM components of the network. According to Telefonaktiebolaget LM Ericsson (i.e. In this application, the telecommunications management and operation support (TMOS) series of network management systems provided by the application's implementation subsystem (75〇) manages the ATM network. Such steel road management may include various functions such as routing algorithms and congestion control implemented in accordance with the disclosure items detailed below. Virtual path capacity configuration: = A practical model that sets the capacity of each path in the telecommunications network is to consider the problem of configuring path capacity as a problem related to the two-layer structure. This two-layer structure includes: one with a discontinuous topology and The first physical network layer specifying the link capacity and a second virtual roadblock layer with several virtual paths and their specific routes provide traffic demand to each virtual path in this model. In Bezhazhong, which only deals with network capacity allocation, in fact, the order of bypassing each virtual path is already gone. Each virtual path passes through a number of physical detours, but the simulation of each virtual path contains only the height of a single path. Each virtual path will have a characteristic blocking value d) The varL printed by the Consumer Labor Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs and a characteristic allocated capacity value (al1. She d—), and the number of variables in this transformation is The number of virtual paths is the same. The term " provided message flow is biased t- the demand for transmission capacity on the virtual path that changes with time. Every two: "dwu" message traffic demand ,, (', trafflc demand ,,) means the time average value of the provided message traffic. When it can be single-parameter A7 _______ B7_____ V. Invention description (19) of Bwason When Poissonian distribution establishes the characteristics of message traffic on an ATM network, such message traffic can be called homogeneous single-class traffic. When the provided message traffic is not uniform, it is usually used Multi-level Poisson distribution establishes a model for this type of message flow. A normal distribution can also be used to model the information provided-< information flow. This type of message flow can be referred to as normal message flow ( normal traffic). Finally, the capacity of each path of the network can be configured according to the actual message traffic determined by the measurement. The transmission needs of multiple users can be combined into a collective message flow. For example, several users can be simultaneously Send messages from Dallas to Stokehlm. At this time, these multiple transmissions are not managed individually, but more efficiently via a broadband The line combines these messages and transmits them as a group. As mentioned above, a virtual channel is a dynamically assignable end-to-end connection. A virtual path is a logical structure that can treat multiple virtual channels as Process and exchange together as a single unit. This unified exchange reduces the overall processing requirements and also speeds up transmission. Because the management of virtual paths is simpler and more efficient than the management of virtual channels or individual physical circuits , So this technology can greatly improve the use of network resources. The basic model considered by the virtual path capacity allocation model is-a connection-oriented network that operates under a fixed route. If we define the physical network as including- Group] arbitrary connection < link, then each virtual path (VP) or path r is an ordered list, and the elements of the ordered list are a subset of _j. The routing matrix can be used to define it especially Each 319944 Α7 Β7 V. Description of the invention (2 Relationship between virtual path and physical link. The elements of the routing matrix are:

X 1 if 〇 otherwise (Equation 1) FIG. 8 shows the topological relationship between each virtual path and the physical path. In FIG. 8: The virtual path VP1 includes physical links and the virtual path Vp packet

With physical link P

And P '= the quantitative relationship between the capacity allocated to each VP and the corresponding capacity allocated on each physical link can be dragged. The array form is expressed as follows;

Phys (Equation 2) where A: is the routing matrix of the type defined in Equation 1 above, is the virtual path capacity to Wei, (: Chong ^ is the physical link capacity vector. It represents the capacity that has been assigned to various VPs * The physical link capacity vector Cphys must not exceed the actual capacity available on any physical link. The following simple limit relationship can be used to express this limit: phys c (Equation 3 _) printed by the Employees Consumer Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs Cphys is a description of the actual capacity available on each physical link of K physical links. Because Equation 3 is a vector inequality, it is important to note that the corresponding components of the two vectors must satisfy this inequality. Figure-23- This paper wave scale is applicable to the Chinese National Standard (CNS) 4% grid (210X 297mm)

Five 'invention description (21 A7 B7 8 for example, the routing matrix is especially (Equation 4) The relationship between the volume here each virtual path capacity: the capacity and the capacity assigned to the corresponding physical link capacity are equations 10. "Ί c. 1 11 • C, + rc 1 2 0 1 2 VP .v phys (Equation 5) where is the virtual 'path capacity vector, c. And C1 + c2 α phys minus consumption of the Central Standards Bureau of the Ministry of Economic Affairs The cooperative prints the assigned vectors of each entity's key. The call request procedure to a specific path r can be any stubborn procedure that we know assumes that the occupancy is allocated. The premise is that the procedure has an acceptable all calls and avoids All blocked unlimited, capacity resources. Xr represents the assumption of the occupancy level of unlimited capacity resources, and is usually referred to as "provided information flow" in this technology. The problem of virtual road capacity configuration is in the system of the present invention. And the method is defined -24. This paper scale is applicable to China National Standards (CNS) A4 specifications (21 〇X 297 mm) 312944 A7 ------_ V. Invention description (22) has two goals. No.—must The amount of transmission allocated to each virtual path is optimized, so the transmission cost function is reduced as much as possible. Second, for each physical link, the efforts of the various virtual paths allocated through this link must not exceed the physical chain The physical transmission capacity limit of the path. It can be used in the range [0, cphys] as a continuous function of any real number value as an approximate function of the actual capacity that can be assigned to various virtual paths. Therefore, resource optimization does not require discrete Optimization and its accompanying complexity. In the configuration path capacity problem proposed in this application, load sharing among different VPs is not considered. We assume that the traffic provided by each virtual path boundary is In addition, because the network has a fixed route, the selection of each virtual path fixes the route of the provided message. &Quot; Push-down "path capacity configuration technology in the present invention regards the virtual path capacity configuration work as a load The balance problem includes the proper selection of the blocking measurement value of the load, and the best solution is equivalent to the selection of the allocated VP capacity, where Distribute the congestion on each virtual path as evenly as possible. One way to make the congestion distribution uniform is to measure the dispersion of the congestion value on each virtual path, and then try to minimize this printed by the Beigong Consumer Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs Seed decentralization. You can use any minimization algorithm such as the conventional simulated annealing technology to implement the above method. —The relevant method is: first identify the virtual path with the highest blocking value, and then reallocate capacity from other VPs. Reduce the congestion of the virtual path 'until the virtual path is no longer the VP with the largest congestion. This kind of constellation phase is used to optimize the problem of taking a small value_maximum value, and can be analyzed as described below Optimize it. This paper scale applies to China National Counties (CNS) A4 ^ (210X297 company director) 31S944 A7 B7 V. Description of the invention (23)-If we express the blocking of the i-th virtual path as B (vPi), it has the maximum blocking VP is maxCBNPi).), Here is for all 7 ;? Take the maximum value. The maximum value of the congestion measurement on this set of virtual paths defines the objective function (also called cost function) of the path configuration problem of vp configuration. Therefore, the goal of this optimization procedure is to extract the minimum value of the objective function, which is equivalent to: '* ·. Min max Ib (VP.))

1 (Equation Q here defines the minimum value for all feasible configurations. Because this technique involves lowering the maximum blocking value in all VPs considered, this technique is used to solve the optimization problem. The method is called a "push-down" algorithm. This algorithm follows the fact that uniform blocking allocation is equivalent to the unrestricted optimal solution of the νP allocation and placement path capacity problem. Therefore, this optimal The solution is to allocate each capacity to every two Vp, so that the blocking on each VP within the -error limit is the same. However, this optimal life is not necessarily achievable because of the capacity limitations of various physical links.— The limited capacity of the physical link must be shared by all VPs that pass through the physical link. Printed by the Employee Consumer Cooperative of the Central Prototype Bureau of the Ministry of Economic Affairs. Figure 9 shows a virtual network configured on a physical network. An algorithm illustrates the steps of the embodiment. The path capacity configuration procedure starts at step (902), at which time the connection topology of each vP will be defined. Each vp is combined into a vp path. Capacity configuration group Then in step (9〇3), the VPs are aggregated according to the order in which each physical link passes. Then in step (904), the initial transmission capacity is allocated. In step (9〇5), the resistance scale is selected for China National ΰ " quasi " TcNS) specifications (21 ^ 7 ^ mm) A7 B7 V. Description of invention (the target number for the reduction of 24 plugs. For τ π 令 _, a, the horse is 5 and the target-we must first Choice-Congestion Measurement. In the preferred embodiment of the present invention, the average message volume ratio function detailed in the next section will be used as the congestion measurement. The target value is used to set the path capacity configuration calculation The termination condition of the method. In step (90.6), the blocking on each VP is determined for each physical link. If the vp passing through the single-sin link does not encounter the same level of blocking, then In step (907), modify the capacity currently allocated to each vp in order to balance the vp < blocking values within the -error limit. By allocating the actual capacity that has not been allocated, or the capacity already allocated by the VP with lower productivity Reallocate to more productive VPs Added to each VP. This type of capacity re-adjustment is performed without violating the capacity limit of any real money path. This procedure can identify one or more physical links in step (908) in this optimization procedure The bottleneck. The physical link that has the highest VP congestion and cannot be reallocated with capacity to reduce congestion is called a critical link. Each critical link determines the minimum congestion available on each VP that passes the physical link. The main work of the following algorithm in each stage of the optimization process is to identify the set of key links of a certain set of virtual paths. Once printed in step (908) by the Central Consumers Bureau of the Ministry of Economic Affairs, the employee consumer cooperative prints When a critical link is identified, the actual capacity can be reallocated among the virtual paths passing through the critical link by balancing the blocking value of each virtual path. Please note that when a physical link is found to be a critical link, in fact the physical link does not have unallocated capacity. Therefore, after the algorithm reaches the stage of this configuration path capacity program, it is possible to only reallocate the capacity between each V p through a critical link. -27- The Chinese national standard (CNS) A4% # (210X: 297mm) is applied to each paper size A7 B7 5. Description of invention (25

ίι, READ * Read back attention items, and then fill in.! Install 1 I and then in step (9〇9), remove the VPs of the allocated capacity from the VPs of the group that still need to configure the path capacity. Correspondingly in step (910), the available physical link capacity reduces the allocated capacity of each VP removed in the previous step. The problem of configuring road technology capacity is thus reduced to the problem of minimizing the highest P and the probability of jamming in the VP set. Therefore, a recursive reentry algorithm can be used to implement this procedure. Now use the blocking value obtained in the previous step as the starting value for the remaining configuration path capacity problems. This optimization procedure is repeated recursively in step (911) until all the capacity of each physical link is allocated. In summary, this greedy-% algorithm starts by allocating its path capacity to all V p of the entire group. When the virtual path set that has yet to be configured with path capacity becomes a zero set, the process is terminated at step (912). Algorithm. We emphasize here that when implementing the path capacity allocation algorithm detailed in this article, the broadcast must also follow the sequence shown in Figure 9. Depending on implementation details and convergence considerations, the execution order of some steps of the path capacity allocation algorithm may be different from that shown in FIG. 9. Printed by the Employee Consumer Cooperative of the Ministry of Economic Affairs of the People's Republic of China to identify critical link problems in a specific VP set, it is also clearly a difficult task. At present, there is no known technology that can directly determine the critical link by using the provided message flow and the capacity limit of the physical link. Therefore, the push-down algorithm uses an iterative method to identify critical links. At the beginning of the algorithm, assign a uniform large blocking value to all, and set the starting value of all VPs. The selected initial blocking value must be so large that the sum of the initial allocation values of the capacity of each VC will not exceed the available real link of each physical link

31S944 A7 B7 Fifth, the description of the invention _ (26) international capacity. All the remaining virtual paths in the optimization process slowly and evenly reduce the degree of blocking, and in each phase, the link that first violates the actual capacity limit of the physical link passed is identified as a critical link. Path capacity configuration using average traffic volume congestion measurement In configuring path capacity, the speed and efficiency of the above procedure used to identify critical links at each stage of the process depends to a certain extent on the amount of congestion used in the model. Traditionally, ’㈣blocking measurement (also known as time congestion blocking) has been used to determine the optimal configuration of each capacity in a network. The technique of the present invention that uses the average message volume ratio function as the blocking measurement produces better results than can be obtained using the Oulan blocking measurement. When using the average message volume ratio function, it is only possible to assign a message flow to each other, and in most cases, this calculation is much faster than calculations based on other congestion measurements. We also found that: mainly because the average message volume ratio function is a convex function, it can greatly improve the iterative search for critical links. Before the Mingyi 2 path capacity allocation algorithm using the average message volume ratio function, we first describe the characteristics of the average message volume ratio function to help understand. Using the average message volume ratio function as the blocking measurement Printed by the Beigong Consumer Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs As mentioned above, the choice of blocking measurement is the key to the push-down algorithm. The general, expression of blocking measurement based on the average m ratio function will be derived below, and then applied to an example scenario in which single-level and multi-level Bwason distributions are used to establish the provided message flow model. The average message volume ratio function is known in this technology, and has been used to __-29 I paper scale is applicable to the national standard A4 specification (210 > < 297 public phantom A7 B7 V. Invention description (27 built For the congestion model at the inter-physical link level, please refer to the "Congestion Measurement for Call Permission and Bandwidth Designation for Multi-layer Message Transmission" by JY Hui, for example. This article is published in Internati〇nal..Journal of Digital & Analog

"Cabled Systems" (1990), but this article does not use the average message volume ratio function at the virtual path level or network level. It is used as a congestion measurement to solve the configuration path capacity or planning problems. In addition, the average message volume ratio The function has been used to define a physical link, the concept of "effective capacity". However, please be sure / idea that the path capacity configuration technique using the average message volume ratio function detailed in this article is not limited to follow the Bwason allocation The provided message flow, and the system and method of the present invention are also applicable to any type of provided message flow distribution, including the use of measurement to determine. The saturation blocking probability can be defined as a requirement that the message flow demand exceeds the transmission capacity The chance. The saturation probability is also known as " tail probability. &Quot; because the probability means the probability group at the tail end of the provided message traffic distribution. The practice of this tail probability will be derived below. Knowing the approximation method, that is-Chernoff limit. Suppose X is an randomly assigned random variable, C is a specific value. Then for all positive values. It is clear that the following limits exist: x> .C) = p (e E (e :; x) Printed by the Ministry of Economic Affairs Central Standards Bureau Negative Consumer Cooperative (Equation 7) where P (X> C) is a random variable The probability that the value taken by X is greater than c. The derivation of this limit is based on the Markov inequality. The term ln (E (esx)) means the logarithm of the torque generation function, also known as the cumulant function-30- This paper is also applicable China National Standard (CNS) A4 specification (2! 0X297mm) V. Description of invention (28 A7 B7. The maximum value of the index SC- " (s) for S, that is, with the tightest limit f (Called Chernoff limit). When s = s *, read Zhaoshan and transported to this maximum value. This imaginary s * is the equation, the only positive solution of (s). This grace is easy to prove " '(s) As s increases, increasing g increases and thus guarantees the uniqueness of its root. Because 〆 '(s) 2〇, so " · () increases with the increase of thinking s. This is because the second derivative is equal to ( (After moving) The index of the assigned variation is represented by Ix (C). It is called the average message volume rate function. The average message volume ratio function h_lant function is convex Conjugate transform (― conjugate transform), and can be expressed by the following formula -lnP (X> C) ^ SUsP [sC ~ lnE (esX)) = I {C) (Equation 8) Because lnP (xsC) = 1ηΡ (_χ ^ ), And Inp (_x > _c) = lnp (e_sx ^ se) ', so one can use the following equation to derive a similar formula for the left end of this distribution: -lnP (^ 〇 ^ S3P (-sC-ln'5 (e- ^)) = XJ ~ C) (Equation 9) Printed by the Employee Consumer Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs can use the parameter S to express the average message volume * the right and left tails of the rate function and number as (s)) C (~ s) (Equation 10) Therefore, when changing the parameter s (sign), you can switch from the right tail to the left tail of the average message volume ratio function, and vice versa. -31-This paper scale is applicable to the Chinese National Standard (CNS) 84 specifications (210X297 by the director B7. V. Description of the invention (29 average sentence bwason distributed message flow average message volume ratio function first consider the uniformity of the provided message flow The average message volume ratio function is used to configure the capacity of each virtual path. The characteristic of uniform Poisson distribution of message traffic is: a bandwidth demand parameter (the average call arrival rate r, and the average per call .. duration h ( (Also known as average busy time). Therefore, the message traffic demand P is the product of the average call arrival rate and the average busy time. That is, r * h. The cumuiant function of uniform message traffic can be described by the following relationship: (Equation 11) P ( s) = p (esp-1) Therefore, the allocated capacity of the uniform message flow. c and the average message volume ratio function I can be expressed as follows: C = U / (s) = ppe. (Equation 12) and J ( C) = 5 * (C) -CP (S * (C)). (Equation 13) The Ministry of Economic Affairs, Central Bureau of Standards, negative labor consumption cooperation, Du Yinze, and g labor (C (s) > = sppe sp-p (e sp-1) (Equation 1 4) If the substitution relation (s-1 n C-1 (0 p)) is the solution to equation 丨 2, then it can also uniquely average the average message flow of the type described in equation 14, -32 This paper scale is applicable to China's national standard (CMS) A4 specification (210X 297 PCT) ^ 9944

、 Invention description (30 information volume ratio function is expressed as

The allocated capacity C l (c) = ~ (in c _ p,-1ηίρρ))-ρ-1 · ΡΡ (Equation 1 5) shown in the following formula The average jp represents a number of different H ratio function curves provided. Fig. 11 shows a function diagram of the relationship between the moving parameter and the amount c of the knife distribution valley. Special: As shown in Figure ⑽U, the average message volume ratio function has three significant two:-, the function is a convex function, which is reached when the average number of distributions is ~. Second, when increasing. Value to the level of distribution:,: That is, when C = P), the movement parameter s changes from negative value to positive value. As shown in Figure 11, when c < p, the movement parameter 5 is a negative number; at that time, 5 is a positive number. Third, the movement parameter s is a monotonic function, and increases as the capacity allocated to a virtual path increases. Therefore, the transformation parameter s can be regarded as a movement parameter of probability distribution. When the movement parameter takes a negative value, the probability distribution is shifted to the left compared to when the movement parameter is zero. When the movement parameter takes the positive value ^., The probability value moves to the right. The multi-level Bawasong distribution of information flow rate for the β-industrial consumption and 4 it is printed by Zhongti quasi. The ratio of abundance and average volume of information flow can extend the message flow model to the provided information flow characterized by multi-level Bwason distribution The average message volume ratio function corresponding to this provided message traffic model will be derived below. If the average k-volume measurement of multi-level distribution is used to replace the average-volume measurement of single-level distribution, the method will be generated and then specified by the allocated capacity C-:-33-this paper. The scale is applicable to the Chinese National Standard (CNS) A4 specification (2Ι0Χ ^^^ A7. V. Description of the invention (31 B7: the difficulty of expressing the average message volume ratio. To avoid this problem, describe the movement parameter S to table 7F The average message volume ratio function, and also uses the shift parameter S as a control parameter. As long as the absolute value of this parameter is increased, it can be changed to the assigned capacity f, so that the average message volume measurement value can be clearly increased. We: Let Xi denote the isocentric random message traffic (where i is the 'broadband bandwidth demand P i per class I' average call arrival rate Γ i, and the busy time hi, so that the expected value is :

P, and random variable \ cumulative value (Equation 16)

If (C) multi-level message flow ^ the average message volume ratio function h is the estimated negative logarithm of saturation probability

In >. (Ο (Equation 1 7) Wen read back if.

Printed by the Ministry of Economic Affairs, Central Bureau of Standardization, Employee, Consumer Cooperative, Printed, Converged, etc. where k is the number of message flow levels, and C is the capacity of the message assigned to the VP. The following relationship expresses the general characteristics of the ratio function of the average message volume applicable to all distributions: -34 This paper scale adopts the Chinese National Standard Falcon (CNS) A4 specification (2 ΙΟχ297 mm A7 B7 )

I ΧΛ Ε t X.

= J (Equation 1 8) (Equation 1 9) d Printed by the Beigong Consumer Cooperative of the Central Bureau of Economics of the Ministry of Economic Affairs. Gong She 'We can also prove that it is convex on c. The logarithm of the moment generating function of the evening-level Bwason ’s distribution of message flow is expressed by the following relationship: = = ρ, (θ ^ -1) because the average message volume ratio function has the following general form

Ix {C [s)) = 5 ^ P (sC- So we can use shift per parameter s to express the average message volume ratio function of multi-level message traffic as follows: I ^ x / C (s)), t {spiPie ^ P; eSPj -l (Equation 21) = The allocated capacity C and the shifting parameter S can have a function as shown in the following formula -35-

This paper scale is applicable to the W national standard (CNS) A4 ^ i ^ 7i〇 ^ 29? S A7 B7 V. Description of the invention (33 (Square 22) If multi-level average message volume measurement is used instead of single-level average message volume Test, it can make the problem more complicated in analysis, because it is unpredictable

? Allocation capacity C clearly indicated. Average message volume measurement. This complexity is due to the inability to remove the movement parameter .s from equation 22. However, because Equation 21 is a function of the ratio of the average message volume of the cousin of the movement parameter s, we can change Aya5 without changing c. Therefore, using formula 2 2 ', the capacity value can be calculated in each iteration of the algorithm. Please note that the zero value of s is equivalent to the average, and the zero value of the measurement of the amount of information. By setting s to zero for all VPs, the path capacity allocation algorithm can be started. Average message volume ratio function for normal distribution of message flow Printed by the Staff Consumer Cooperative of the Central Bureau of Standards of the Ministry of Economy The average message volume ratio function is also acceptable. It is used in conjunction with other message flow models that are not based on the message flow distribution provided by the Bwason model. Two important message traffic models will be described below. The first message flow model is based on the ratio of the average message volume to the normal distribution of the provided message flow. The corresponding formula of the average message volume ratio function will be derived for this message flow model. The first message model is based on the average message volume ratio function derived from the actual measurement of message flow, rather than based on explicit assumptions about the distribution of message flow provided. For the meaning, please refer to RS Ellis, "Average Message Volume, Large Deviation, and Statistical Mechanics 9 (Springer-Verlag, 1985). The average message volume of the normal distribution message flow can be determined using the following formula .. Ratio function 36- This paper uses Chinese National Standard (CNS) A4 ^ grid (210X297mm) 31S944 A7 B7 V. Description of invention (34

(Equation 23) where m is the average number of normal distribution N (...), & the number of variations. In addition, s (C) c (Equation 24) The following simple formulas can be obtained from the above two relations. # 读 财 ^ 表 ㈣The average information volume ratio function of the following simple formula ·· JX (CU)) (Equation 2 5) Printed by industrial and consumer cooperatives. Therefore, in the case of g-state distribution of message flow, it can be proved that the average message makes the ratio function a simple (and convex) quadratic equation. The average message volume ratio function for measuring message traffic will be provided on the network in the future. Various services must be able to provide many more names at present, so it is very likely that the future broadband network will provide the type of message traffic on the network. Great changes. Therefore, all models based on a certain-ideal representation of message flow distribution may no longer be applicable. This is because this method is inherently unable to adjust elastically. The average message volume derived from the self-message test measurement. The ratio function can solve this difficult problem of window flow model estimation. The information flow model we have discussed so far has been defined in Hu w-37- This paper scale is applicable to the Chinese National Standard (CNS) A4 specification (21 OX 297 mm) _________B7 _ V. Invention description (35) Level Time scale. Conversely, the statistics of the message flow model are defined on the time scale of a standard atm unit packet. This call-level time scale can be regarded as an approximation of the unit packet-level time scale. Therefore, one or more parameters describing the fixed bandwidth requirement on the time scale of the call hierarchy can be used to extract the characteristics of the message traffic demand that changes randomly during a call. 0 Recently, it has been suggested that the message traffic at the unit packet level can be utilized The ratio function of measuring the estimated average message volume. See, for example, by NG Duffield et al., "Average Message Volume of AT M Message Flow: A Tool for Estimating Quality of Service Parameters" (Dublin Institute for Advanced Studies, 1994) o Ratio of Message Flow Overload to Average Message Volume The impact of the function · If the average provided message traffic on each physical link is less than the corresponding available actual capacity on the link (that is, if E (Xk) < 0 two), then the average message volume ratio function is converted to the blocking volume The test can work normally. However, in some actual overload conditions, this condition may be reversed. We will now consider the following example based on the uniform Poisson distribution message flow and time congestion congestion measurement (ie, the Oulan congestion formula). Printed by the Employee Consumer Cooperative of the Central Department of Economics of the Ministry of Economic Affairs. Table 2 lists the three allocated capacity values and the corresponding message flow requirements calculated for a fixed congestion value of 0.03 sf. Please note that in the last case, even though the blockage is small, but the flow of information provided is greater than the allocated capacity 3 ____- 38- This paper standard applies to the Chinese National Standard (CNS) 8.4 regulations ^ (210X297 mm ) 5. Description of the invention (the average message flow provided by the allocated capacity of 36 A7 B7 15.45 10 109.29 100 699.00 700 This example shows that when the condition of E (Xk) $ cphys is violated, the average message volume blocking measurement must be extended to This overload condition is covered. Mathematically, this extension can be easily completed. As mentioned above, the average message volume ratio function is a convex function with a minimum value of E (Xk). The left end of the average message volume ratio function Define this overload area (please refer to Figure 10 & n). In this area, the increase of the ratio function of the information volume of the Pacific is equivalent to the decrease of the allocated capacity. Also, each control parameter adopts a negative value. The sign of Becco is easy to extend the push-down algorithm based on the average message volume ratio function to cover the overload area. This extension only requires a small modification of the original path capacity matching algorithm The rest of the questions on the mostly conceptual in nature, that is, how to interpret this (iv) extend the left area ratio function of the average amount of information equivalent to the population of the left tail probability approximation 2: Central Bureau of Ministry of Economic Affairs rub quasi HIGHLAND consumer cooperatives printed

In P (X &lt; C) &gt; Ix (C (-s)) (Equation 2 6), and the left end of the average message volume ratio function can be regarded as wide, τ can be regarded as the utilization of the allocated capacity C ι average message volume Measurement. -39- This paper standard home standard (CNS ^ see grid (21〇 × 297 Gongdong A7 318944 V. Description of the invention (37)) ~~-Because in the initial state, the allocated resources exceed the actual resource capacity, also And E (Xk) is lacking, so it is necessary to reduce the utilization of resources, because the utilization is equivalent to the increase of the average message volume ratio function. The optimization target in the overload area can be analyzed as described below. In order to improve the average message in this area The utilization of the measured value of the allocated H identifies the largest consumer of the resource (that is, ν with the lowest average message volume), and reduces this extreme representative of the resource utilization of the VP. The reduction in the resource consumption of the largest consumer is equivalent to The increase of the ratio function of the average message volume in the overload area. Therefore, this method is equivalent to the formulation of the maximum value and the minimum value of the optimization problem. Note that now it starts with each number value that initially falls in the overload area. Second limit = condition. Ordering us to reuse the average message volume at this time. The even distribution of the measurement value is equivalent to the best use of the available resources (even if the best use is not feasible). Follow the same reason as the right end area to meet the restriction conditions, so that the average message utilization measurement value of each critical link can be even. In addition, when allocating the capacity of all vp passing through the critical link, make use of these average message volumes Values are the same. — Printed by the National Bureau of Standards of the Ministry of Economic Affairs, the Negative Work Consumer Cooperative. As mentioned above, the purpose of optimizing in the area at the right end of the average information volume curve is to make the V ρ assigned to the largest blockage The amount of information blocking measurement value (VP) is increased. This method is equivalent to formulating the maximum value and minimum value of the second optimization problem. Please note that as long as the expression in the left end area is first expressed in the optimization goal, the term "blocking," , Replace the term &quot; use, and replace the term "most resource consuming" with the term "νρ that has the most blockage", and the optimization goal of the left Lu domain can be transformed into the right end area -40- This paper music scale is applicable to the China National Standardization (CNS) A4 ^^ 771〇x2 ^^ Printed by the Ministry of Economic Affairs, Central Bureau of Standardization, Consumers and Workers Cooperatives, and description of invention_ (38 Optimization problem. 最 # The two ratios of the left and right regions of the m ratio 1 ^ number: the target expression will get the same optimization procedure. In both cases, the average message volume ratio function will be increased. Increase the absolute value of the control parameter 5 ' The above goal can be achieved. Knowing that if the load does not exceed the available resources, the hyperactive parameter will be positive, and can be quickly assigned to each of the virtual paths of the Maguba 1,-. End all available actual resources. On the other hand, if the load exceeds the available resources, the movement parameter will be negative. In this case, the allocated capacity must be gradually increased until these capacities reach the limit of the international resources. Through the VP path capacity allocation algorithm that utilizes the average message volume ratio function, we can use the characteristics of the average message volume ratio function detailed above to efficiently solve the VP configuration path capacity problem. As mentioned above, the goal of the vp configuration path capacity problem is to allocate limited actual network resources to a plurality of predetermined νρs under a given message traffic allocation. An example of a ν ρ path capacity allocation algorithm using the average message volume ratio function is shown in FIG. 12. , This program starts with a series of initial steps (1202-12〇6). In step (1202), all vp of the path capacity to be configured are concentrated in the -ν ρ path capacity configuration set. In step (1203), the transmission capacity limit of each physical link in the network is specified. In step (1204), a set of upper limit ι Αχ of the average message volume ratio function is selectively specified. Please note that step impatience (1202-1204) can be performed in any order depending on implementation considerations. In addition, only in reality may there be a truncated right tail -41-This paper scale is applicable to the Chinese National Standard (CNS) Α4 ^ # (210 X 297mm) Please read first · Read the back ▽ South * · 'Item I and then Call to fill &lt; binding ^ 19944 A7 ____________B7 5. The distribution of information flow provided by the invention description (39) (that is, only when p (x> c) is greater than a limited value cMAXu value is zero) Manic ⑽ 4) Chinese phantom mirror. If the provided message traffic distribution has a truncated right tail, the distribution may theoretically be a time scale. Therefore, one or more parameters that describe the fixed demand on the time scale of the call hierarchy can be used to extract the characteristics of the message traffic demand that changes randomly during a call. After the initial step ⑴〇2-'12〇6), this path capacity configuration technique performs a recursive step (12〇7-1215). The recursion technique shown in FIG. 12 shows a two-layer recursion, in which, first, as shown in step (12〇7_121〇), the vp path configuration algorithm repeatedly allocates capacity to the vp path capacity configuration set Each of the VPs until one or more physical links reach full (ie 100%) utilization. A physical link that has fully allocated capacity is called a critical link. Therefore, the secondary effect of step (1207-1210) is to recursively identify each key link. Although the critical link identification process may identify only one physical link as a critical link at each stage of the iterative process, the implemented algorithm can also identify and process more than one physical link at a specific time Critical link. In an embodiment of the present invention, the Employee Consumer Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs uses the function formula of the average message volume ratio function according to the provided message flow model to convert the existing average message volume in step (12〇7) The ratio function estimate value is incremented by a fixed amount to perform the identification of critical links. Examples of such formulas are Equation 15 for uniform Boisson distribution message flow, Equation 2 for multi-level Boisson distribution message flow, and Equations 2 3 and 2 5 for normal distribution message flow ____ -42- This paper scale is applicable to the Chinese National Standard (CNS> 319944 A7 B7 V. Description of the invention (40). Please note that the estimated value of the average message volume ratio function can sometimes be changed in value. If the road control capacity is configured The algorithm exceeds the optimal value and is excessive: When the amount of valley is allocated, the above situation may occur. In step (12〇8), the -v movement parameter s value in the set is configured for the path capacity. Please note that for Corresponding to vp, the moving parameter value represents the slope of the π average message volume graph shown in Figure 10. In step (simplified), each incremental material using the average message volume ratio (iv) is assigned to the incremental capacity of each VP in the path capacity configuration set Depending on implementation considerations, the steps (12〇7 · 12〇9) are executed in an order different from the π shown in Figure 12. Then, in step (12UD), the capacity allocated to each νρ is accumulated for each physical key , In step (1211), the accumulated capacity is compared with the physical link <total capacity. If the actual capacity of the link that has not been allocated is less than a predetermined limit, the link is determined to be a critical link. Central Ministry of Economic Affairs Printed by the Bureau of Consumer Labor Cooperatives If the above comparison results in the identification of a physical link as a critical link, the calculation procedure proceeds to step (1212). If no physical link is found to be a critical link, it is repeated Step (12〇7_121〇) or its equivalent steps until a critical link is found. In the rare case where the provided message traffic model has a truncated right tail, this iterative procedure may sometimes fail to identify any More critical links. In this case, when the average message volume ratio function reaches the maximum value I Μ Α χ specified in step (1204), this calculation procedure will automatically terminate. After identifying a critical link, As shown in steps (1212-1213), this V Ρ road control capacity configuration algorithm outputs the calculation results and re-formulates the problem. Whenever one or more entities are included in step (1211) Road identification is off-43-M's Zhang scale is applicable to China National Standard (CNS) Α4 specification (210Χ297 mm) S18944 V. Description of invention (41 A7 B7 ^ m Path capacity configuration algorithm is entered into the step (1212) ^ output ,, detailed listing the current allocation of each vp through a critical link. Then, in step (1213), remove all VPs from the critical link of Tongxian from the route #capacity configuration set. If there is no V p If path capacity is still to be configured, the path capacity allocation algorithm terminates at step (1216). Printed by the Employee Consumer Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs. If there are still one or more VPs waiting to configure path capacity, then at step (1215) Redefine the set of path capacity configurations to include only such VPs. Because each of the critical links has been removed from the path capacity configuration set and because these removed VPs use a portion of the physical link capacity, the work of configuring the two-path valley volume is simplified into a pair of each remaining in the path capacity configuration set VP allocates the physical link capacity that has not been allocated. In step (1215), the available valley of each physical link is deducted and allocated to the removed capacity corresponding to each key link found in step (d 1) last time. In an alternative embodiment of the invention, freezing the average message volume ratio function value of each VP that has been removed from the V p path capacity configuration π Ο can also achieve the same effect. Because the capacity allocated to the removed V P is It is generated in step (1212), so it is easy to perform this calculation. After reformulating the problem in step (1215), the algorithm returns to step (1207), and as described above, the pair remains in the path capacity All ν ρ of the configuration set will increase the average message volume ratio function by a fixed amount. VP path capacity configuration using adaptive search technology will now show a way to increase the computational efficiency of the basic ν ρ path capacity configuration algorithm to use average messages The volume ratio function is used as a method above the efficiency that can be obtained by blocking measurement. This method is based on the average message volume ratio function -44- This paper scale is applicable to the Chinese National Standard (CNS ) ΑΊ specification (210Χ297 mm) A7 B7 V. Description of invention (42 general characteristics, and can be applied to any type of message flow demand model, including models based on actual message flow measurement.

Using an iterative linear approximation method similar to Newton-Raphson technology, and using the convex function characteristic of the average message volume rate function, the convergence speed of the basic VP path capacity allocation algorithm can be greatly accelerated "when the message traffic of a virtual network When the demand disturbance is small, a linear approximation method can be used to improve the adaptive VP path capacity allocation algorithm. Subscription Virtual network may be disturbed sometimes. Such disturbances may be caused by changes in spatial message traffic patterns, changes in message traffic distribution, or small changes in network operating parameters. For example, when multiple virtual networks are first defined on the top layer of a single physical network, and then a small amount of actual resources are redistributed among different virtual networks, such changes may occur. The "movable boundary" problem is a special case of the network configuration path capacity problem, which will be described in detail below. The Ministry of Economic Affairs Central Standards Bureau's Beigong Consumer Cooperation Du Printed Adaptive VP Path Capacity Configuration Technology is based on The vp path capacity configuration technology detailed in conjunction with FIGS. 9 and 12 provides greater advantages over conventional technologies in the final stage of configuring path capacity. In order to understand some of the key elements of this adaptive technology, we It is assumed that the basic algorithm is used to obtain an approximate optimal state. In this approximate optimal state, the capacity allocated to each virtual path is close to the upper limit of the actual capacity of the critical link. In this case, the convexity is not used. The repeated trial and error procedure shown in 2 to find out the status of each critical link can be fully utilized, and it is not necessary to equalize the load of each VP passing through the critical link, but the following general approximation method can be used to greatly Accelerated basic path capacity cylinder placement algorithm: -45 This paper size is applicable to China National Standard (CNS) A4 specification (210X 297mm siS944 at --- B7 V. Invention Ming (43

AI (C) = a C

SAC (Equation 2 7) This approximation system follows the following easy-to-derivable, one-to-one relationship of the average message volume ratio function: ’.

dl (C) dC (Equation 2 8) Then we must calculate the average increment of the average message volume ratio function for the link] 4 in step t: [〆 'so that the unallocated capacity of the critical link k can be allocated to All VP-through this link. We show the incremental capacity allocated to the i-th VP. Because in each iteration step {中 八 ^ = △ Ik, = ί = ΔΙ / Σ — S 1 J -J \ Si), the following formula will be followed: ΔΤ AC, Σ s (Equation 29) Central Standards of the Ministry of Economic Affairs The relationship between each △ i and the corresponding △ s calculated by the bureau employee consumer cooperative printing equation 2 9 is as follows: MC (5)) «AS —dJ (j (S)) = AS ^ L- ^ 1 £ I = S do a ⑶ Λ… V (sl OS dC ds dsz s Therefore, as long as the differential formula ϋ $.) = S replaces C = &quot; (S), and applies ν (-46- this paper's scale to Chinese countries Standard (CNS) Α4 specification (210 X 297 mm) A7 A7, then we can get △ s 5. Invention description (44 = d2 μ (S) ^ ds2 is defined as the variation after distribution, and △ I The relationship is as follows: SV (S) (Equation 30) The adaptive VP path capacity allocation algorithm requires the instantaneous slope s of the average message volume ratio function to be non-zero, and you can use the t0000 program 29 to estimate the average message volume ratio for a larger number of values The function increments ^ Ik. 'Is the ratio of the average message volume. The function is convex on c, and c increases monotonously with the increase, so this calculation technique will be underestimated. Therefore, if you repeat the execution With this procedure, the procedure will converge steadily. Another example of a vp path capacity configuration algorithm that uses the average message volume ratio function as the congestion measurement is shown in Figure 13. The basic VP channel 4 valley configuration shown in Figure 12 The trial-and-error technique used in the sub-algorithm is different. This path capacity allocation algorithm uses an adaptive search technique somewhat similar to the Newton_Raphs technology. The basic vp path capacity allocation algorithm shown in Figure 12 is shown in Figure 13. The adaptive VP path capacity configuration algorithm also begins with a series of initial steps (1302 — 1304). In step (13〇2), each vp of the path capacity to be configured is concentrated in the _VP path capacity configuration set. In step (13〇3), the transmission capacity limit of each physical link in the network is specified. The basic average message volume VP path capacity configuration algorithm shown in FIG. 12 is executed in step (12〇4). You can also optionally specify the average message volume ratio letter at this stage _____- 47- This paper standard is applicable to China's National Standards (CNS) Α4 specifications (2 丨 0Χ297mm) Please read the back: Wild · * Things 1 ^ 1 S 1 ▲ page binding printed by the Ministry of Economic Affairs, Central Bureau of Accreditation, Workers and Consumers Cooperative η: description of invention (a A7 B7, the upper limit of the number of employees ’cooperatives printed by the Central Bureau of Standards, Ministry of Economy, ιΜΑχ. As mentioned above, oral The right end of the mention: there may actually be a heart ⑽ quantity that can intercept the enemy. It is only stipulated that please note that the visual test (⑽) and ⑴〇3). As mentioned above :: While the first step is executed sequentially by mjL ^ r MQ, only if the provided message flow distribution has a possible opportunity to truncate the right tail T. It is only 4 hours and 20+ hours that it is specified in step (12〇4) that MAX is the same as the basic average message volume. VP Road Lulu County * Class. The VP Road Valley Volume Configuration Algorithm in Li. This initial procedure includes: —, Κ ^ &gt; τ, the information flow model as supplied; the corresponding relational expression for deriving the average message volume ratio function I to the flood barrier, double 1, the allocated capacity, and the movement parameter S. These steps are not shown in Figure 13. Other initial steps include: In step (⑽), select the starting value of the set of average message volume (i) function estimation value is small and not zero, and the converted parameter S (corresponding to non-zero value. As mentioned above These initial estimated values are equivalent to larger initial blocking values. _ Like the basic average message volume νρ path capacity allocation algorithm, this path capacity allocation algorithm allocates capacity repeatedly as shown in steps (13G5 _ 1311) until One or more physical links reach full utilization. When it is determined in step (1312) that a physical link is a _critical link, as shown in step (1313 • 1316), remove the path from the path capacity configuration set. All VPs of critical links, and re-formulation of the configuration path capacity problem. Figure 13 Key link identification of 7F adaptive VP path capacity configuration algorithm. The sequence is slightly different from the basic algorithm shown in Figure 12 When using the relations derived previously, the capacity allocation% of each VP is calculated for each physical link in step (1305). In step (13〇6), these starting points are accumulated for each physical link Capacity allocation , And get each physical link. In the step reading · reading back · face ·, * write the page after the matters I binding 48- This paper size is applicable to China National Standard (CNS) A4 specifications (210X297 mm) A7 B7 five 3. Description of the invention (in milk step (1.07) 'compare this sum with the transmission capacity specified in step _ (i3G3)' in order to determine the unallocated capacity of each physical link. If the initial allocation makes one or more entities There is an overload on the link as in the basic model. The subsequent calculation steps need to be modified in order to use the formula for the movement parameter s based on the negative value. ′ Using equation 29, and calculating the dynamic parameters for all vp leaves passing through each physical link k The sum can be calculated in step () for the entity-calculate the average amount of incremental information in the t-th step of the iterative process △ L t. Addicted, ancient 咅 · dry Cheng Juxinkou tk clearly note that each The incremental value calculated by the physical link may be different. In step (13.9), the average message volume ratio function of each physical link is compared to estimate the incremental value, and the lowest value is selected. Step ⑽ Μ Use the following update equation to increase the existing value = Ratio letter Estimated Zhi It:.., &Quot; (Equation 31) * Please read the read back 'side of the note · ·

Item Zai / &gt;! Fill in% to break the clothes I Ordered by the Ministry of Economic Affairs Central Bureau of Accuracy Bureau of Consumer Labor Cooperative Printed from 1 Ϊ: ^ Step (13 11) using the specific C_lant function to provide a specific amount of information for the K function &quot; (1) Modify the motion parameter estimation U step (1311) to update the s value using the following relationship:

As. (Equation 32) where S j is lucky when his father is small, it is possible to calculate the transformation parameters of -49-297 mm by Nan Nan tian s _L · 〇Λ Λ 10000 Formula 30) A7 B7 318944 V. Description of invention ( 47 Change and repeat the repeated steps (1305-1311) 'until the unallocated capacity of at least one remaining physical seam decreases below the preset error or termination condition. When such condition is determined in step (1312), it has been found Out of one or more real links are critical links. After detecting a customs link, the operation of the adaptive vp path capacity configuration algorithm is the same as the basic algorithm, and the calculation procedure enters the step (13 13 ). If no physical link is found to be a critical link, repeat the steps (1j05-1311) or its equivalent steps until a critical link is found. The provided message traffic model has a cut In the rare situation at the right end, this iterative procedure may sometimes fail any other critical links. In this case, when the average message volume ratio function reaches a preset maximum value of 1 MAX, the calculation procedure is automatically terminated. whenever When one or more physical links are identified as key links in step (1312), the vp path capacity configuration algorithm enters step (1313) to generate output, detailing the current status of each vp passing through a key link In step (1314), the allocated capacity removes those VPs that pass through the critical link from the path capacity configuration set. If there is no vp in step (1315), wait for the path printed by the consumer labor cooperative of the Central Bureau of Standards of the Ministry of Economy Capacity configuration, the path capacity configuration algorithm ends at step (1317). If there are still one or more VPs waiting for path capacity configuration, then redefine the path capacity configuration set in step (b 16) so that only such types are included Because each node passing through the critical link has been removed from the path capacity configuration set, and because these removed VPs are allocated to part of the physical link capacity, the standard scale (CNS) Ege (g 50 X 297 mm ) A7 B7 V. Description of the invention _ (48) The path capacity configuration work is simplified to allocate the unallocated physical link capacity to each VP that remains in the path capacity configuration set. In step (1316) , The available capacity of each physical link is reduced to the capacity corresponding to the key link found in step (1312) that has been removed last time, and the above goal can be completed. Restart in step (13 16) After formulating the problem, the adaptive vp path & valley configuration calculation; the sound returns to step (1305), and as described above, the capacity allocation is performed for all VPs that remain in the path capacity configuration set. Refer to Figure 14 And 15, you can intuitively understand the operation of this adaptive vp path capacity allocation algorithm. Figure 14 shows that even though different νρs pass through a common physical link, the characteristics of these VPs may be different Average message capacity curve ... As shown in Figure 14, the average information capacity curve of different v p has its minimum value at different allocated capacity values. This result is due to the fact that the message traffic provided on these different VPs is different. Therefore, each VP has its smallest non-average message volume on different allocated capacity values. For the situation of two virtual paths VPi and VP2 passing through the same physical link, the capacity allocation procedure of this adaptive vpt path capacity configuration algorithm is shown in Fig. 15. At the beginning of this private sequence, the average message volume ratio function. The starting value of a group of small numbers I v t rather than zero is selected. The corresponding capacities allocated to V p i and V p 2 are C i and C 2, respectively. Then draw tangent lines on the bases k (c 1 ′ I k t) and (c 21, I k t) of the average information capacity curves of V p i and V p 2 respectively. The symbols s / and s / in Figure J5 indicate these tangents, and these tangents correspond to the existing iterative estimates of the movement parameters of these virtual paths. The unallocated capacity at the beginning of the t-th iteration △ ckt is equal to the total capacity and has been ________- 51-The paper scale is applicable to the Chinese National Standard (CNS) A4 specification (210X297 mm). Please read first. ·, Winter I it, fill out this page, binding, printed by the Ministry of Economy, Central Bureau of Standards, Employee's Consumer Cooperative. V. Description of invention (49 A7 B7 distribution capacity ΓΓ and the difference between c / sum. And s2t, it can be profitable. The incremental increment of the known slope st is as follows: The average alum heart rate ratio function estimates ΔΧ?

C

Phys cc + — + -IL · Ren st 1 -2. (Equation 3 2 A) The Ministry of Economic Affairs, Central Standards Bureau, Beigong Consumer Cooperative Printed Model Penalties' 1 1) In the case of Qiqi, it uses the -specific way; The estimated value of the specific chain k without the average message volume ratio function then identifies the minimum value of the incremental increment Δν of the second physical link, and the minimum value increasing average message volume ratio function estimate value. Of course, the corpse has the ratio of the average message volume of the physical link. The function is to increase the same volume. The corresponding incremental capacity allocations for V Ρ 2 are a 'c · 11 and △ c t. / Then the average message volume ratio. The function estimates the value of γ by 2 and increases the value of δι ^, and gets 1k. .Sorry / idea, we do not calculate a different V + 1 for each physical link, and use △. At first, it is the co-incremental increment of the previously estimated value of the average message volume ratio function. The inverse relationship of the average message volume ratio function is used to determine the capacity c / "and c / q to be allocated to the virtual VPI &amp; Vp2, and at the same time determine the new slope_Sit + 1 &amp; S2t + 1. Then repeat This iterative step, in order to get the next increment of the average message volume ratio function estimate value ΔΙ / -1. This calculation step is stable 'because △ I kt is underestimated. This is because if you ask II READ-ί Reading back "I iS ··-I 'ί I iΙΑ Fill in I to write a book

K 52 This paper scale is applicable to the Chinese National Standard (CNS) A4 specification (210X297 mm) A7 B7 5. Description of the invention (50 The incremental increment of the allocated capacity shown in Figure 15 is Δ (: ι4Δ (ν are less than each Calculate △ in the iterative steps; [/ values used AC ”and △ C, *. Use the average message volume ratio function to derive other operating parameters 2 In some environments, network operators may wish to specify such as blocking of allocated resources It is different from the network parameters of the average message measurement. At the call level, on the scale, the Bwason distribution provides the blocking performance specification of the message traffic. The most commonly used measurement is the amount of time congestion based on the Oulan blocking formula The Eulerian blocking formula E (p, C) expresses the characteristics of blocking as a function of two parameters, which are the load p and the allocated capacity c. Because a path measured based on the average message volume has been found The capacity allocation technology is much faster than other path capacity allocation methods, so the most efficient way to calculate the final value of the time congestion measurement is to use the average message volume measurement to obtain the best value of the allocated capacity, and then The best value of these capacities and the known value of the message traffic load are used to calculate the Oulan congestion value. The same method can be used for time congestion and congestion measurement of multi-level (ie non-uniform) Bwason distribution message traffic Other congestion measurements. If other congestion measurements are used to configure the path capacity, a complex and inefficient path capacity configuration algorithm can be obtained. The average message volume ratio function can be used as the basic VP path capacity configuration for the congestion measurement The algorithm or the adaptive VP path capacity configuration algorithm is used instead. The comparative results using the average message volume ratio function can be more easily understood by considering the following examples. Figure 16 shows one with six An example of a physical network of nodes (1601-1606). As shown in Figure 16, the node (1601) is composed of links (1611) and (1616) -53- wooden paper scale applicable to the Chinese National Standard (CNS) Α4 Specification (210X297mm) Please fill in the i. Δ *. * I

IT, -Ά Printed 3i S944 Α7 Β7 by the employee consumer cooperative of the Central Bureau of Standards of the Ministry of Economy V. Description of the invention (51) and connected to the nodes (1602) and (1606), and the nodes (1604) are via links (1613) And (1614) and connected to the nodes (1603) and (1606). Nodes (1602) and (1605) are connected to nodes (^ 03) and (16〇6) via links (1612), (1615), (1617), and (1618). It is assumed that these links of the physical network can carry two-way messages. However, it is assumed that each virtual path defined on this physical network is unidirectional. ... Table 3 described below defines ten different virtual paths V P i to V P i on the physical network of this example of six nodes. Therefore, such as the virtual path V P 3 is from the node (1602) to the node (16〇5) via the node (1603). The connection topology of these virtual paths is shown in Figure 17. The nodes (170 1-1706) in Figure 17 are equivalent to the nodes (1601-1606) in Figure 16. The virtual path v P3 shown is from node (1702) to node (1705) via node (1703). In the considered example, it is assumed that the transmission capacity of each physical link of the eight physical links (1611 1618) shown in FIG. 16 is a transmission capacity of 50 units. The unit transmission capacity per physical link can be arbitrarily selected, for example, 20 million bits per second. VPi VP2 VP3 vp4 VP5 vp6 VP7 1-2-3-4 1-6-5-4 2-3-5 2-6-5 6-2-3 6-5-3 2-3-4 Central Ministry of Economic Affairs The quasi-bureau employee consumer cooperative printed

6-5-4 First consider the situation when the provided message flow is uniform. The first column of Table 4 shows the traffic requirements of the ten virtual paths (VPl-VPl〇). As shown in Figure 1, this means that the message traffic demand is evenly distributed between each physical link / situation, because the change in message traffic demand between different virtual paths is more than -54-the paper standard applies the Chinese National Standard (CNS) A4 specification (210X297 Variant A7 B7 V. Description of invention (52) More consecutive results, not a dual-modal or multi-modal situation. The path capacity configuration results obtained by three different algorithms are listed in Table 5. Table 4 VPi VP2 VP3 VP4 VP5 vp6 VP7 VP8 VP9 VPi〇 Message flow requirement 4 6 3 1 6 4 3 6 5 2 Virtual path capacity algorithm I 9.89 13.64 7.99 5.10 12.86 10.44 7.99 13.64 11.24 7.16 algorithm II 9.75 13.53 8.19 5.29 12.64 10.50 8.19 13.53 11.18 7.11 algorithm III 9.72 13.83 8.07 4.79 12.82 10.59 8.07 13.83 11.30 6.96 Ouland blocking probability (%) Algorithm I 0.623 0.330 0.863 0.443 0.607 0.388 0.863 0.330 0.715 0.350 Algorithm II 0.743 0.366 0.743 0.366 0.743 0.366 0.743 0.366 0.743 0.366 Algorithm III 0.768 0.275 0.818 0.890 0.635 0.338 0.818 0.275 0.689 0.439 Read · read the back, if · 3 notes fill in I write Printed Sheet Sheet Printed by the Central Standard Falcon Bureau Employee Consumer Cooperative of the Ministry of Economic Affairs 5 Link Cl-2 c2 -6 Ci-6 C2-3-5 C3 -5 C3 -4 C5 -4 Link Capacity 50 50 50 50 50 50 50 50 Bond lifetime 1 utilization rate (%) Algorithm I 42.28 35.92 41.60 100 100 36.96 35.78 54.56 Algorithm II 41.86 35.88 41.28 100 100 37.40 35.88 54.14 Algorithm III 42.04 35.22 41.58 100 100 37.31 35.58 55.32 -55- L Applicable to the Chinese National Standard (CNS) A4 specification (210X297 mm) A7 B7 319944 ----- —__ V. Description of invention (53 Subalgorithm I is a convex function programming type algorithm, its goal is to make use of the European L: The total flow of carrying messages calculated by the plug formula is the largest. The type of this algorithm is the type described in the conference paper "Quality Bandwidth Management of the Best Virtual Path" by A. Farago, s. Blaabjerg, w. Holender, τ · Henk, and LMalomsoky: The paper It was sent to US,% in May 1994, and the present invention is hereby incorporated by reference. The calculation is the same-using the push-down algorithm described above for the European blockage measurement. Algorithm ΙΠ is an average message volume VP path capacity allocation algorithm also mentioned in the previous section. In order to make the results of these three algorithms comparable, the bandwidth demand parameter ρ of all three algorithms is set to have a unit value. In addition, it is also assumed that the message traffic demand A value is lower than the actual capacity of each link. Table 4 also Shows the blocking probability obtained by the three algorithms using the Oulan blocking formula under the allocated capacities calculated by the algorithm ΠΙ. After comparing the path capacity configuration results of algorithms II and III, we can see: ^ The path capacity configuration result of Algorithm II (the push-down algorithm using Oulan's congestion measurement) is close to the path capacity configuration obtained using Algorithm 111 (adaptive path capacity configuration algorithm using the average message volume ratio function) Results. Printed by the Beigong Consumer Cooperative of the Central Bureau of Economic Affairs of the Ministry of Economics. Therefore, the configuration path capacity issue is obviously not sensitive to the choice of congestion measurement. Table 5 7F shows the corresponding results of the link utilization of each virtual path. Table 5 It can be seen that the c2_3 &amp; C6 links corresponding to the physical links (1612) and (1615) are critical key links because the utilization rate of these physical links is 1QQ%. Table 6 shows The overall performance of the two algorithms with characteristics and comparable to the three algorithms -56-The paper method scale is applicable to the Yin Guo National Standard (CNS) A4 specification (210X297 mm) Ά19944 A7 V. Description of the invention (54) Parameters: Maximum VP blocking Machine 4, and the total traffic carrying urgent messages. Table 6 Comparison Algorithm I Algorithm II Algorithm III Maximum VP Blocking Probability 0.8639 0.7437 〇8901 Total Carrying Message Traffic 39.7843 39.7748 39.7834 Also corresponding to the demand for message traffic between virtual paths Scattered message traffic patterns. Investigate the effectiveness of the same three algorithms. Table 7 shows a comparison table of the allocated V p capacity that can cause low blocking for uneven message traffic requirements. As shown in the first example of Table 7, the virtual The message flow provided on the path νρ ^ is very different from the message flow provided on the virtual paths VP2 to VP1〇. Table 7 VP! VP2 VP3 VP4 VP5 VP6 VP7 vp8 VP9 VPi〇 Message flow requirements 12 1 1 1 2 2 1 2 1 2 Virtual road rendering volume algorithm I 23.80 5.73 5.73 5.73 8.00 8.00 5.73 8.00 5.73 8.00 algorithm II 23.89 8.47 5.98 8.47 8.15 10.99 5.98 10.99 5.98 10 99 Algorithm III 24.81 8.21 5.69 8.21 8.09 11.19 5.69 11.19 5.69 11.19 Printed by the Ministry of Economic Affairs, Central Bureau of Customs and Industry Beigong Consumer Cooperatives. The comparison results of the same three algorithms described above for the link utilization on 10 virtual paths are shown in the table. 8. We can see that when using the algorithm j 工 -57- this paper scale is applicable to the Chinese National Standard (CNS) A4 specification (21〇X 297 mm) A7 ______B7 V. Invention description (55) and ΠΙ to configure the network In path capacity, physical links (〗 612) and (1618) are critical links. The comparison of the overall performance evaluation parameters of these three algorithms is shown in Table 9. The results listed in Tables 7, 8, and 9 show the situation where the non-uniform message traffic demand causes low congestion. Table 8 Table 9 Comparison Algorithm I Algorithm II Algorithm II I Total Carrying Message Traffic (; Σ DEMAND = 25) 24.971 24.985 24.986 Link Cl-2 ^ 2-6 Ci-6-3 ^ 6-5 C3-5 C3 -4 C5-4 Link capacity 50 50 50 50 50 50 50 50 Link life rate (%) Algorithm I 59.1 27.5 27.5 98 71 27.5 59.1 27.5 Algorithm II 59.8 38.9 100 100 34.0 59.8 38.9 Algorithm Π 61.0 32.6 38.8 100 100 33.8 61.0 38.8 Ministry of Economic Affairs, Central Falcon Bureau Beigong Consumer Cooperative Society. On the contrary, Tables 10, 11, 1 and 12 show the situation of high blocking caused by uneven provision of information traffic. Table 10 shows the capacity allocated to the message flow requirements shown in the first column. Table 11 shows the corresponding link utilization results for 10 virtual paths. Table 12 summarizes the comparative analysis of the three algorithms by using the total message traffic as a parameter for overall performance evaluation. ______- 58--This paper scale is applicable to China National Standard (CNS) A4 specification (210X 2 mm) A7 B7 V. Invention description (56 κ —— 1 VP9 VPi〇VPl VP2 VP3 VP4 VP5 VP6 VP7 VP8 message flow Requirement 16 2 3 20 19 2 4 2 1 2 Virtual way 4 to Rongo algorithm I 18.00 5.25 3.99 28.97 21.02 5.25 4.99 5.25 1.99 5.25 algorithm II 16.98 5.65 4.75 27.37 19.71 5.65 5.73 5.65 2.82 5.65 algorithm III 18.11 5.09 3.94 28.69 21.30 5.09 5.08 5.09 1.56 5.09 Please pay attention to please fill out the table on the clothing page 1 1 Link C Bu 2 -6 C Bu 6-3 ^ 6-5 C3 -5 C3-4 C5 · 4 Link capacity 50 50 50 50 50 50 50 50 Link utilization rate (%) Algorithm I 40.0 100 21.0 100 100 18.5 46.0 21.0 Algorithm II 39.6 94.2 22.6 100 100 20.8 45.4 22.6 Algorithm III 39.4 100 20.4 100 98.1 18.1 46.4 20.4 Schedule 12 Economy Printed comparison algorithm of the Central Bureau of Standards of the Ministry of Labor and Economic Cooperation I algorithm II algorithm m Total message flow (Σ DEMAND = 71) 64.8137 64.2233 64.6251 As can be seen from Tables 6, 9 'and 1 2 as a whole In terms of sexual performance, these three calculations The method obtains quite similar results. -59- This paper scale is applicable to the Chinese National Standard (CNS) Α4 specification (210Χ 297 mm) A7 B7 319944 V. Invention description (57 Virtual network path capacity configuration can be configured with VP path capacity The problem is generalized to a virtual network's path capacity allocation problem 'because the virtual path is a special case of the virtual network. From a practical point of view, this generalization is very important because there are many applications of the virtual network concept For example, the same service can be provided on different virtual networks. Some users may require a dedicated virtual network that can provide guaranteed service quality ((iv) parameters. Other users may require a certain level of transmission resources and security Virtual network, so these virtual networks must be independent of other virtual networks. In order to provide these services, we must solve the problem of virtual network (VN) path capacity allocation. The concept of virtual network can also be applied to the situation where a telecommunications operator transfers the transmission capacity from another operator. The first player may want to allocate the transmission capacity mentioned to its customers in a way that maximizes or maintains service quality standards. In order to achieve the above objective # 1, the first operator will need to configure the path capacity of its leased virtual network. A reseller's customer may be a lower-level reseller. Therefore, for example, a large organization that uses a long-term contract to purchase a large amount of transmission capacity may want to redistribute transmission capacity to each department based on changes in the information flow requirements of each department in the organization. This kind of redistribution is a problem of one V N path capacity configuration. , Another example of the application of the concept of virtual network-a simple example is related to the transmission of Internet protocol via an atm network (Café ^ purchase heart abbreviation ⑺ packet). One way to perform the above work is to allocate a set of VPs on the ATM network to complete the connection-oriented transmission of packets. However, it can also be regarded as: allocating a virtual network instead of only allocating a set of virtual paths, which can be applied to the China National Standards (CNS) Α4 specification (210Χ297 dong) at the virtual -60 M Zhang scale. Please read back面 ί Matters and then fill out this page to be printed by the Ministry of Economic Affairs, Central Bureau of Economics and Technology, Beigong Consumer Cooperative Printed $ Α7 _______ Β7 V. Description of Invention (58) Different packets are sent around the network through different paths. This less-connected transmission mode will provide more versatile gain than connection-oriented technology. Abstract model of VN path capacity configuration problem The description of the entire path capacity configuration problem can be decomposed into a multi-layer representation, which can describe the path capacity configuration problem on each layer in a form independent of the description of other layers. In an embodiment of the present invention, a three-layer model is used to solve the path capacity allocation problem of a virtual network (VN for short). This two-layer model is shown in Figure 18. The lowest layer of this three-layer model is the physical layer, on which physical network topology and physical link resources have been specified. Usually these regulations constitute the limiting conditions for VN path capacity allocation issues. However, the solution technique detailed in the present invention is not limited to the case where the physical network is unchanged, but can also be applied to solve the network planning problem. In the usage of this article, the network planning problem is related to finding the optimal capacity allocation on each physical link of a network under a given message traffic pattern. The Central Falcon Bureau Employee Consumer Cooperative printed the three-layer model shown in Figure 18. The middle layer is the virtual network layer. The virtual network layer describes the topology of one or more virtual networks and virtual links. . Each virtual network is composed of one or more virtual links. Please note that each virtual link on the virtual network layer represents a set of one or more physical links on the physical layer. In the model problem solved by the present invention, it is assumed that each virtual network is different. The top layer of the three-layer model shown in Figure 18 is the virtual path layer. The virtual paths defined on the virtual path layer are described in an abstract way in the virtual networks on the lower layer ----61 -__ This paper scale is applicable to the Chinese National Standard (CNS) A4 specification (210X297 mm) A7 B7 V. Invention Description (Motion on 59. Please note that the virtual paths on the top layer that correspond to different virtual networks in the lower layer are not duplicated in the illustrated model. However, this model can be extended to cover different virtual networks in common. The case of lower-layer resources, or cover the case where the virtual paths corresponding to different virtual networks are integrated into a virtual trunk carrying integrated message traffic. Α Generalize the problem of capacity allocation of Layer 3 VN paths to Layer N. Please note that the three The layer abstraction model is generalized into an N layer abstraction model without limitation, where N is a positive integer. In this case of layer ^, the technology described in the present invention can be configured under the resource limitation of the N_2 layer. The capacity of each link in the layer. In order to understand the generalization of the VN path capacity configuration problem from the third layer to the ν layer, we will now consider a model scenario in which an operator maintains an ATM network. This ATM network can be used It is regarded as the bottom layer of a multi-level hierarchy (that is, &quot; infrastructure ... and assumes that the topology and actual resources of this ATM network are fixed, and thus constitutes the limiting condition for the entire path capacity allocation problem. &Amp; Central Bureau of Economics Printed by employee consumer cooperatives, we can recursively build one or more layers on top of the ATM network layer, "sandwich" layers. The "sandwich" used in this article can be defined as several The integration layer of adjacent layers. Each interlayer can be regarded as a three-layer structure that is the same as this model and used to solve the problem of VN path capacity configuration. Second, the layer under the interlayer is regarded as recursive Formulation of MN: The limited physical infrastructure of the capacity allocation problem. An example with a single mezzanine structure can be used to illustrate this concept. It is assumed that a number of dedicated lines are defined above the AT M network layer (these dedicated lines are equivalent to (Virtual path or logical link). These special lines can now be regarded as the construction of -62- dagger paper scale applicable to the Chinese national standard ^ milk) six 4 ^ grid (2 丨 0 297 mm) A7 B7 V. Description of invention ( 60) A new mezzanine physical infrastructure that does not change over the duration of the mezzanine. These logical links (logical links; LL for short) can then be regarded as the limited resources of the virtual network (VN) layer built on the LL · layer This vn layer is different from the LL layer and is variable, that is, various reconfigurations can be accepted during the duration of the sandwich structure. Users who change at any time may require this reconfiguration. For example,-home has distribution The telecommunications needs of large companies in several departments in various regions may change at any time due to the growth of some departments or the downsizing of other departments. The transmission services are defined on the top layer of each mezzanine. The value of the relationship. The virtual networks of this mezzanine model can represent, for example, interconnected regional networks or wide area networks. Above each virtual network, there can be a layer of Vp corresponding to the integrated message flow path. These mapping VPs will be defined in the chapter that discusses the basic three-tier VN path capacity configuration later in this application. Please note that the logical link layer is not necessary when analyzing path capacity configuration issues. The main reason for still including this layer is to isolate the sandwich structure from the ATM network layer. In this way, the mezzanine operator can perform virtual network reconfiguration without caring for the operator. This method simplifies the management of the network layer of the Indian Consumers' Cooperative of the Central Sample Bureau of the Ministry of Economic Affairs, because the duration of each logical link will be equivalent to the duration of the mezzanine structure. The V Ν path capacity configuration problem can be expressed as follows: The virtual network (VN) path capacity configuration problem solved by the present invention is as follows: we start with a physical network, which has a component physical link (abbreviated physical link; abbreviated as PL) and the specified physical link capacity (paper size 110X297 mm) '--- 3iS944 A7; ________ B7 V. The specified topology indicated by the invention description (61). A 1 &amp; define a number of shared actual resources on this physical network and lack of different virtual networks. File w I ^, you specify the virtual link of each virtual network with the virtual link of the virtual network &amp;

Complex, structure. The VN topology is related to the PN topology, and VP 'is related to the corresponding VNs. Finally, it is also known that each virtual path is provided with (heartbeat traffic. The goal of the VN path capacity configuration problem is to allocate transmission capacity to each without violating the topology or the source system. Segmentation architecture Generally speaking, paths The ❽ shift, ^ effort allocation goal is to divide the physical infrastructure into several segments, and the way of segmentation is different-there is no interaction between the segments (that is, evening industrialization) 'and there is a certain within each segment These interactions (that is, multi-tasking). When the transmission capacity of the whole message is smaller than the sum of the required capacity of each component stream, the rural industrialization gain is helpful. + Considered in the present invention In the model VN path capacity allocation problem, it is assumed that the resource separation between different virtual networks has been completed. Therefore, even though two or more virtual networks share the same actual resources, this path capacity allocation type does not Handle the multiplexing of message streams belonging to different virtual networks. Only allow the actual resources available for multiple message streams within a single virtual network. The Central Sample Bureau of the Ministry of Economic Affairs-Industrial Consumer Cooperatives prints at least two types of virtual The partition architecture of the proposed network path capacity allocation problem conforms to the multiplexing rules of the above model: a_ virtual road fe (vp) partition architecture, where each-each virtual path on the virtual network is from the middle of other virtual paths Isolate it so that there is no multiplexing between any two different virtual paths. B. Virtual link (VL) partition architecture, where each virtual link on each virtual network is from another virtual link Separated to make different virtual -______ B7 V. Description of the invention (62) There is no multiplexing between the proposed links, and different vp through the same virtual link are allowed to share the lower layer resources. C. Physical link (PL) area Separate architecture, where each virtual network is separated from other virtual networks. Here, there is no multiplexing between different virtual networks' and allows virtual links within a single virtual network to share underlying resources From a network management point of view, the VP partition architecture is the least complicated. However, this kind of architecture will make the utilization of available actual resources lower than the other two partition architectures. On the contrary, the PL partition architecture is The most computationally complex, It provides the highest available actual resource utilization rate among the three partitioned architectures. The three partitioned architectures will be described in detail below. First, consider the VN path capacity configuration problem using the VP partitioned architecture. Then, the virtual links are not allowed to share When considering actual resources, consider VN path capacity configuration. This method is equivalent to using the v L partition architecture defined above. Finally, in the case of allowing different virtual links to share actual resources, consider VN path capacity configuration. This method is equivalent to using the PL partition architecture mentioned above. As shown below, the vn path capacity can be reduced to the printing and placement problem of the negative work consumer cooperative of the Central Bureau of Standards of the Ministry of Economy to one for all three partition architectures. Equivalent vp path capacity configuration problem. You can use any of the solutions detailed in other articles to solve this equivalent Vp path capacity configuration problem, and get the resource allocation of each virtual link. Using the _ average message volume VP path capacity allocation algorithm to configure the path capacity of a group of virtual networks using the PL partition architecture, and exemplify the system and method of the present invention. The performance of the algorithm for allocating the average message volume path capacity and other values are applied to the Chinese National Standard (CNS) a ^-(-21〇x29 ^ j 319944 A7. ___——-------- B7 Five 3. Description of the invention (63) '~~~-A comparison of conventional algorithms, that is, ordering the order to estimate the accuracy and speed of the former. The concept of VN path capacity configuration is to deal with sentences and efficiently For the hierarchical description of the VN path capacity configuration problem, we first need to define the model and some notation. A special VN configuration matrix can be used to describe the membership relationship between each virtual seam of the virtual network and each corresponding physical link. This The structural matrix is similar in concept to the routing matrix previously defined in solving the VP path capacity allocation problem. We still remember: routing matrix description-between the virtual paths defined on the basic network and the links of the basic network Membership relationship. The VN path capacity allocation problem decomposed according to hierarchy can be simplified to the VP path capacity allocation problem that will be explained in the following. This simplified procedure is similar to the compiler procedure. The virtual network structure and this structure can be defined of The traffic demand is regarded as a high-level specification. The problem of vp path capacity configuration can be regarded as a low-level specification. The system and method of the program of the present invention compiles a high-level specification into a low-level specification. The Vp structure is called the post-compilation structure, and this similarity is emphasized. Please note that this compiler can be used in conjunction with any kind of VP path capacity configuration algorithm. The Ministry of Economic Affairs Central Standards Bureau employee consumption cooperation du printed virtual network (VN ) The structure matrix describes the topology of each virtual link that constitutes a virtual network. Because each virtual network is a set of virtual links, and because each virtual link in any virtual network is defined on the underlying physical network The abstract concept of each physical link, so the VN structure matrix describes a set of paths on the physical network. "The VN structure matrix contains a number of zeros and ones and describes the group membership. The number of columns in the structure matrix is the same as the number of physical links. The number of lines and the virtual link are based on the Chinese paper hoarding standard (CNS) A4 specification (210X297mm) B7 5. Invention description (64) Each row of the matrix defines a member of this type and represents a single virtual link. ,, and Zhanren Jiakui list this member relationship as all physical links: # 集 &amp; 心 一 subset. Therefore, the row The virtual link VL in the vector column represents the physical link pLp through the sound value, and the zero value represents the VL. It goes through the physical link PLp. 1 and the physical link elements that constitute the corresponding subset are mathematically The way to define each pseudo-link. This subset is the mathematical notation of virtual links. Each virtual network has its own structure matrix to describe all virtual links belonging to the virtual network. Please note that this type of (iv) Predictable materials-virtual paths become the abstract concept of any physical link aggregation (such as subnet). A virtual path is just a special type of subnet. Therefore, the vp path capacity configuration problem is just a special case of the VN path capacity configuration problem. By defining the link capacity vectors of the virtual network layer and the physical network layer, a quantitative relationship between the capacity of each virtual link and the corresponding capacity allocated to each physical link can be obtained. If expressed in matrix notation, the form of the relationship is as follows:

C ck phys (Equation 3 3) Printed by the Consumer Labor Cooperative of the Central Prototype Bureau of the Ministry of Economic Affairs where S k is the structure matrix of the virtual network k, and C k is the definition of the virtual link capacity of the virtual network Capacity vector 'Ckphys is the corresponding vector of the actual capacity of each virtual link allocated to the virtual network k. The cumulative actual capacity allocated to all virtual networks must not exceed the available capacity of the physical network. The above restriction conditions can be expressed as the following restriction relationship: -67- This paper music scale is applicable to the Chinese National Standard (CNS) (210X297mm) A7 B7 V. Invention description (65

i2 (sk-ck) ^ C phy.r (Equation 34) The CphysS printed by the Employee Consumer Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs-a vector describing the available actual capacity, and this is performed for all ^^ virtual networks total. Because this is a vector inequality, the corresponding component chapter of each vector must satisfy the inequality individually. VN path capacity configuration problem using the VP partition architecture When using the VP partition architecture, the Vn path capacity configuration problem becomes: determining the capacity allocation of each virtual path and virtual link, so that the actual network can Road resources for maximum use. The structure of the VP area and partition structure of the VN path capacity configuration problem is the same as that of the standard Vp path capacity configuration problem. Because the virtual link capacity on each virtual network is not known in advance, it is necessary to map a vp path capacity configuration problem specified on the VN layer to a converted vp path capacity configuration problem specified on the physical network layer. · Use a structure matrix Sk that associates the structure of each virtual network with the structure of the underlying physical network, and a routing matrix X k that describes each VP on each virtual network with a virtual link. Complete the above mapping operation. Note that each virtual link can be considered as paths on the physical network, just as each virtual path is considered as paths on the virtual network. This recursive definition can be extended to the path capacity configuration and model of the N-layer network. Because the routing matrix maps the virtual paths to the virtual links, and the VN structure matrix Sk maps the virtual links to the physical routes, the two moments, Sk and ck obviously have some overlapping relationship. However, the Chinese National Standard (CNS) A4 specification (210X 297mm) can be directly applied to the actual -68 paper standard (please read the notes on the back ^ and then fill out this page)

T outfit

, 1T 3iS944 V. Description of the invention (66) A7 B7 body links define virtual paths. 'Suo Tang Ai Yu &lt; overlapping relationship between these two area matrix Boolean product (B〇〇lpr〇duct), expressed as follows: Γ k (Equation 3 5) If two are known to have matching rows and ranks can be normal Matrix multiplication reduces MS 'and all elements of these two matrices are zero or _, then N represents the two moments Boolean product, which is defined as the re-conversion of daisy arrays M and N After the product of the positive matrix, each non-zero element in the resulting matrix is converted back to one. The combined structure matrix Γ k is also known as a superimposed structure matrix (structure matdx), and the combined structure moment p car can be regarded as a designated _-fixed route on the virtual network projected to the physical layer. Using this kind of procedure, the problem of allocating the vp path capacity of a specific virtual network on the VN layer can be mapped to the problem of VP path configuration after conversion on one of the physical network layers. Because all the virtual networks on the virtual network layer compete for the resources allocated on the same physical network, all the superimposed structural matrix rk must be concentrated into a structural matrix: printed by the Concat Central Bureau of Economics and Economic Cooperation of the Concat Ministry of Economic Affairs Γ phys

k = l, K Γ (Equation 3 6) The sequential operation means: Γ phys is a combination of all the ensemble structure matrices ρ k. This matrix can be called the structure matrix after compilation. Physically, the sequence of the moments produces a new matrix containing all the rows of the sequenced matrix. Although) This paper size is applicable to the Chinese National Standard (CNS) A4 specification (210X297 mm) Printed by A7 --------- B7 of the Employees' Consumer Cooperative of the Central Prototype Bureau of the Ministry of Economic Affairs 5. Description of the invention p) This order The order of the links is not important for the path capacity allocation program, but it is still important to save the information about the order of the links in order to calculate the accuracy. The above definition of the compiled structure matrix Γ p ㈣ can be regarded as: describing all the set of mapping paths defined on the physical network layer, and each superimposed structure of six vehicles r A subset. After compilation, the paths of the ν Ν path and volume configuration problems are similar to the virtual paths of the basic V ρ path capacity configuration problems. All subsets of these antipodal paths must be collected, in order to get the full set of all antipodal paths, as obtained in the structure matrix Γ phys after compilation. Because r phys has already been defined, any vp path capacity can be applied with a one-time algorithm to configure the path capacity of each integrated vp and decide to allocate the capacity of each V P defined by Γ phys. Using the routing matrix zk, it is possible to calculate k virtual. The corresponding virtual link capacity ck allocated on each virtual network of the network is as follows: (Equation 3 7) where Vk represents the allocation to each defined on the virtual network k The capacity vector of the path. Equation 37 describes the output of the VN path capacity allocation algorithm. Compiler A three-layer model is shown in Figure 18, in which a virtual network is defined above the physical network, and a number of virtual paths are defined on the virtual network. The routing matrix that associates the VN layer with the vp layer is defined in Equation 1 and the related text 4 *. For one with three links Li, B2, and '\ ^^ layer -70- ^ paper standard General Chinese National Standard (CNS) A4 specifications (210 X 297 mm) —-------- -(Please read the precautions on the back side 4 · Fill in this page), 11

T ^ 19944

V. Description of the invention (in terms of 68, the VN structure matrix has the following form

S 10 0 110 0 11 0 0 1 (Equation 38) Taking the example shown in Figure 18 and + ^, ', the physical matrix is used to directly define the edited structure matrix of each virtual router. Human aw and the superimposed structure matrix that associates each virtual link with each real link, such as 7XT, because in this example there is only one

VN

[10 0] 110 0 1 1 0 0 1 (g) Γ. (Equation 3 9) ------ (If you want to read it, please read the "Notes on &quot; before filling out this page)

, 1T Printed by the Ministry of Economic Affairs Central Bureau of Precision Industry Beigong Consumer Cooperative can regard Wancheng 39 as a representation: The virtual network layer contains three virtual links Li j, L2, and Ls, each virtual link is two entities A collection of links. If the virtual link VLj passes through the physical link pp, the VN structure matrix § column jp and one of the j rows are used. One indicates this situation ^ Each virtual path vh and: 丨 VP2 is three of the four physical links A collection of physical links. If the virtual · 丨 1 path νρ 〖passes the physical link Pp, then this is represented by one of column p | and row i in the structure matrix factory after compilation. | The compiler of the VP partition architecture is outlined in Figure μ. First in step (19〇2) |

Choose one of the top three layers of the N-layer VN path capacity configuration problem, and the boundary I ------- · -71-This paper scale is applicable to the Chinese National Standard (CNS) 84 specifications (21〇297 297 mm ) A7 ______________B7 printed by the Employee Consumer Cooperative of the Central Bureau of Economic Development of the Ministry of Economic Affairs 5. Description of the invention (69) One ''-* Three-tier structure. The lowest layer correspondingly describes the extension and limitation of the physical network, the middle layer describes the virtual network of the path capacity to be configured, and the top layer describes the routing structures (that is, vp) defined on each virtual network. At the same time, the corresponding v N structure matrix and v p routing matrix are also constructed in step (1902). Then, in step (1903), the three-layer problem concept is superimposed into the corresponding two-layer problem. Due to the fact that each path on the top layer of the virtual network (that is, VP) uses the same actual capacity of the physical network, a corresponding overlapping routing matrix can be derived to describe the allocation of each path to the overlapping structure. The actual resources of the mapping path. Then, in step (19〇4), the superposed structure matrix of each antipodal path is sequentially connected into a single compiled structural matrix. The compiled structural matrix describes the topology of each antipodal VP with each physical link. Then in step (19〇5), any v P path valley configuration algorithm can be applied to the post-compilation description of the v N path capacity configuration problem in order to determine the transmission capacity to be allocated to each overlapping v p. Examples of such VP road valley configuration algorithms include (but are not limited to) ·· The push-down algorithm shown in Figure 9, the basic average message volume vp path valley configuration algorithm shown in Figure 12, and the graph The adaptive vp path capacity allocation algorithm shown in i 3. Finally, in step (1906), the VP routing matrix is used to calculate the capacity value of each virtual link to be allocated to each virtual network. Introduction to VL and PL partition architecture: Generally speaking, different virtual links can interact with each other, that is, these virtual links can share the same actual resources. Therefore, the multiplexing of the message traffic allocated to these links is allowed. In the case where the virtual links do not interact with each other, the integrated message traffic of each logical link must be defined. On the other hand, —__- 72- The standard size of this paper is f SI S home standard rate (CNS) Α4 · (21GX297 company director) (please read the precautions on the back and fill in this page). 3iS944 A7 --------- B7 printed by the Bureau ’s Consumer Co-operative Society V. Description of the invention (70) One '' ~ — 'If each virtual link is allowed to interact with each other, it is necessary to define the integration of every physical link 々 Message traffic. In a situation where multiple links between virtual links are not allowed, the virtual network (operation will be similar to a leased physical network. The management of such a network is relatively simple, but this network is allocated to The utilization rate of resources is lower than the utilization rate of the multiplexed network allowed between the virtual links. The VL partition architecture can treat each virtual link without interaction as each vp above the physical layer. The task of valley configuration is to determine the integrated provisioned traffic to be allocated to each virtual link. This integrated provisioned traffic can be obtained by using the routing matrix, because it is provided by the virtual path regulations on each virtual network The amount of information. The next step is to get the overall set of all virtual links and determine the relationship between the overall set and the set of all physical links. This overall set is the virtual link of each virtual network Aggregate. Please note that the structure of the virtual network only affects the integrated message traffic provided, but has nothing to do with the task of configuring the virtual link capacity. From the perspective of the physical layer, each virtual link constitutes a virtual path Therefore, the VN path capacity configuration problem using the partition architecture is also simplified to-vP path capacity configuration problem. Therefore, any v P path capacity configuration algorithm previously discussed can be used. The information provided on the R mapping paths can be used The traffic Xr is aggregated into a path message flow vector X. Similarly, the message flow U j provided on the j virtual links and the river-link message flow vector u can be used. The routing moment barrier can be used to make the path message flow vector The following is related to the message flow vector of this link: -73- This paper scale is applicable to the Chinese country ^ M CNS) A4 specification (210X 297mm) —Clothing --- (please read the t. 面 的 注解 #Fill in (This page)

'1T 1 V. Description of the invention (71 U,

ht: X (Equation 4 0) In the vector ° method, Equation 4 becomes the following formula u = A:

• X (Equation 4 1) The square dj is the capacity allocated to the virtual link j, then &amp;, bj is the pair d or more allocated on the path r. Both will follow the formula, Χ. (Equation 4 2 ) ---------- ^-installation-f 箕 命 闻 靖 背 Φ ''s notes and then fill in this page} In the vector and self method, equation 42 becomes the following: The Standard Bureau's consumer cooperation cooperation has printed the virtual link volume and the average message volume of dj. The equation is expressed as: ,, If the provided message flow U ”is known, the average message volume measurement on the virtual link” can be calculated. Therefore, if it is known to be in the physical network Homepage (CNS) Eight specifications (210 X297 mm 7 ^ 19944 A7 B7 V. Invention description (72 sets of all virtual links defined as virtual paths, you can also use VN paths with VL partition architecture The capacity configuration problem is mapped to the _vp path capacity configuration problem, and __ solves the VN path capacity configuration problem. Comparison of VP and VL partition architecture: If we define-allocated The link resource vector d corresponds to the path resource vector b: then dVL &lt; dvp, where dvp represents the allocated link resource vector of the vp partition architecture, and, represents the allocated resource vector of the VL partition architecture. The inequality reflects the gain obtained by multiplexing the integrated message traffic provided on each virtual link. The PL segmentation structure. 'Printed by the Staff Consumer Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs and then considering the differences belonging to the same virtual network The virtual link can be multiplexed. Now it is necessary to determine the corresponding integrated message flow provided by each physical segment, and the corresponding integral of the set provided by all the segments :: information flow, using the routing matrix Z And the virtual network structure matrix § can perform the above-mentioned &lt; procedure. Each physical partition provides a lack of integrated message traffic defined as a collection of integrated virtual link message traffic. This can be obtained using the .VNg matrix Extra integration; because the VN structure matrix defines the membership relationship between each piece of pseudo link and the corresponding physical network. If the vector is redundant, j = each entity in the physical link With the integrated message traffic provided on the road, and the vector g representing the equivalent capacity allocated to each of the J physical links, the following formula can be obtained using the aforementioned notation: Z. U, (Equation 4 5; ) -75- The wood paper shape degree is suitable for the family standard (CNS) M specification (2 丨 ^ V. Invention description (73 When expressed in the corresponding vector notation, the above formula becomes:

z = s · U

X · X (Equation 46) where the composite matrix s.

Lgi (Equation 4 7) -------- Strong clothing-f ^^ Jing Jing back-face considerations and then fill out this page} Use is defined as &amp; k Θ 尤 k 之 in equation 3 5 The superposition structure matrix "k" can be used to determine the member matrix used to define the physical partition. Therefore, it is proved again that the VN path wheat configuration problem using the PL partition architecture can be simplified to the vp path capacity configuration problem. VL and PL · Comparison of segmentation architectures, 'we will now compare the VL and PL segmentation architectures by way of example to understand the difference between the two methods and show the effect of multi-tasking integrated message traffic on the resulting capacity allocation. Now will Consider the network structure shown in Figure 2 and Table 13-D-°

T &quot; Printed by the Ministry of Economic Affairs, Central Bureau of Preservation and Consumer Cooperatives-76- This paper scale is applicable to the Chinese National Standard (CNS) A4 specification (210 X 297 mm) A7 __B7 V. Description of invention (74) Table 13 Virtual chain Road path 2-5 2-1-6-5 3-6 3-4-5-6 5-6 5-6, -3-5 3-5 2-6 2-6 2-3 2-3 (please #Read the precautions of the reed and fill in this file. Assume that multi-level message traffic with the characteristics shown in Table 14 is provided to all virtual links. Table 14 Effective bandwidth (million bits / second) Call arrival rate (1 / sec) — 1 Busy time (sec) 1 0.25 30 1 0.01 240 1 0.051 90 The capacity of each physical link is assumed to be 15 million megabits per second. The result of the PL partition architecture If we use the PL partition architecture, there are eight physical partitions for the VL path capacity configuration problem. These eight physical partitions correspond to the eight physical partitions shown in FIG. 20 and listed in the-row of Table 15 below. Link (2001-2008). Derivation of the traffic provided on each segment from the information traffic provided on each virtual link? Recording rule CNS) Α4 · (210X297mm) ~~, 1r Central Ladder of Ministry of Economic Affairs Bureau of Consumer Cooperatives HIGHLAND printed B7 V. Description (75) invention integrated message traffic. For example, the integrated message traffic provided to the segment on the physical link P5.6 is 43.47 units. This value is calculated in the following manner: the scalar product of the call arrival rate vector and busy time vector listed in Table 14 is three times, that is, 3 * (0.25 * 30 + 〇.〇1 * 240 + 0.051 * 90). The integrated message traffic provided is three times the scalar product of these two vectors, because there are three virtual links passing through this area g. The average message volume V p path capacity allocation algorithm has been used to solve the example VN path capacity allocation problem using the PL partition architecture, and the results are shown in the following table 15: Table 15 The physical links after the partition are allocated to The capacity of the segment 2-3 23.0 3-4 32.3 4-5 53.2 5-6 48.7 2-6 37.5 3-5 37.5 _ 6-1 53.2 2-1 53.2 Employee Consumer Cooperative of the Central Bureau of Economic Development of the Ministry of Economic Affairs Printed VL Results of the partition architecture If a VL partition architecture is used for the VN path capacity configuration problem, the message traffic provided to each virtual link is the same as described above, that is, 14.49 units (0.25 * 30 + 〇_01 * 24〇 + 0.051 * 90). When the virtual --- 78-__-this paper scale is applicable to the Chinese National Standard (CNS) A4 specification (210X297 Gongdong) ^ 19944 A7-~~ -_ _B7 5. Description of the invention (76) The seam is regarded as a segment At the time, the capacity allocated to different virtual links is listed in the following table. 16 The capacity allocated to virtual links 2-5 16.7 3-6 16.7 2-3 21.4 2-6 69.1 3-5 69.1 6-5 16.7 The routing structure of the virtual network is the same as that shown in Figure 2 and Table 3a. The effect of multiplexing can now be studied. We will consider the example of a physical link ^ ^. In the case of the PL partition architecture, the total capacity allocated to the two virtual links using this physical link is 48.7 units, which is less than the sum of the individually allocated capacity in the VL partition architecture (ie, 16.7 + 16.7 + 16.7). Comparison with conventional VN path capacity allocation algorithm. Printed by the Central Industry Bureau of the Ministry of Economic Affairs

The performance of the V N path capacity allocation technique detailed in this algorithm has been compared with other algorithms. The results shown below are related to a VN path capacity configuration problem that uses a p L partition architecture (also sometimes referred to as a complete partition architecture). Now consider the five virtual networks (21〇1- 2105). The topology of the lower layer physical network is the same as that described above with reference to FIGS. 16 and 17. Suppose __ -79- This paper scale is applicable to the Chinese national standard (€ milk) 4 wash grids (210 \ 297 mm) A7 B7 V. Description of invention (77 shared by the five virtual subnets with the topology shown in FIG. 21 Figure 16 shows the resources of a single physical network for a long time / shown = 'The virtual network (2 丨 03) shown in Figure 21 represents an example of multiplexing between virtual links. The following three virtual links Through the physical link ρ ·. Node 3 via nodes 4 and 5 to node 6 L3 _6, from node 2 via points 5 and 4 to node 4 L2_4_, and L5_6. In the case of PL partition architecture, can be The provided message flows from these three virtual links are integrated into one message flow on the physical link P5 · 6. The path capacity allocation algorithm also corresponds to the two to allocate capacity to the physical link P5_6. "In this example, it is assumed All the physical links have the same capacity of 2400 units. The n-level Poisson distribution message traffic can be provided to these five virtual subnets. In the model problem analyzed in this paper, those with the characteristics shown below Each virtual link and each virtual subnet defines four levels of Urgent traffic demand. Assume that the message traffic demand pi of level i (i. Is an integer value from! •) is expressed by the following functional relationship: Applicable to a 3-node virtual subnet-_ 碕 读 背 Binding P, · 1 14000 (Equation 4 8) Printed by the Consumer Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs for a 4-node subnet__

Pi 1 7000 (Equation 4 9) -80- This paper scale is applicable to the Chinese National Standard (CNS) A4 specification (210X297 mm) Printed by the Ministry of Economic Affairs Central Standards Bureau Staff Consumer Cooperative A7 &quot; --'__ _ B7 5. Description of the invention (78 ^ ~~ ---- ~ At the same time, the bandwidth requirement of the message flow level 1 can be expressed by the following formula Η:

Pi = 0.064 /? I (Equation 5 0) The parameter defined here is provided by Lu definition * —coercion and unevenness of volume and non-uniformity, that is, it can carry multiple levels of message traffic by itself at each = Xin level In this case, the difference between frequency buying and demand at different levels is 0-the next / two dry &amp; ± ^, and the early value of P represents uniform message flow. The parameter «boundary distribution of message flow requirements among the four message flow levels. : Two other path capacity allocation methods have been compared with the methods described below. These other methods represent two extreme approaches, one of which is slow in time but based on a rigorous theoretical framework; the other is faster in time but based on rough theoretical simplification. The first reference path capacity allocation algorithm uses a fixed-point algorithm, which is detailed in A. Farago, S. Blaabjerg, W. Holender, T Henk, L. Ast, A. Szentesi, and Z_Ziaga " Resource Separation Species: An Effective Tool for Optimizing AT Μ Network Configuration ", this article is published in, NETWORKS94 International Network Planning Symposium Proceedings 83-88 (Budapest, 1994), the algorithm is also detailed in SP Chung and K.W_ Ross, "Approximation of load reduction for multi-level lossy networks", published in "41 IEEE / ACM Transactions on Communications 1222-31 (1993). This algorithm represents a kind of A known method for configuring the path capacity of a virtual network. However, the speed of this algorithm is not very fast, so it is mainly used to evaluate the accuracy of the results obtained by using the new V ΝPath capacity configuration technology described in this application _ -81-_ _ _ _ This paper scale is applicable to Chinese countries Kneading (CNS) Α4 specification (210Χ: 297mm) 1 I n H-I-ί ^ ί- I — (Please read the note on the back of Sr before filling in this page) Order

T B7 V. Description of the invention (79). The theoretical framework of this benchmark algorithm is quite rigorous. However, this architecture implements two assumptions that oversimplify the link independence of the path capacity configuration problem and reduce the load. The second reference path capacity configuration method uses the equivalent link blocking (equivalent link blocking; ELB;) algorithm, which is described in detail in the paper examined by A. Fargo "A method based on the probability of equivalent link blocking Simplified Path Valley Configuration Model, "This paper is published by the Department of Telecommunications and Telematics of Budapest University of Technology. This algorithm represents a rough and ready-made virtual network path capacity allocation method based on the theoretical assumptions of path capacity allocation problems. This algorithm is mainly used to evaluate the calculation speed (that is, execution time) of this new v N path capacity allocation technology. We use the following three different evaluation parameters to compare these three methods: network revenue, maximum path congestion, and Execution time. Because the new VN roadbolt volume configuration technology does not generate network revenue and maximum path congestion (as a measure of time congestion), these parameters have been calculated extra. Printed by Beigong Consumer Cooperatives, Central Bureau of Standards, Ministry of Economic Affairs The allocated capacity obtained by adopting the VN path capacity allocation technology of this application is used as a benchmark The input of the radial capacity configuration technology. In this way y, using the same parameter relationship as the reference algorithm, you can calculate each material calculation parameter as a time congestion congestion measurement &amp; income function. ^ This application ; ^ As detailed in the previous article, other network operating parameters can be obtained from the average message volume ratio function. I have already decided for each of the three VN path capacity configuration algorithms for each of the three VN path volume configuration algorithms. The profit function and blocking parameters of the provided information flow examples. The results of these comparisons are shown in Tables 17 to 20. Three I--. 〇〇 The paper size is applicable to public 1). ) Table 17 in the dimension chart

Please read the I-i first, fill in 5 pages, and order it.

83- This paper scale is applicable to China National Standard (CNS) A4 specification (210X 297mm) A7 B7 V. Description of invention (81) Table 19.

Table 20 (please ^^ read the back *-the precautions and then fill out this page). Pack. Order Printed by the Beige Consumer Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs

t Shui-84- This paper scale is applicable to the Chinese National Standard (CNS) A4 Sige (2 丨 0 X 297 mm) Λ7 B7 Printed by the Employees Consumer Cooperative of the Central Bureau of Economic Development of the Ministry of Economic Affairs-85- V. Description of the invention (82) We found that the average message volume VN Lujiao Jiaowu Xianmiao &amp; ^ A, the calculation speed of the slightly-combined quantity allocation algorithm is about twice as fast as that of the fixed-point benchmark algorithm, and ^ -times faster than that of the ELB benchmark algorithm. Conclusion The system and method of the present invention have slaved {裎 + τ 〇,... Two out and solved two related problems: VP path capacity configuration problem and v M ^ t and VN path fe various configuration problems. In the solution technique detailed in this application, I use the ratio function of λ- · τ and the average message volume as the blocking measurement, and load balancing as the path capacity allocation technique. Using several examples and comparing the two benchmark algorithms, the feasibility of the method adopted by the present invention has been proved. The calculation method of this new path capacity M method is twice as fast as the conventional path capacity configuration technology. This obviously faster calculation speed comes at the cost of calculation accuracy. Because of &amp; this new method provides better results than conventional techniques. This approach is also the best overall. The child method is based on end-to-end congestion measurement and does not require the assumption of link independence and load reduction of conventional techniques. This method is of great importance in the practicality of configuring ATM network path capacity, because this method does not assume any specific message traffic distribution, but can be applied to general message traffic. In the future, it is possible to configure the path capacity of the A τ Μ network based on actual measurement of message traffic. The path capacity allocation system and method detailed in the present invention provide a practical architecture for implementing a path capacity allocation algorithm that can handle the general provision of message traffic and the provided message traffic determined by measurement. Although the preferred embodiments of the method and apparatus of the present invention have been shown and described with reference to the drawings and the detailed description above, we should understand that the present invention is not limited to the disclosed embodiments, but can be found in Don't deviate from the following application for the new financial standards: Shi Cai Guan Jia Ti Zhun (CNS) ~ 秘 状 297 公 # (please read the precautions before filling in this page) • Install · A7 B7 5. Description of the invention (83 〉 Under the spirit of the invention as described and defined in the patent scope, make various reconfigurations, modifications, and substitutions. --------- i outfit — (please read the notes on the back cup and fill out this page) The size of the paper printed by the Employee Consumer Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs is applicable to the Chinese National Standard (CNS) A4 (210X297mm)

Claims (1)

A8 B8 C8 D8 ----------... / ½ 年 7 取 / eiPuizheng No. 851〇85〇6 Patent Application Application for Patent Fantai ⑽ 年 7 E,) Patent Application Scope 1. A variety of configuration-carrying general information traffic telegrams. The method of defining the capacity of each virtual network on the network. The network has multiple interconnected channels with limited transmission capacity. The path capacity configuration The method includes the following steps:. Define a hierarchy of at least three layers as a model of network resources and needs. The first layer contains physical networks, the second layer contains multiple virtual networks, and the third layer contains multiple virtual networks. Path, each of the virtual networks on the second layer is associated with a plurality of physical links on the first layer, and each of the virtual paths on the third layer The virtual road network is associated with a plurality of virtual networks on the second layer; select a partitioned structure and define the scope of resource sharing; stack the three-tiered hierarchy into a two-tiered hierarchy, in which the new One layer contains the physical network, and the new second layer contains multiple superimposed virtual Path; derive the corresponding superposition structure matrix for each superposition virtual path; connect the superposition structure matrices in sequence to generate a compiled structure matrix, which is used to represent simplifying the virtual network's path capacity configuration problem into one The problem of effective virtual road traffic capacity allocation problem; The Beigong Consumer Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs prints and allocates the capacity to each superimposed virtual path under the transmission capacity limitations of these multiple physical links; The routing matrix of each virtual path_the relationship between each logical link determines the capacity of each virtual link to be allocated to each virtual route 0. 2. Virtual network path capacity configuration according to item 1 of the patent scope Method This paper scale applies the Chinese National Standard (CNS) A4 specification (210X297mm) 319944 as Β8 C8 Printed by the Employee Consumer Cooperative of the Central Bureau of Economic Development of the Ministry of Economic Affairs. The partition structure used to define the scope of resource sharing is the virtual path partition structure. 3. According to the virtual network path content of the patent application item No. 丨, the partition structure used to define the resource sharing range is a virtual link partition structure. 4. According to the method of virtual network path capacity allocation according to item〗 of the patent application scope, the partition architecture used to define the resource sharing scope is the physical link partition architecture. 5. According to the patent application scope item 1, the method of allocating network path capacity allocation method, in which an algorithm performs the step of allocating capacity to each virtual path, the algorithm includes the following steps: selection-proper averaging The message volume ratio function, as a model of the load on each virtual path of the telecommunications network; choose a solution algorithm, which can be calculated to solve the load balancing problem of the general message flow; and use the ϋ average The load balance algorithm of the M interest rate ratio function performs calculations on a computer system in order to generate as uniform a load distribution as possible on these virtual paths. 6_ The virtual network path capacity allocation method according to item 丨 of the patent application scope, in which an algorithm performs the step of allocating capacity to each virtual path. The algorithm includes the following steps: Select an appropriate average Message volume ^ humble function, as a model of the load on each virtual path of the telecommunications network; choose a solution that uses the average message volume ratio function as a blocking measurement ΑΑ specification (2 丨 Οχ297 Gongqing) (please read the note Ϋ on the back and then fill in this page) -install- * A8 B8 C8 D8 The scope of the patent application The answer algorithm can be calculated to solve the general message flow ^ load balancing problem; and, = the load balancing algorithm using the average message volume ratio function% brain system Ή to perform calculations' A 4a uniform load distribution is generated on the path. ΤVirtual network path capacity allocation method according to item 6 of the patent application scope, which uses message traffic measurement to determine the average message volume ratio function 0 8 The average message volume ratio function is determined by idealizing the characteristics of the message traffic provided on a telecommunications network. 9_ The virtual network path capacity allocation method according to item 8 of the scope of the patent application 4, in which the average message volume ratio function is idealized for a uniform Poisson distribution of the provided message traffic. 10. The virtual network path capacity allocation method according to item 8 of Shenlong's patent scope, in which the average message volume ratio function is idealized for a multi-level Poisson distribution of the provided message traffic. 11_ The virtual network path capacity allocation method according to item 8 of the patent application scope ’where the average message volume ratio function is idealized for the normal distribution of a provided message flow. · 12_ The virtual / pseudo-network path capacity allocation method according to item 8 of the patent scope ’which is based on the binomial distribution of the amount of information provided to make the average &lt; flrv average information volume ratio function idealized. 13. Virtual network path capacity allocation method according to item 8 of the patent application scope -3-Use Zhongguanjiabiaofeng (CNS) A4% # · (21GX297mm) {Please read the precautions on the back before filling this page ): Packing-Order A8 B8 C8 D8 printed by the Consumer Labor Cooperative of the Central Standards Bureau of the Ministry of Economic Affairs The scope of the patent application. The average message volume measurement used as the provided message flow model is the average message volume ratio function Ix (c), which is the average message ^ Please read the notes on the back before I install the X ratio function Calculate the approximate method of the negative logarithm of the probability that a randomly assigned random variable X is greater than or equal to-the preselected value C, and the average message volume ratio function is a minimum convex function obtained on the average of the distribution. 14. The virtual network path capacity allocation method according to item 丨 of the patent application scope, in which an adaptive virtual path capacity allocation algorithm performs the step of allocating capacity to each virtual path, the algorithm includes the following steps: : — ^ 一定 ^ The plurality of physical links are mapped to one or more virtual paths, each of these virtual paths provides a switchable between a pair of nodes of the telecommunications network Connection; specifying the transmission capacity of each physical link; concentrating the selected plurality of these virtual paths into a path capacity configuration set; using the point average message volume ratio function as a blocking measurement to determine the initial value of the transmission capacity Allocated to each virtual road in the set of capacity allocations of this path by the Ministry of Economic Affairs Central Bureau of Standards, the Immigration Consumer Cooperative Printed Path, each of these start values is the same and is selected to make the blockage larger; The reciprocal approach will complete the valley volume (assigned to each virtual route through a physical link. Those physical links are mostly classified as key links. Do step further comprises the steps of: using a previous average ratio of the amount of information and the previously calculated function feedstock Zhi 'Patent application scope — the mobile parameter value calculated, to calculate the capacity to be allocated to each virtual path; add up the capacity of all such virtual paths allocated to each physical link, so as to obtain the allocation on each physical link The total capacity of the child; increase the estimated value of the average message volume ratio function by a calculation amount; recalculate the movement parameters of each virtual path in the child path capacity configuration set; the total capacity allocated to each physical link and the physical link Comparison of the various quantities in order to determine whether the unallocated capacity of the physical link is approximately zero. Output the capacity currently allocated on each virtual path identified as an off-amp; from the child path capacity configuration set Remove all virtual paths through each critical physical link; and redefine the specified actual capacity of these physical links to compensate for the capacity allocated to these removed virtual roads. 15. A system with a path capacity for each virtual network defined on a telecommunications network that carries general message traffic. The network has multiple interconnected links with limited transmission capacity. The system includes: Ministry of Economic Affairs Printed by the Central Standard Bureau's Beigong Consumer Cooperative, a device that defines at least three layers as a model of network resources and needs, where the first layer contains physical networks and the second layer contains multiple virtual networks. The third layer contains A plurality of—virtual paths, each of the virtual networks in the virtual networks on the second layer is associated with a plurality of physical links on the first layer, and those on the third layer Each virtual path of the virtual path is associated with a plurality of virtual networks on the second layer; -5- This paper standard is applicable to the Chinese National Standard (CNS) A4 specification (2H) X297mm) A8 B8 C8 D8 VI 、 Patent application range selection device, used to select-separate architecture, and define the scope of resource sharing; 4-in device, used to superimpose the three-tiered hierarchy into a two-tiered hierarchy, of which the new- The layer contains this Volume network, and the new second layer contains a plurality of superimposed virtual paths; a derivation device to derive the corresponding superimposed structure matrix for each superimposed virtual path; σ sequential devices to sequentially connect the superimposed virtual paths Structure matrix, so as to generate a compiled structure matrix, which is used to represent ^ the virtual network path capacity configuration problem is reduced to an equivalent virtual path path capacity configuration problem; the distribution device is used in the plurality of physical links The capacity is allocated to each superimposed virtual path under the limitation of the transmission capacity; and a decision device is used to determine each to be allocated to each virtual network using a routing matrix that defines the relationship between each virtual path and each logical link The capacity of the virtual link. 16-Virtual network path capacity allocation system according to item I5 of the patent application scope. The partition structure used to define the resource sharing range is a virtual path partition structure. 17_ The virtual network road capacity allocation system according to item 15 of the patent application scope, wherein the partition architecture used to define the resource sharing scope is a virtual link partition architecture. _ 18.According to # 15 of the scope of patent application, the proposed network path capacity allocation system, in which the partition structure used to define the resource sharing scope is a physical link partition structure. Please read first, 'Read back ¢-ί Matters and write this page Bound by the Ministry of Economic Affairs Printed by the Central Bureau of Industry and Commerce M-6 Consumer Cooperatives-6- A8 B8 C8 ---------- D8 VI. Application ^ Scope ~ 1 ~-1M Gen According to the patent application No. 1! A five-item virtual network path capacity allocation system, where the device that allocates capacity to each virtual path includes: a device that selects an average message volume ratio function to select an appropriate average message volume ratio function as the telecommunications network The model of the load on each virtual path of the road; select the device that answers the algorithm, which can be calculated to solve the load balance problem of far-general message traffic; and the device that performs the calculation to use the average message The load balancing algorithm of the volume ratio function performs calculations on a computer system in order to produce as uniform load distribution as possible on these virtual paths. 20. The virtual network path capacity allocation system according to item 15 of the patent application scope, wherein the device that allocates capacity to each virtual path includes: a device that selects an average message volume ratio function to select an appropriate average message The volume ratio function is used as a model of the load on each virtual path of the telecommunications network; a device that selects a solution algorithm is used to select-the average message volume ratio function is used as the solution algorithm for the blocking measurement, the solution algorithm The method can be calculated to solve the load balancing problem of the general message flow; and the Central Standards Bureau of the Ministry of Economic Affairs of the Ministry of Economic Affairs and Consumers Cooperative Printed the device to perform the calculation 'to use the load balance algorithm using the average message volume ratio function. Perform calculations on the two computer systems in order to generate a uniform load distribution on the virtual paths. 21. The virtual network path capacity allocation system according to item 2 of the patent application scope 'in which' uses' the information flow measurement to determine the average message volume ratio function. The paper scale is applicable to the Chinese National Standard (CNS) A4 said grid (210χ297 Mm) A8 B8 C8 printed by the Beigong Consumer Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs ^ Application for patent scope 1, = virtual kiss path capacity allocation system for item 2 of the patent scope of patent application 2 The characteristics of the provided message flow are ideal to determine the average message volume ratio function. The virtual network path capacity allocation system according to item 22 of the β-Shenzhen patent scope: 'Among them, the average message volume ratio function is idealized for the uniform Poisson distribution of the provided message traffic. 24_According to the 22nd item of the patent scope, the virtual network path capacity allocation system is aimed at the multi-level Poisson distribution of the provided message traffic to make the average speech volume ratio function ideal. 25. The virtual network path capacity allocation system according to item 22 of the Chinese patent application, where the average message volume ratio function is idealized for the normal distribution of a provided message flow. 26. The virtual network path capacity allocation system according to item 22 of the patent application scope, in which the average message volume ratio function is idealized for the binomial distribution of a provided message flow. 27. The virtual network path capacity allocation system according to item 22 of the patent application scope is used as the average message volume congestion measurement for the provided message flow model. The average message volume ratio function X (c) is the average message The volume ratio function is calculated as an approximation of the negative logarithm of the probability of a randomly distributed random variable χ greater than or equal to-the preselected number value c, and the average message volume ratio function is again a minimum value obtained on the distribution average Convex function. 28 'The virtual network path capacity allocation system according to item 15 of the patent application scope-8 This paper standard is applicable to the national government (CNS) 84 4 threat (21〇 297 public director) (please read the notes on the back first (Fill in this page again) 、 1T • —. ^ 1--The scope of patent application The device that allocates capacity to each virtual path includes: a mapping device to map these multiple physical links to one or more evenings A virtual path, each of these virtual paths provides-a connection that can be individually switched between the nodes of the telecommunications network; a device that regulates the transmission capacity of each physical link, centralized Install "to collect multiple selected virtual paths into a path capacity configuration set; / distribution device" to use the average message volume ratio function as a blocking I test to set the initial value of transmission capacity ^ Assigned to each virtual path in the path capacity configuration set, each of the starting &amp; starting values are two and selected to make the blockage larger; the recursive identification device is used to recursively Capacity is over The f ^ channels assigned to the virtual members passing through the real path are the key links, and the recursive identification device further includes: a computing device for estimating the value using the previous average message volume ratio function and calculating W first Dynamic parameter value, to calculate the capacity to be allocated to each proposed path; the Central Standard Falcon Bureau of the Ministry of Economic Affairs, the Consumer Cooperation Cooperative Institutional Summing Device, adds up the capacity of all such virtual paths allocated to each physical link, so that Obtain the oil capacity on each physical link;. Estimate the average message volume ratio function as a device; two different quantities and recalculate the device of each virtual virtual parameter in the path capacity configuration set;- 9------- This paper scale is applicable to China National Standard (CNS) Α4 specification (210Χ297mm) ABCD 6. Patent scope comparison device to compare the total capacity allocated by each physical link with that entity The specified capacity of the link is compared to determine whether the unallocated capacity of the physical link is approximately zero; an output device is used to output each virtual link that is identified as a critical link The currently allocated capacity on the proposed path; a device to remove the virtual path to remove all virtual paths through each critical physical link from the set of path capacity configurations; and a device to redefine the actual capacity to redefine the path The actual capacity of the physical link is required to compensate for the capacity allocated to the virtual paths that have been removed. Please read the notes on the back-first Printed by the Employee Consumer Cooperative of the Central Bureau of Standards of the Ministry of Economic Affairs -10- This paper scale applies the Chinese National Standard (CNS) A4 specification (210X297 mm)