GB/T 15274-1994 Internal organization of the network layer of information processing systems Open Systems Interconnection
Some standard content:
National Standard of the People's Republic of China
Information processing systems-Open systems Interconnection: Internal organization of the network layer
GB/T 15274-94
ISO8648-1988
This standard is equivalent to the International Standard 1ISO8648-1988 Internal organization of the network layer of information processing systems-Open systems Interconnection, including the technical error table 1 of 1ISO8648-1988 published on March 15, 1991.0 Introduction
This standard defines the organizational structure of the network layer architecture of the OS1 reference model. It involves the organizational structure of the network layer entity functions in the open system and the method of mapping the organizational structure to the "real world" components (such as "real network, switching equipment, transmission media, etc.). This standard links the "real world" objects that must be studied with a set of abstract requirements. There may be various mappings between abstract elements and the physical devices used to implement them. Describing these mappings requires a clear distinction between architectural and real-world terms. The architectural organization defined by this standard identifies and categorizes the ways in which network layer protocols perform functions within the network layer. In doing so, it provides a unified framework for describing how different network layer protocols can be used to provide OSI network services, either individually or in concert. While focusing on the functional elements common to network layer protocols, the framework is intended to: simplify the use of network layer protocols to provide network services in different contexts; limit the proliferation of uncoordinated network layer protocols with overlapping functionality; clarify requirements for future network layer protocol standards and guide their development. In the context of networks, this detailed internal structure is necessary for two reasons: 1) The network layer provides a uniform network service to its users, regardless of the various variations in the underlying "real world" network services, technologies, and regulatory bodies. It is important to understand how the underlying fabric is organized and interacts within the network layer and how they can be used effectively and efficiently. 2) GB9387 stipulates that the network layer performs routing and relay functions and may contain entities in intermediate systems and end systems. In both types of entities, it is necessary to describe the events that occur within the network layer. These two types of entities refer to the end system [providing network services to network service (NS) users and intermediate systems (network entities that provide these relay and routing functions do not provide network services to NS users). 1. Content and scope of application
This standard provides an OSI network layer architecture model as a framework for OSI network layer standardization to enable existing networks to be incorporated into OSI
This framework encourages the design of real subnets that fully support OSI network services, and also adapts to other subnets that do not fully support OSI network services to join the OSI environment.
This standard covers the design and application of ten network layer protocols. These protocols operate between network entities in end systems that provide OSI network services or in intermediate systems that provide routing and relay functions. This standard applies to:
Approved by the State Administration of Technical Supervision on December 7, 1994 and implemented on August 1, 1995
GB/T15274-94
Provides a Set of common concepts and terms used in network layer standards (these network layer standards should reference this standard): analyze the functionality of the network layer and classify network layer protocols; specify how the "real network" should be used to support or provide OSI network services, especially when multiple "real networks" are to be interconnected and used.
The organizational structure defined here does not involve network layer management, especially the relationship between network layer entities that may be required for network layer management purposes. In addition, this standard does not specify the operational requirements for relay functions within intermediate systems, nor does it involve how specific combinations of allowed functions can be actually and effectively utilized.
2 Referenced standards
GB9387 Basic reference model for interconnection of information processing systems Note, also refer to CCITT X.200 Recommends the open system interconnection model applied by CCITT. GB11593 Interface between data terminal equipment (DTE) and data circuit terminating equipment (DCE) working synchronously on public data network GB11595 Interface between packet data terminal equipment (DTE) and data circuit terminating equipment (DCE) connected to public data network by dedicated circuits
GB/T15126 Definition of network service for data communication of information processing system Note, also refer to CCITTTX.213 Recommends the network service of open system interconnection (OSID) applied by CCITT Service definition. GB/T15129 Information processing system Open System Interconnection Service Agreement ISO7498/Supplement 1 Information processing system Open System Interconnection Basic Reference Model Supplement 1: Connectionless transmission ISO8208 Information processing system data communication X.25 packet level protocol for data terminal equipment ISO8348/Supplement 1 Information processing system data communication Network service definition Supplement 1 Connectionless transmission Note, see also CCITTX.210 Recommendation Open System Interconnection (OS1) layer service definition agreement. Part 2: Information processing system data communication Kitchen area network logical link control ISO ISO 8802
ISO8802
Part 3: Carrier Sense Multiple Access with Collision Detection (CSMA/CID) Access Method and Physical Layer Specifications Part 4: Token Passing Bus Access Method and Physical Layer Specifications ISO 8802
Part 5: Token Ring Access Method and Physical Layer Specifications ISO 8802
Part 7: Slotted Ring Access Method and Physical Layer Specifications ISO8802
3 Terminology
3.1 Reference Model Terminology
This standard uses the following terms defined in GB9387: (XSI Network Layer; a
OSI Network Service OSI Network Serviceib.
OSI Network Service Access Point OSI network-service-access-pointtc
OSI Network Service Access Point Address (Network Address) OSI network-service-accegs-point-address（network-d.
address)
OSI network entityOsInetwork-entitynetwork-relay；
routing，
serviceservicel
protocolprotocol，
protocol-control-informatian。 3.2 Service Conventions
This standard uses the following terms defined in GB/T 15129: service userservice user;
service providerservice provider. b.
3.3 Network Layer Architecture Terms
The following terms apply to this standard:
a. Real subnetwork
GB/T 1527494
A collection of devices and physical media that form an autonomous entity and are used to interconnect real systems for communication. b. Subnetwork
An abstraction of a real subnet.
c. Interworking unit One or more components or a part of a component in a device, whose operation provides a network relay function (that is, a real system that receives data from one network entity and forwards it to another corresponding network entity). In general, such a device can be nested in an interconnecting real network. d. Network relay system network-relay system An abstraction of a device that forms an interworking unit. e. Intermediate system An abstraction of a real system that provides a network relay function. f.Data transmission service A set of capabilities available from a real network that can be used by users of the real subnet to receive and send data. Note: This service is defined between each user connection to the real subnet and the points of access to it (e.g. DTE/DCE interfaces), and should not be confused with the concept of services at layer boundaries as defined in GB 9387. 4 Authorization
The following abbreviations apply to this standard:
CL-NS-
Connection-mode network service;
Connectionless-mode network service:
DCE data circuit-terminating equipment;
Data terminal equipment:
V working unit,
Local area network;
Network layer:
Network service access point:
PC1-Protocol control message;
SNAcP-
SNDCP-
SNICP-
Subnetwork access protocol;
Network-dependent convergence coordination protocol
Subnetwork-independent convergence coordination protocol.
5 Network layer overview and terminology
This chapter gives precise meanings to several of the following terms used in the OSI network layer standard, but these terms have many interpretations in more general contexts.
5.1 Real-world objects and abstract elements
In this clause and in Chapter 8, diagrams are used throughout to illustrate various interesting configurations in the network layer. Each diagram has two components. Real-world objects are depicted at the top of each diagram, while representations of the corresponding abstract elements are depicted at the bottom. Table 1 shows the graphical representations of real-world objects and abstract elements used in each diagram. CB/T 15274—94
Figures 2, 3 and 4 depict the basic combination of real-world objects (and their corresponding abstract elements) interconnected. From this basic set of interconnections, any number of allowed interconnections can be established by repeating and/or recursively invoking the basic interconnections shown in Figures 2, 3 and 4. Note: In Figure 5 and those that follow it, the graphical representations of abstract elements contain additional details (e.g., the layers of SI, protocol roles, etc.). 5.2 End Systems and Intermediate Systems
End systems and intermediate systems are abstractions of devices that implement the requirements of the OSI standards that are necessary to provide the functions of open systems (see GB9387). Intermediate systems only perform the functions of the lower three layers of the reference model. Interworking units and real subnets are real-world examples of devices that implement the role of intermediate systems. End systems must additionally provide the functions of layers above the network layer. A real end system is any device or collection of devices that implements the role of the end system. The network layer entities within the end system should provide the inter-system communication functions of the network layer, which are the same as those provided by the network layer entities in the intermediate system, except for the relay functions that are only performed by the intermediate system. The same collection of physical devices may have the ability to work as a real end system or as a working unit or both at the same time, depending on their use. The functions enabled in specific situations correspond to the functions of the abstract end system or intermediate system. 5.2.1 End System Considerations
An end system can communicate with another end system in one of two ways: a. directly, without the intervention of an intermediate system (see Figure 1) b. through the intervention of one or more intermediate systems (see Figure 2). In this context, "directly" refers to communications involving only network entities in the two communicating end systems. Communications "through one or more intermediate systems" refer to communications involving, in addition to the network entities in the two communicating end systems, one or more network entities performing network relay functions.
In either case, there is no restriction on the low-level relay (e.g., data link layer relay or physical layer relay) methods that may be used to provide communications between network entities.
For example, the physical subnet directly connecting the two end systems, as depicted in Figure 1, can be as simple as a direct point-to-point link between the two end systems, or as complex as a set of interconnected local area networks (LANs). 5.2.2 Intermediate System Considerations
An intermediate system is an abstraction of one of the following: a. A real network (see, for example, 5.3 and Figure 2) that is not directly connected as described in 5.2.1; h. An interworking unit that connects two or more real subnetworks (see 5.4 and Figure 4, which only shows two real subnetworks:
A real subnetwork (see 5.4 and Figure 3) with an associated interworking unit. Any combination of the above may also be referred to as an intermediate system. In some cases, it may be useful to identify the abstract functionality of a real subnetwork and distinguish it from the abstract functionality of an interworking unit. In this case, the abstract terms "subnetwork" and "relay system" may be used to refer to specific types of these intermediate systems (as described in 5.3 and 5.4).
5.3 Real subnets and subnetworks
A real network is a collection of equipment and physical media that form an autonomous whole and are used to interconnect other real systems for communication (see Figure 2). Examples of real subnets include:
a. Public networks provided by recognized communications departments; b. Networks provided and used by specialized departments; e. Local Area Network (LAN);
d. A real subnet formed by taking into account any of the above networks and related interface devices and (or) interworking units. All equipment that constitutes this real subnet It must be pointed out that the collection of such physical devices listed above and represented here as "subnets" is usually called a "network"..comReal world
Real system
Real open system
Interworking unit
Real subnet
Real end system
GB/T15274—94
Table 1 Correspondence between real-world objects and abstract elements
Graphical representation of real-world objects
Corresponding
Abstract elements||t t||Open system
Network relay system
Peak system
Abstract key space
Graphical representation
Intermediate system
Intermediate system"
When the system
Note: 1) For directly connected end systems (see 5.2.1), their interconnection will be shown as a single link rather than an intermediate system. The externally visible characteristics of some real subnets are described by their respective standards, such as: GB11593 circuit switching and GB.11595 packet switching standards or local area network standards. In the interconnection of other real systems, more than one real subnet may be involved. The term "subnet" can be recursively applied to a collection of subnets that are interconnected in such a way that, for a connected system, they can be regarded as a single network and treated as a single subnet.
A real subnet can be implemented in a way that fully supports the OSI network services. In other words, the functionality provided by the real subnet and appropriately utilized by the real end systems connected to it is abstractly equivalent to the functionality of the OSI network services (see Section 5.5). A real subnet with this property is said to support all the required OSI network services. If a real subnet has full support for OSI If an interworking unit is to provide the capability of providing network services, then it must provide all the necessary elements of OSI network services. It may also provide any OSI network service provider options, but is not required to do so.
5.4 Network Relay Systems and Interworking Units
A network relay system is an abstraction of a device (or a collection of devices) that is usually called an interworking unit in the real world. The purpose of an interworking unit is to facilitate the interconnection of various real subnets. An interworking unit may belong to the management department of the real network provider or may be managed by some other department. In the case where the interworking unit and the real subnets interconnected with it are physically separated, the interworking unit may be represented separately as an intermediate system (see Figure 4). The interworking unit may also be included as an integral part of each interconnected real subnet. In this case, each real subnet and its associated interworking unit may be represented as a single intermediate system (see Figure 3a). Alternatively, each real subnet may be represented as an intermediate system and connected by another intermediate system that represents the intervening interworking unit maintenance functionality (see Figure 3b). 5.5 Data transmission services and subnetwork services
Any real subnetwork provides a certain set of functions to the real systems that use it. Some real subnetworks perform functions defined above the network layer of the OSI architecture (or outside of OSI entirely), which are outside the scope of this standard. The abstraction of a set of capabilities available at the interface between a real end system and a real subnetwork is called a data transmission service. Different types of data transmission services are available for different types of real subnetworks (e.g. packet switching and circuit switching data transmission services). In the context of the OSI network layer, an abstraction of the functions provided by a subnetwork and those functions that need to be performed in an open system using these subnetwork functions is called a subnetwork service. A subnetwork service includes a data transmission service, or a subset of it (see Figure 5). Thus, there may be three types of network services: a. equivalent to an OSI network service in all cases of use (see 5.3); b. not equivalent to an OSI network service in all cases of use; e. not equivalent to an OSI network service in some cases of use. Network services, but in some other usage scenarios are equivalent to OSI network services. When a real subnet supports subnet services of type b or c, additional network layer functions need to be implemented in order to provide the full range of USI network services in all cases. These additional functions are implemented in the end systems and/or network relay systems that enable subnet services on the real subnet.
GB/T15274-94
A special case of type c subnet connection mode service is that a real subnet is defined so that its access protocol contains the functions required to support the establishment and release of network connections, but does not contain all the functions required to support data transfer on these network connections. Therefore, for such a subnet, some enhanced protocols are required during the data transfer phase, rather than the establishment or release phase. 5.6 Service Types
OSI network services include two basic service types: connection mode service (CO-NS) (see GB/T 15126); b connectionless mode service (CL-NS) (see GB/T 15126 and ISO 8348/Supplement 1). Likewise, a subnet may provide either connection-mode service or connectionless-mode service, or both. Any combination of subnets that can provide either CO-NS or CLNS (see Section 8 and Table 2) enables the operation of the network layer protocol to provide either CO-NS or CI-NS, or both services.
For a given communication situation, the same service (i.e., connection-mode or connectionless-mode) is provided to both NS users. The choice of providing connection-mode or connectionless-mode network service is subject to the requirements of 1SO7498/Supplement 1. 6.2. 6 Network layer organization structure
There are cases where some of the interconnected physical subnets do not fully support OSI network services. Therefore, it is necessary to describe a network layer architecture framework to provide OSI network services using such physical subnets. This standard provides such a framework, which can be used in the development and application of standards related to the interconnection of physical subnets. 6.1 Factors affecting the internal organization structure of the network layer When providing OSI network services, there are cases where all the physical subnets involved are subnets that support all elements of the network service. In these cases, the OSI network service can be provided using a single network layer protocol associated with each subnet (now 6.7). However, in other cases, because the subnet services are different, it may not be enough to use only a single network layer protocol for each subnet. Therefore, in order to realize the OSI network service, multiple network layer protocols have to be used to operate together. Therefore, for a given combination of interconnected end systems and intermediate systems, a specific network layer coordination can perform all functions necessary to provide SI network services or only some functions. In order to understand when and how multiple network layer protocols are used to provide OSI network services, it is therefore useful to introduce the concept of a "role," which is the role that a network layer protocol performs when used in a given combination of systems. The role of a network layer protocol is a definition of the function it performs when it helps to provide an OSI network service. The "role" is determined by the combination of protocols used and the relationship of that protocol to any other protocols that may be operating. Therefore, the role is always relative. NOTE The elements that determine the role of a protocol are both the agreed-upon services that the protocol is to provide and the underlying services that are envisioned. Any protocol (for the network layer or any other layer) is defined relative to a particular underlying service; part of the definition of a protocol is the way it makes use of that service. 6.2 Description of possible roles of network layer protocols Three roles of network layer protocols have been identified to describe how multiple network layer protocols can be used to establish OSI network services (see Figure 6): The role of the Subnet Independent Convergence Protocol (SNICP): The role of the Subnet Dependent Reception Protocol (SNDCP): b. The role of the Subnet Access Protocol (SNAcP): c. In a particular combination of systems, a protocol at the network layer may perform one of the three roles described above; the same protocol may perform the same or different roles in different combinations. Each role is not necessarily performed by a separate protocol. The case where a single network layer protocol provides OSI network services can be described as a case where the single protocol performs all defined protocol roles. An example of such a configuration is described in Section 6.7 and Figures 7a, 8a, 9a, and 9h. Sections 6.3 to 6.5 and Figures 10 to 21 describe configurations that utilize multiple network layer protocols.
It is generally not required that a separate protocol be used to perform each role in Figure 6, for example: a single network layer protocol that performs all network layer protocol roles simultaneously can also provide OSI network services. The notations "SNICP", "SNDCP\ and \SNAcP\ in Figure 6 refer to the case where a separate protocol performs each of the SNICP, SNDCP. and SNAcP roles. Those labeled "1\ may not be functionally equivalent to those labeled "2".
Real son-in-law system anti
end system Yao
female love music system
Rui system
GB/T 15274-94
Figure! "Vertical" communication between end systems (no intermediate system) Real network
Intermediate peach
End system communication through intermediate system
Real increase in departure
Additional system
Real end system
Full system
Real wax system
Tide system
Real end system
Tuning system
Real end system
Real two
Top example
Intermediate system
GE/T15274-94
Real subnet
Real subnet
Intermediate system
aRepresent each real network and interaction as an intermediate system and two
*国*
Intermediate system
Intermediate control system
bThe combined interworking functions are represented as a single integrated intermediate system Figure 3 The interconnection of subnets including the complete interworking unit functionality is implemented in the network
Intermediate system
Interworking unit
Intermediate system
Intermediate system
Figure 4 The interconnection of subnets implemented with a single interworking unit is implemented in the end system
Lake system
End system
End system
British Swiss system
Travel line
GB/T 15274-94
Point within an end system that can access data transmission services
Real subnet
Intermediate system
Point within an end system that provides subnet services
Figure 5 Relationship between subnet services and data transmission services Note: This figure depicts situations where network services are not equivalent to ()S1 network services, SNICPI
SNAcPI
Independent of the effects of subnet
Routing and intermediate links
(see note)
Depends on the two||t Figure 6 Network layer protocol rules
End system
Link bzxz.net
SNICP?
SNDCP2
SNACP2
Note: The rectangle marked "Routing and Relay" indicates the necessary functions to link the protocol information on both sides of the slash. This enables information to be forwarded from one intermediate system or end system to another, while supporting (SI network services. 6.3 Inter-Subnet Access Protocol
A protocol that performs the SNAcP role and operates under the constraints of characteristics that are clearly defined as being specific to a particular network. The operation of this protocol facilitates the provision of network services that are specific to the relevant subnet. Such subnet services may or may not be consistent with (OSI network services. The manner in which this protocol facilitates the establishment of OSI network services is governed by the definition of the corresponding subnet service and the manner in which other network layer protocols (if any) are selected to utilize that service in a particular configuration. Within a given subnet, there may be relay and routing functions that govern the forwarding of information entirely within the subnet itself. The fulfillment of the routing and relay functions of the subnet is associated with the operation of the protocol that performs the SNAcP role. 6.4 Subnet-Independent Convergence Protocol
A protocol that performs the SNAcP role establishes an OSI network service by operating on a well-defined set of underlying capabilities that are not necessarily based on the characteristics of any particular network service. GB/T 15274—94
The specification of a protocol intended for use in the SNICP role begins with the definition of such a set of lower-layer capabilities that it uses. Such capabilities may be obtained by the operation of other network layer protocols (or by the provision of data link services). The selection of an appropriate set of lower-layer capabilities is based on: a. the types of subnets and combinations of subnets that the protocol is intended to use; b. the technical and economic advantages and disadvantages associated with using one set of lower-layer capabilities rather than another; c. the similarity of the capabilities required to be conveniently provided by the operation of other network layer protocols (or data link services); d. the reasonableness and desirability of using a set of lower-layer capabilities. The complexity of the protocol mechanisms in SNICP itself is expected to be supported. A notable feature of a protocol that performs the SNICP role is that the underlying capabilities upon which the protocol depends do not have to be based on any network service-specific features.
6.5 Subnet-Dependent Protocols
A protocol that performs the SNDCP role by definition operates on top of a protocol that provides the SNAcP role. It is used to provide the capabilities envisioned by a protocol that performs the SNICP role, or to provide OSI network services by performing both the SNICP and SNDXCP roles. The protocol in the SNDCP role is dependent on the features of a particular subnet service and the SNICP 6.5.1 Relationship of SNDCP to SNICP The introduction of the SNAcP role allows the assumptions made by SNICP to be decoupled from the detailed operation of a particular SNAcP, thereby accommodating the generality expected of protocols implementing the SNICP role. The provision of lower-layer capabilities required by SNICP for use of a particular subnetwork service may require the operation of an explicit protocol in the SNICP role (i.e., a protocol involving the explicit exchange of protocol control information between peer network entities). However, there may be cases where the "protocol" in the SNDCP role simply consists of a set of rules for handling the subnetwork service. This set of rules should be implemented in the individual end systems and intermediate systems as required, but should not involve any explicit exchange of PCIs. 6.5.2 Relationship of SNDCP to OSI Network Services In some cases, a protocol may operate directly on a particular subnetwork service (and its associated SNAcP) and provide capabilities that are completely equivalent to the OSI Network Service rather than being a collection of capabilities of some different underlying network service intended to satisfy the SNICP requirements). When the protocol is used in this way, it will be used to perform both the roles of SNICP (providing network services) and SNDCP (operating on a specific subnet service). 6.6 Relay and Routing
The relay function enables a network entity to forward information received from one network entity to another corresponding network entity, and the routing function determines the appropriate route between network addresses. In the case of multiple real subnets in a given configuration, the relay and routing functions can be distributed to each network entity in one of the following two ways:
, the relay and routing functions are located in the network entity outside the real subnet, in which case the functions are performed by identifiable interworking units connecting the two real subnets (see Figure 4); b, the relay and routing functions are located in the network entity within the real subnet, in which case it is considered that one or more [WUs in each real subnet perform these functions, and these real subnets can be represented as a single abstract subnet from the perspective of routing and relaying (see Figure 3a).
If a combination of physical subnets that perform relay and routing functions outside the physical subnets provides the necessary interworking functions with a combination of relay and routing functions within the physical subnets, then the two combinations can be interconnected. 6.7 A single network layer protocol that performs all protocol functions to support network services. The distinguishing feature of all required subnets is that there is no additional protocol (e.g., the protocol described in ISO 8208) between two corresponding (SI) end systems or intermediate systems directly connected to those physical subnets. A network layer protocol for a physical subnet with these characteristics is capable of conveying all the information necessary to provide OSI network services between the appropriate combination of end systems and intermediate systems connecting it. ..com7 Application of the internal organization of the network layer
GB/T 15274—94
The concepts of internal organization of the network layer introduced in Chapter 6 can be used to describe some of the special methods used in interconnecting subnetworks, which are the subject of OSI standardization.
Such methods of interconnection are concerned with the provision of OSI network services through the use of a combination of interconnected subnetworks. Functions such as internetwork relaying and routing are combined with functions performed by individual subnetworks to provide SI network connection-mode or connectionless-mode transport between end systems that may or may not be directly connected to the same subnetwork. The internal organization of the network layer applies to three methods of interconnecting subnetworks, which are described in the following items. These three methods are: a. Supporting all elements of network services. b. Coordination on a segment-by-segment basis on each subnet; c. An Internet Protocol method used on multiple subnets. In the case of connection-mode services, these three methods can be applied independently to each of the three stages of network connection. Note that the term "Internet Protocol" is used in the term "interconnection" to identify a style of interconnection rather than to refer to any specific protocol. 7.1 Supporting OSI Network Services A method of interconnecting subnets that supports all elements of a network service, assuming that all real subnets involved fully support OSI network services. A single network layer protocol is used between the intermediate systems and end systems connected to each such subnet to provide OSI network services (see clause 6.7). These subnets are interconnected with the help of relaying and routing functions (see clause 6.6) to support end-to-end OSI network services. In this type of interconnection, NSAP addresses are transported with the help of a single network layer protocol used. 7.2 Hop-by-Hop Coordination
The hop-by-hop coordination method involves subnets. The present invention relates to a method for interconnecting two subnets, each of which includes at least one subnet that does not support all elements of a network service. In this method, the subnet service of each subnet that does not support all elements of a network service is "coordinated" to be equivalent to an OSI network service. Additional functions are necessary to convert the subnet service into an OSI network service, which generally may include the function of shielding certain features of the subnet service and the function of extending and enhancing the subnet service. In many cases, the addition of these coordination functions will require one or more separate coordinated operations, which perform SNICP functions and possibly SNDCP functions on top of the SNAcP protocol. When each subnet is thus coordinated to support the required OSI network service, the coordinated subnet services are connected by means of routing and relay functions to support the required end-to-end network service. NSAP addresses are carried on each subnet either as PCI in SNAcP or as PCI in SNDCP or SNICP. The choice of protocol for carrying addressing information depends on the capabilities of the underlying subnet. An important consequence of this interconnection method is as follows. To ensure the possibility of communication between all end systems when hop-by-hop coordination is extended across a given subnet, it is necessary that all real end systems and any interworking units connected to that real subnet support the same protocol used to provide hop-by-hop coordination.
7.3 Use of the Internet Protocol Approach
The Internet Protocol Approach defines protocols that can operate on a series of interconnected subnets and on different types of subnets. OSI network services may then be provided by a single arrangement that operates through the combined action of several interconnected subnets and intermediate systems. An Internet Protocol, like any SVICP, must rely on some defined set of capabilities on which to operate. The specification of such a set of capabilities is an integral part of the Internet Protocol specification, and is an important consideration in selecting subnets on which the Internet Protocol can be used.
In practice, various design choices can be made regarding the degree to which intermediate systems participate in various Internet Protocol functions, as well as the complexity of the Internet Protocol itself. For example, intermediate systems may provide functions such as Internet flow control, or may fragment protocol data units, reassemble previously fragmented protocol data units, or both. An Internet Protocol may be used on a subnet that provides services other than the defined set of capabilities on which the Internet Protocol is specified to operate. In these cases, the protocol operating in the SNDCI role (i.e. an explicit protocol or simply a set of rules for handling subnetwork services) is used to convert subnetwork services into the capabilities required by SNICP, which can then operate over SNDCP and SNAcP of the subnetwork (or each subnetwork involved in the interconnection). The NSAP address in this case is carried as an element of the internetworking protocol on all subnetworks involved. When this method of interconnecting subnetworks is used, the possibility of communication between real systems, real end systems and interworking units connected to these real subnetworks is only guaranteed in those cases where these systems use the corresponding internetworking protocol. 7.4 Combination of methods of interconnecting subnetworks
Although the application of three different methods of interconnecting real subnetworks is described in 7.1 to 7.3, it is in principle possible to use any combination of different methods in the combination of interconnected subnetworks. Where a combination of Internetworking Protocols is used but does not operate end-to-end, the Internetworking Protocol acts as a coordination protocol over the segments. B. Interconnection Schemes
The figures in this chapter show various interconnection schemes that illustrate the application of the internal organization of the network layer described in the previous chapters. Table 2 summarizes the schemes depicted by the type of subnetwork service used and whether CO-NS or CL-NS is provided. NOTE: The same subnetwork may be capable of providing both CU-NS and CL-NS, although the service provided in any case of data transmission shall be the same for both users of the subnetwork.
In these figures, the notations \SNICP\, SNDCP\, and \SNAcP\ refer to the respective identifiable protocols that perform the corresponding roles. Where it is indicated that one or more of these protocols may be absent, one or more of the remaining protocols provide all the necessary functions that should be provided by the absent protocols.
The figures summarized in Table 2 depict some specific combinations of protocols used in the network layer in a given communication situation. Such combinations, for connection-mode transmission, occur at any time during the existence of a network connection between the real end systems shown in the figure. Where CL-NS is provided between the real end systems, such combinations may occur at any time during connectionless mode transmission. NOTE: Because these figures apply only to a given situation during data transmission, different schemes can apply to different stages of connection-mode service. 8.1 Single data link/single subnet interconnection Figures 7 and 8 depict the possible range of interconnections via two configurations: 1. single data link;
b. single subnet.
For both configurations, the following two cases are illustrated: 1) using a single network layer protocol;
2) using multiple network layer protocols.
The term single data link applies to the various configurations described in clause 5.2.1. Table 2 Summary of interconnection scheme
Subnetwork service provided by subnetwork A
C/L or C/
Of which; N/A
Subnetwork service provided by subnetwork B
Not available, subnetwork does not exist, connection mode service;
To be provided (OST
Network service
CO-NS or CI-NS
C..NS or CO-NS
Figure 9.10.11
Figure 9.12.13
Figure 11, 15
Figure 16, 17
Figure 18, 19
Figure 20, 2113
Figure 11, 15
Figure 16, 17
Figure 18, 19
Figure 20, 2113
Figure 11, 15
Figure 16, 17
Figure 18, 19
Figure 20, 21
Tip: This standard content only shows part of the intercepted content of the complete standard. If you need the complete standard, please go to the top to download the complete standard document for free.