4.4.9 Interfaces

A node supports several types of interfaces. Apart from the client-signal interfaces, such as SONET, ATM, IP, and so on, there are internal interfaces between bays, shelves, and units for control and other functions. Similarly, there are external interfaces between the node and the network for operations, administration, management, and protection (OAM&P).

4.4.9.1 External Interfaces

The traditional or legacy network (Chapter 3) was designed for switched-voice traffic and was based on copper loops and a network optimized to process calls. However, this network is evolving to offer integrated access, transport, and multiservices (including switching of voice) at a huge aggregate bandwidth.

Networks are composed of a wide variety of NEs from different vendors. NEs are transmission distribution systems, switching systems, access systems, and the like. This is also known as the NE layer. Each network of NEs is complex and is accompanied by its own element management system (EMS). The EMS manages the elements that comprise the network. This is known as the EMS layer. The communication protocol between NEs in the multivendor network and its corresponding EMS varies. It may be proprietary or standard, such as TL1, signaling network management protocol (SNMP), common management information service element (CMISE), and so on. Conversely, the NE-specific EMSs communicate via an open, standard, northbound interface to a higher-level network management system (NMS) (Fig. 4.32).

NMSs manage the network and systems for capacity, congestion, diversity, and so on. This functionality is known as the network management layer. Above that is the service management layer (SML) that is responsible for service quality, cost, and so on. Above the NML is the business management layer (BML), which is responsible for market share, and so on. This hierarchical management responsibility is known as the telecommunications management network (TMN) five-layer structure (Fig. 4.33). Thus, the TMN is a system that consolidates functionality related to network resource management, monitoring, and controlling and ensures the consistent performance of network and services.

Figure 4.32 A multivendor network is connected with the NMS via EMSs.

Figure 4.33 The TMN five layering structure and typical interfaces.

The most representative external interface agents between network elements (NEs) and the element management system (EMS) are:

• TL1: transport language 1 sends filtered alarms to a network fault management system. It is also a bidirectional interface for flow-through provisioning from a network management layer (NML) system. TL1 has been traditionally preferred by competitive local exchange carriers) CLECs.
• ODBC: open database connectivity sends bulk data transfer to either an EMS report generator or to an external analysis and reporting application. It is also an interface for flow-through provisioning from an NML system.
• SNMP: simple network management protocol is for less complex NE-EMS systems to send faults (traps) to an NML fault management system. It is also a bidirectional interface for flow-through provisioning from an NML system. SNMP does not define network topology nor network auto-discovery. It is a specification of an application layer protocol (ALP) and a management information model (MIM) intended for TCP/IP, but it also supports IP, ATM, FR, ISDN, SMDS LAN (Ethernet, FDDI) and other protocols, and therefore it is currently preferred by Internet service providers (ISP).
• Q3/CMISE: common management information service element is a bidirectional common object request broker architecture (CORBA) interface to send filtered alarms to an NML fault management system and to enable flow-through provisioning from an NML. However, CORBA is a general object distribution framework, and not network management protocol, and it consists of an object request broker (that is, a shared library or middleware), servers and objects.
• CMIP: common management information protocol is used by EMSs to communicate with their NEs.
• Native protocols: these are proprietary protocols used by NEs and their corresponding EMSs.

Figure 4.34 Agents and managers between TMN layers.

This large variety of agents may be accommodated within a TMN framework by one of several ways. One is to use a mediation device known as a Q-adapter; A Q-adapter takes a message from a management application and translates it into a language an agent application understands. In turn, the Q-adapter translates alarms and reports from the agent back to the manager (Fig. 4.34).

In the preceding discussion, we have avoided focusing on the management aspects of the various infrastructure layers, such as an IP, ATM, SONET, and so on, and the multilayered infrastructure of, for example, IP over ATM over SONET over optical. Specific networks have specific requirements that deserve to be treated in their own details and standards, which are beyond the scope of this section. However, the network management system has the same hierarchical structure, and if, for example, there are ATM elements to be managed, additional parameters would be monitored (QoS, VBR, ABR, CBR, cell rate loss, etc.) to ensure that the network complies with the service level agreement (SLA). We should, however, conclude that there is a great deal of effort to consolidate as much as possible this protocol mosaic, and many manufacturers have developed network management solutions that claim to provide an open platform that easily interfaces with different protocols so that they become transparent to the applications layer.

4.4.9.2 Internal Interfaces

Internal interfaces provide the communications link between bays, shelves, and units and their controllers. The type of information exchanged over such links is related to:

• performance parameters and quality metrics
• fault or degradation indication
• alarms
• provisioning at startup
• dynamic provisioning
• automatic discovery of components
• automatic discovery of units and unit type
• testing
• statistical data
• data store
• others, as design requirements demand

The choice of the internal interface, custom versus standard and off-the-shelf (OTS), depends on the system designer. However, this choice is not arbitrary but it is carefully made after examining the cost-performance of standard and available versus custom-designed (proprietary?). The preference is to use a standard interface such as a Gigabit Ethernet (GbE) between the main control and the shelf controllers, scaled down between shelf and unit controllers. However, the communications link between components on the same unit or within a module (a small cluster of units) may be customized to meet specific needs of the node. Typically, customized interfaces are used where standard OTS interfaces do not satisfy the design requirements of the system. In general, customized interfaces on the unit level are implemented with either ASICs or FPGAs.

All communications systems provide a craft interface (through a RJ-45 connector) used by technical personnel to "enter" the management command space of the system using a terminal (such as a portable computer with a GUI interface), known as craft interface termination (CIT). Such commands are useful to initiate system tests on the unit and shelf level and on various signal levels. Via the CIT, technical personnel can also download software upgrades from remote locations using SNMS/CIT file transfer protocol (FTP) or to manually provision various units.

4.4.9.3 Network Management

A telecommunications network consists of resources such as access systems, switching systems, cross-connects, and so on. In TMN these resources are referred to as network elements (NEs). TMN is a standard way to manage a telecommunications network, and it enables communication between operations support systems (OSS) and NEs.

In the previous section we described the five telecommunications management network (TMN) layers:

• the network element layer (NEL)
• the element management layer (EML)
• the network management layer (NML)
• the service management layer (SML), and
• the business management layer (BML)

The TMN model provides a hierarchical approach to network management and segregates the management responsibilities. This makes it possible to use different operating systems, different databases, and different programming languages. The TMN enables telecommunication service providers to achieve interconnectivity and communication across operating systems and telecommunications networks. The TMN is represented by several building blocks:

• The OS performs operations system functions, including monitoring and controlling; the OS can also provide some of the mediation, Q-adoption, and WS functions.
• The MD performs mediation between local TMN interfaces and the OS information model.
• The QA translates between TMN and non-TMN interfaces. For example, a TL1 Q-adapter translates between a TL1 ASCII message and the CMIP, the TMN interface protocol.
• The NE has a standard TMN interface and it contains manageable information that is monitored and controlled by an OS.
• The WS performs workstation functions. WSs translate information between TMN format and a GUI (graphical user interface) displayable format.
• The DCN represents OSI layers 1 to 3 and is the communication network within a TMN.

ITU-T has partitioned the general management functionality offered by systems into five key areas, Fault, Configuration, Accounting, Performance, and Security, known by the acronym FCAPS. FCAPS is not the responsibility of a specific layer of the TMN architecture but portions of FCAPS functionality are performed at different layers. For example, as part of fault management, the EML logs in detail each discrete alarm or event. It then filters this information (it abstracts) and forwards it to an NMS (at the NML layer) that performs alarm correlation across multiple nodes and technologies and root-cause analysis. In addition to fault management, there are configuration, performance, security, and accounting.

The three lower levels of the TMN model describe how a network's NEs are managed by their corresponding EMS and how EMSs are managed by an NMS. It is important to notice that each NE in their network communicates with their respective EMS. Similarly, one EMS is deployed for a group of NEs of the same type. EMSs control and manage all aspects of their domain and ensure maximum usage of available resources. From the detailed knowledge about the NEs and the network, the EMS abstracts relevant aspects, which communicates via a northbound interface to NML. Adjacent TMN layers interface to provide communications between applications. The TMN M.3010 document allows for the use of multiple protocols and thus open standards such as SNMP and CORBA. The TMN model has been effectively used to represent graphically complex relationships within network-management architectures.

The EMS plays a key role in maintaining both NEs and transmission facilities. It is the primary repository of detailed history of NE-specific faults, QoS, events, technicians' actions, and performance data. The EMS model includes service provisioning, service assurance, EMS and NE operations support, and automation enabling. Some of the tasks of an EMS are:

• Installing the NE (load parameters, auto-discover the NE equipment, establish and verify connectivity).
• Collection of data used to determine whether the service provided matches subscribers' usage characteristics and to forecast demand.
• Provisioning and planning capacity (auto-discover NE components, provide inventory information and information on available capacity).
• Upgrading the NE (auto-discover new equipment, download software upgrades, maintain concurrency between EMS and NE software and hardware releases.
• Protecting NEs and EMS database integrity (back up and restore databases, monitor loss of NE-EMS connectivity, resynchronize database when connectivity is lost).
• Service assurance to ensure that the purchased service is provisioned as agreed and delivered. This includes network maintenance and restoration and network monitoring and control.
• Periodic collection of quality metrics to characterize the performance of network resources and discover degradation trends.
• Fault management support to ensure that service remains available and at the agreed QoS. This involves the proactive monitoring of network resources to detect degradations and faults, performance, and utilization parameters.
• QoS assurance to ensure that the quality metrics characterizing the network performance remain within the agreed limits.
• Service usage to measure subscriber usage of the resources for billing. It applies only to those NEs that provide a chargeable function, such as connection and call setup.

The NML has three primary functions: fault management and root-cause analysis, integrated end-to-end service provisioning of multivendor and multitechnology networks, and integration between the EML and the SML.

• Service provisioning encompasses tasks such as equipment installation, capacity planning, capacity provisioning, and NE database integrity upgrading and protecting.

NML, SML, and BML perform high-level management processes, such as network inventory, development and planning, network configuration, and network provisioning.

• Inventory management keeps a detail record of all NE resources in the subnetwork; locations, types of equipment, model numbers, serial numbers, versions, installation dates, and more.
• Configuration management performs gross control of subnetwork resources, topologies, and redundancies. It includes the installation and turn-up of new equipment resources, the assignment of resources to trunk routes or service areas, and network protection switching. It may also include the partitioning of resources into virtual private networks (VPNs).
• Provisioning involves the creation of specific connections or the enabling of specific subnetwork features, including QoS, which are assigned to a specific subscriber for an agreed period.