15.8 T1 Restoration

---

T1 facilities constitute the backbone of many private networks. Companies have a variety of means at their disposal to protect their T1 networks from congestion and link failures. Traffic can be offloaded to standby links or the switched digital services of various carriers. The CPE that handles this is usually the T1 multiplexer.

T1 multiplexers with advanced transport management systems can satisfy the instantaneous restoration requirements of the high-capacity facilities on today’s global WANs. For hybrid networks consisting of a combination of dedicated private lines and switched digital services, this means having a multiplexer that fully supports ISDN and other switched digital services. For dedicated (non-switched) networks, it is the inherent routing and rerouting capabilities that are important because they determine whether the T1 multiplexer will make a viable alternative to carrier-based restoration methods.

15.8.1 T1 Backup Via ISDN

Primary rate ISDN offers a viable solution to backing up private T1 facilities. Under ISDN, the typical time required for call setup is 3 to 10 seconds. Thus, the PRIequipped multiplexer permits traffic to be rerouted from a failing T1 line to an ISDN facility in a matter of seconds, rather than hours as is required by other restoration methods. With ISDN, the customer pays a monthly charge for the primary rate local-access channels but pays for the interoffice channels only when they are used. Figure 15.3 illustrates how primary rate ISDN can be used for backing up a T1 private line.

Figure 15.3: Configuration for ISDN backup for a T1 private line.

ISDN can also be used to back up fractional T1 (FT1) links. If a fractional link goes down, dial backup to ISDN can be implemented in a manner analogous to placing a standard telephone call. Each T1 multiplex node equipped with ISDN PRI could then be programmed with ISDN dial numbers for use in the event of a system failure. Placing ISDN calls would result in 64-Kbps or 384-Kbps channels being placed into service within a few seconds, and the network could then reroute private network voice and data traffic over the newly allocated channels.

15.8.2 Automatic Routing and Rerouting

A T1 multiplexer can automatically implement circuit routing through specific network circuits. The circuits to be routed are identified by the operator on a per-circuit or user group basis. The T1 multiplexer then chooses the best path based on a match between the circuit profile and the attributes and parameters of the aggregate.

The characteristics of each aggregate are determined during network configuration. The data entered is used in conjunction with circuit profiles to ensure optimum routing. In this way, a routing scheme is executed based on quality considerations to ensure the integrity of the applications. The quality-based parameters of each aggregate include delay, error rate, availability, and user-defined attributes.

A T1 multiplexer’s ability to perform automatic rerouting quickly and efficiently without the need for operator intervention ensures the continuous connectivity of critical applications. The T1 multiplexer maintains full optimization of all routes in conformance with the quality-based routing parameters established during initial circuit configuration. Circuits can even be down-speeded (i.e., compressed at varying rates) to ensure that all users, not just a few, continue to communicate when routes become congested.

When confronted with congestion or impending circuit failure, the system automatically calculates rerouting based on each likely failure. In the event of a failure, the system automatically recalculates optimized routing based on current network conditions. After restoration, the system again automatically calculates the best rerouting in preparation for a possible second failure on the network. The system continues this rerouting evaluation process so that it can handle each new emergency until there is not enough of the network remaining through which to reroute traffic.

To ensure that the applications of the highest priority are always rerouted over the best paths during reroutes, priority rerouting is performed on a network-wide basis, rather than on a node-to-node basis. This feature determines whether unaffected circuits will be “bumped” to accommodate higher priorities.

Because applications require different grades of service to continue operating efficiently during line failures, circuits must be routed to the best path for each application, not just switched to any available bandwidth. An application-based priority scheme ensures that the network continues to support all applications with the best response times. After the emergency has passed, the network can be manually returned to its original configuration or automatically returned on a scheduled basis.

The security of sensitive data is also protected during reroutes. During initial circuit configuration, sensitive data can be assigned to encrypted lines to ensure that there is no breach of security during line failures. Circuits that should not be involved in the rerouting scheme at all are alarmed immediately, time/date stamped, and logged to the system controller when problems occur.

Many automatic reconfiguration schemes result in service denial. However, with a down-speeding capability, T1 multiplexers can implement reconfiguration without bumping users off the network. As configured in the circuit profile, voice transmissions, for example, can be automatically down-speeded on a selective basis during automatic rerouting to ensure that full network connectivity for all applications (voice and data) is maintained. The multiplexer’s ability to support software-selectable optioning of voice-compression algorithms ensures that adaptive down-speeding can be implemented to keep users on-line during automatic rerouting instead of forcing them to get bumped off the network. This permits continued operation with efficient T1 bandwidth fills and saves money, since fewer T1 lines are required for emergencies.

The ability to reroute a full T1 circuit without timing out front-end/host sessions is a paramount concern. Some T1 multiplexers are capable of rerouting a circuit in a few seconds. EIA signaling leads remain frozen in their current state during rerouting to ensure active connections for front-end processor/host sessions during aggregate line failures. This ability to reroute nearly instantaneously ensures that network users do not have to manually restart host sessions after circuits are rerouted.

15.8.3 Dial Backup

The dial backup function of the T1 multiplexer allows the network manager to communicate from the system controller to a remote node that has become isolated from the rest of the network. The network manager can use the secondary port on the system controller as the alternate communications channel, whereas the primary port would be used for supervisory data (see Figure 15.4).

Figure 15.4: Dial backup communications link between multiplexer nodes.

To access a remote node that has become isolated, the local controller initiates dialing of a stored phone number. After establishing the dial-up link, the controller routes supervisory data, which permits the controller to communicate not only with local nodes but with the isolated remote node. When the dial backup function is no longer needed, the controller terminates the link, and the dial backup port returns to normal and is again available for test purposes.

15.8.4 Network Modeling

With the T1 multiplexer’s management system, network planners can simulate various disaster scenarios on an aggregate or node level anywhere in the network and devise effective recovery solutions. Off-line simulation allows planners to test and monitor changing conditions to determine their impact on network operations.

Models can be created that typify network failures and determine the viability of recovery strategies—all without impacting current network operations. The various models can be stored and retrieved for implementation when needed, eliminating guesswork during emergencies.

Through menu selections, the network planner can simulate aggregate failure, which then invokes automatic rerouting. During the simulation, optimum channel routing is determined from the quality-based parameters. The network planner can also simulate a node failure by forcing total aggregate failure to that node. The automatic rerouting algorithm then reroutes channels based on the entire node failing. The simulations do not impact the active network. Any of the stored reroute scenarios can be retrieved to facilitate disaster contingency planning.

Of course, this kind of network modeling can also be done with a variety of third-party tools. Although many do not provide easily verifiable data, they nevertheless make good prototyping tools for capacity planning, what-if modeling, and disaster-recovery scenario development. Some products feature a simple graphical interface, drag-and-drop features, and intuitive modeling functions that allow managers without network design experience to use them successfully. A systems administrator, for example, can use this kind of tool to help decide if switched T1, frame relay, or ATM is best for LAN interconnection over the WAN and what impact there might be on the entire network if one or more of these circuits should fail.

15.8.5 Inverse Multiplexing

Through an inverse multiplexer, a company can set up the appropriate switched digital bandwidth needed to temporarily replace failed higher-capacity private leased lines. By aggregating multiple channels available from the carrier—usually in increments of 56K/64 Kbps—users pay for the number of local-access channels only when they are used to transmit voice, data, or video traffic. Upon restoration of private facilities, the on-demand switched digital channels are taken down and carrier billing stops.

Under this bandwidth-on-demand concept, extra bandwidth can be set up to accommodate peak traffic periods, as well as to reroute traffic from failed private lines. Advantages to this approach include: the immediate availability of bandwidth; the user pays only for bandwidth used, according to time and distance; and the elimination of the need for standby links that are billed for whether fully used or not.

---