Balancing Reliability

Switches form the core of today's networks. Their performance and reliability have a direct effect on overall network reliability. You need switches and associated technologies to deliver uninterrupted service and dependable call handling.

The FCC tracks and rates switch performance, and issues its findings in the annual Automated Reporting Management Information System Service Quality Reports. These reports rate telecom switch reliability based on "occurrences of two minutes or more duration downtime" for carriers that generate more than $107 million in annual operating revenue.

By those governmental measurements, a new class of digital switch has emerged that delivers predictably higher levels of performance and network reliability. The best-of-breed approach employs distributed processing strategies, engineering for reliability and redundancy, innovative connectivity solutions, and a powerful suite of reliability-oriented network tools. You need to understand these new technology capabilities so that you can equip your networks to deliver the highest possible level of service.

Today's most advanced digital switching systems employ distributed processing strategies designed to offload processing from one single point to focus on higher-level operations essential to providing efficient service for large numbers of subscribers. In this approach, each hardware subsystem incorporates its own microprocessor and is programmed with the intelligence needed to perform its specific function. This modular design is used in developing switch-related hardware and software, and contributes to long-term network reliability.

Technicians must engineer switch hardware to ensure that no single point of failure can have an adverse effect on service. In service diagnostics, you must implement routine exercise tests and manual diagnostic procedures to ensure reliable fault detection, isolation and correction.

Overload performance, or a system's ability to handle sudden or excessive traffic volumes, also is a key indicator of switch reliability. Some of the industry's most advanced digital switches employ patented control mechanisms to detect impending overloads and to shift processing loads automatically from the processing elements to peripheral computing systems. This innovative strategy gives priority to in-progress calls and transactions, protects the processing elements and ensures maximum system throughput. Depending on the specific equipment deployed, switches use sync-matched core module processors to provide hot standby protection. During failures, the synchronized mate processor assumes control instantly without affecting network performance. In addition, warm-spare or load-sharing processors provide single point-of-failure protection and non-duplicated, non-sync-matched processors handle projected traffic loads.

The latest generation of switch architectures features multiple processors operating independently so single processor failure does not cause complete switch failure. Because the same hardware and software is used in each processor, code sequences can be executed on any processor with the same result. In this scenario, multiple processors also access shared memory, balancing the workload between processing elements regardless of load type, switch configuration or software variations.

ADVANCED CONNECTIVITYInternal switch messaging and connectivity also can affect overall reliability. An advanced hub bus is used to link the core module processor and the peripheral modules and other network devices. To ensure optimum reliability, this bus contains dual message switches. These switches share the load under normal operating conditions, but each is capable of supporting the entire load during network failures.

By establishing and maintaining 2-way speech and message pathways between the switch and peripheral systems, advanced switches deliver higher availability rates. If any connection fails, inactive backup connections provide instant duplicate links.

An advanced non-blocking switching matrix provides highly secure interconnections between various peripheral modules using time-division multiplexing. To deliver optimum reliability, these key interfaces make use of active independence, duplication and fiber-optic connectivity. Because all mediums entering and exiting the switching matrix carry digital information, this solution switches speech connections via a fully digital network. The threat of network blocking and crosstalk is virtually eliminated because time-division switching improves the flexibility of the switching matrix.

OTHER STRATEGIESThere are many new hardware and software solutions designed to boost switch and RF performance. Some advanced systems use a new type of network data node to further reduce processing-element workload and support increased switch dependability. A data-management peripheral is designed to handle a large volume of operations, administration, maintenance and provisioning (OAM&P) information. By distributing key data-management tasks away from the core processor, this peripheral reduces the stress on the most vital switch operations. Advanced peripherals incorporate several reliability-driven features, including mirrored disk space and redundant Ethernet interfaces for transferring billing data, enhanced file backup processes and improved log file functions.

A stratified software architecture, which separates the platform software from the application software, also contributes to overall network dependability. This modular approach leverages the value of your current software investment while supporting the smooth introduction of increasingly powerful and reliable intelligence upgrades. Automated internal checking systems built into core switch-control software confirm the system's integrity. When there are problems, this failsafe software initiates a warm, cold or reload system restart to minimize potential traffic loss and ensure optimum network stability.

The ability to test, maintain and service individual circuits improves reliability. By monitoring circuit-level performance closely, you can reduce the incidence and length of OAM&P-related downtime and prevent potentially serious network failures.

As you adjust networks to deal with increased subscriber volumes and call-handling loads, you should consider deploying sophisticated software tools such as cell tiering, adaptive channel allocation, bit-error-rate-driven mobile power control, and mobile-assisted channel allocation. These software tools improve capacity and reliability, increase spectrum efficiency, reduce co-channel and adjacent-channel interference, improve uplink performance and extend battery life.

Load balancing is another dependability-driven tactic. It ensures the proper balance of cell-site traffic among intelligent peripherals. By examining current traffic distribution and cell-site allocation, you can ensure maximum performance and prevent serious resource blocking. Path balancing, also known as talk-in/talk-out balancing, can control extraneous RF energy in a high-use wireless system. Path balancing also can reduce interference, call dragging and base-station transmission problems.

You can deploy RF-optimization studies, simulation and re-engineering strategies to increase subscriber-level network reliability. A variety of tools is available to measure and improve RF performance, including site audits, morphology drive tests, in-field path balancing and handoff optimization. You can use specialized RF-optimization techniques to improve performance in small-cluster deployments.

LOOK TO THE FUTUREAs wireless networks transport more mission-critical voice and data, switch reliability is an increasingly crucial issue. Next-generation switches feature fully fault-tolerant processing components. This new and even more reliable approach features multiple parallel processors and dynamically duplicated memory. By adding extra processors and memory cards with hot spares, the network can experience failures without reducing capacity or overall availability.

Live insertion capabilities will allow you to replace processing elements more quickly and easily, which enhances reliability by reducing downtime during routine maintenance or system failures. This flexible, on-line spare approach also supports greater network dependability by providing improved overload protection. You can expect continued refinements in reliability-related connectivity, controls and software utilities.

When evaluating potential solutions, you should consider switches that have been analyzed to the module level for predicted fail rates under expected field conditions. You should perform this analysis using a combination of empirical field data and predictive benchmarks. You also should consider the manufacturers' proven commitment to switch and network reliability. Fault tolerance is one of most important characteristics you should look for in a switch solution. Selecting a less expensive solution may seem like a good choice initially, but it may not provide built-in redundancy and scaleability. Without redundancy, you may face more system outages.

How do you know when it's time to invest in an advanced solution? If your capacity needs are increasing, and you are expanding networks rapidly, or if you are entering new markets, you are a prime candidate for an upgrade. If you want enhanced reliability, you should consider migrating to advanced switches.

Like most carriers, you are probably concerned about capital expenditure, especially because there are so many variables involved in the final price tag. But in the long run, a more dependable network means loyal subscribers, less churn and improved revenue opportunities. Today's advanced support technologies can help keep your networks up and running one step ahead of the competition.