Tuesday, May 10, 2011

Structured Cabling and Standardization

Typical business environments and requirements change quickly. Companies restructure and reorganize at alarming rates. In some companies, the average employee changes work locations once every two years. During a two-year tenure, a friend changed offices at a particular company five times. Each time, his telephone, both networked computers, a VAX VT-100 terminal, and a networked printer had to be moved. The data and voice cabling system had to support these reconfigurations quickly and easily. Earlier cabling designs would not have easily supported this business environment.

Until the early 1990s, cabling systems were proprietary, vendor-specific, and lacking in flexibility. Some of the downsides of pre-1990 cabling systems included the following:
  • Vendor-specific cabling locked the customer into a proprietary system.

  • Upgrades or new systems often required a completely new cabling infrastructure.

  • Moves and changes often necessitated major cabling plant reconfigurations. Some coaxial and twinax cabling systems required that entire areas (or the entire system) be brought down in order to make changes.

  • Companies often had several cabling infrastructures that had to be maintained for their various applications.

  • Troubleshooting proprietary systems was time consuming and difficult unless you were intimately familiar with a system.
Cabling has changed quite a bit over the years. Cabling installations have evolved from proprietary systems to flexible, open solutions that can be used by many vendors and applications. This change is the result of the adaptation of standards-based, structured cabling systems. The driving force behind this acceptance is due not only to customers but also to the cooperation between many telecommunications vendors and international standards organizations.

A properly designed structured cabling system is based around components or wiring units. An example of a wiring unit is a story of an office building, as shown in Figure 1. All the work locations on that floor are connected to a single wiring closet. All of the wiring units (stories of the office building) can be combined together using backbone cables as part of a larger system.

Figure 1: A typical small office with horizontal cabling running to a single wiring room (closet)

This point bears repeating: a structured cabling system is not designed around any specific application but rather is designed to be generic. This permits many applications to take advantage of the cabling system.

The components used to design a structured cabling system should be based on a widely accepted specification and should allow many applications (analog voice, digital voice, 10Base-T, 100Base-TX, 16Mbps Token Ring, RS-232, etc.) to use the cabling system. The components should also adhere to certain performance specifications so that the installer or customer will know exactly what types of applications will be supported.

A number of documents are related to data cabling. In the United States, the standard is ANSI/TIA-568-C, also known as the Commercial Building Telecommunications Cabling Standard. The ANSI/TIA-568-C standard is a specification adopted by ANSI (American National Standards Institute), but the ANSI portion of the document name is commonly left out. In Europe, the predominant standard is the ISO/IEC 11801 Ed. 2 standard, also known as the International Standard on Information Technology Generic Cabling for Customer Premises.

When Is a Standard Not a Standard?
In the United States, a document is not officially a national standard until it is sanctioned by ANSI. In Canada, the CSA is the sanctioning body, and in Europe, it is the ISO. Until sanctioned by these organizations, a requirements document is merely a specification. However, many people use the words specification and standard interchangeably. (In Europe, the word norm also comes into play.) Just be aware that a "specification" can be created by anyone with a word processor, whereas a national standard carries the weight of governmental recognition as a comprehensive, fair, and objective document.

These two documents are quite similar, although their terminology is different, and the ISO/IEC 11801 Ed. 2 standard permits an additional type of UTP cabling. Throughout much of the rest of the world, countries and specification organizations have adopted one of these standards as their own.

Cabling Standards: A Moving Target
Briefly introduces the ANSI/TIA-568-C and the ISO/IEC 11801 Ed. 2 standards, but it is not intended to be a comprehensive guide to either. Networking vendors and specifications committees are figuring out ways to transmit larger quantities of data, voice, and video over copper and fiber-optic cable. Therefore, the requirements and performance specifications for the standards are continually being updated. If you are responsible for large cabling-systems design and implementation, you should own a copy of the relevant documents.

Most of the TIA/EIA documents mentioned are available for purchase through Global Engineering Documents at (877) 413-5184 or on the Web at http://global.ihs.com. Global Engineering Documents sells printed versions of the ISO, TIA, EIA, and ETSI specifications, as well as others. The ITU recommendations are available for purchase from the ITU's website at www.itu.int/.

CSA International Standards documents are available from the CSA at www.csa.ca.

Friday, May 6, 2011

Data Cabling

The Bottom Line

Identify the key industry standards necessary to specify, install, and test network cabling Early cabling systems were unstructured and proprietary, and often worked only with a specific vendor's equipment. Frequently, vendor-specific cabling caused problems due to lack of flexibility. More important, with so many options, it was difficult to utilize a standard approach to the design, installation, and testing of network cabling systems. This often led to poorly planned and poorly implemented cabling systems that did not support the intended application with the appropriate cable type.
Master It In your new position as a product specification specialist, it is your responsibility to review the end users' requirements and specify products that will support them. Since you must ensure that the product is specified per recognized U.S. industry standards, you will be careful to identify that the customer request references the appropriate application and cabling standards. What industry standards body and standards series numbers do you need to reference for Ethernet applications and cabling?
Understand the different types of unshielded twisted-pair (UTP) cabling. Standards evolve with time to support the need for higher bandwidth and networking speeds. As a result, there have been many types of UTP cabling standardized over the years. It is important to know the differences among these cable types in order to ensure that you are using the correct cable for a given speed.
Master It An end user is interested in ensuring that the network cabling they install today for their 1000Base-T network will be able to support future speeds such as 10Gbps to a maximum of 100 meters. They have heard that Category 6 is their best option. Being well versed in the ANSI/TIA-568-C standard, you have a different opinion. What are the different types of Category 6 cable and what should be recommended for this network?
Understand the different types of shielded twisted-pair cabling Shielded twisted-pair cabling is becoming more popular in the United States for situations where shielding the cable from external factors (such as EMI) is critical to the reliability of the network. In reviewing vendor catalogs, you will see many options. It is important to know the differences.
Master It Your customer is installing communications cabling in a factory full of stray EMI. UTP is not an option and a shielded cable is necessary. The customer wants to ensure capability to operate at 10GBase-T. What cable would you recommend to offer the best shielding performance?
Determine the uses of plenum- and riser-rated cabling There are two main types of UTP cable designs: plenum and riser. The cost difference between them is substantial. Therefore, it's critical to understand the differences between the two.
Master It Your customer is building a traditional star network. They plan to route cable for horizontal links through the same space that is used for air circulation and HVAC systems. They plan to run cable vertically from their main equipment room to their telecommunications rooms on each floor of the building. What type of cable would you use for:
  1. The horizontal spaces
  2. The vertical links
Identify the key test parameters for communications cables As you begin to work with UTP cable installation, you will need to perform a battery of testing to ensure that the cabling system was installed properly and meets the channel requirements for the intended applications and cable grades. If you find faults, you will need to identify the likely culprits and fix it.
Master It Crosstalk is one of the key electrical phenomena that can interfere with the signal. There are various types of crosstalk: NEXT, FEXT, AXT, among others. This amount of crosstalk can be caused in various ways. What would you look for in trying to find fault if you had the following failures:
  1. NEXT and FEXT problems in 1Gbps links
  2. Difficulty meeting 10Gbps performance requirements

Tuesday, May 3, 2011

Attenuation-to-Crosstalk Ratio (ACR)

Attenuation-to-crosstalk ratio (ACR) is an indication of how much larger the received signal is when compared to the NEXT (crosstalk or noise) on the same pair. ACR is also sometimes referred to as the signal-to-noise ratio (SNR). It is a calculated value; you can't directly measure ACR. Also, despite the name, it's not really a ratio. It is the mathematical difference you get when you subtract the crosstalk value from the attenuation value at a given frequency. Note that ACR, and its power-sum equivalent, PSACR, is not a required parameter in TIA/EIN-568-B. Technically, SNR also incorporates not only noise generated by the data transmission but also outside interference. For practical purposes, the ACR and true SNR are functionally identical, except in environments with high levels of EMI.

KEY TERM: headroom 
Because ACR represents the minimum gap between attenuation and crosstalk, the headroom represents the difference between the minimum ACR and the actual ACR performance values. Greater headroom is desirable because it provides additional performance margin that can compensate for the sins of cheap connectors or sloppy termination practices. It also results in a slight increase in the maximum bandwidth of the cable.

The differential between the crosstalk (noise) and the attenuation (loss of signal) is important because it assures that the signal being sent down a wire is stronger at the receiving end than any interference that may be imposed by crosstalk or other noise.

Figure 1 shows the relationship between attenuation and NEXT and graphically illustrates ACR for Category 5. (Category 5e, Category 6 and 6A would produce similar graphs.) Notice that as the frequency increases, the NEXT values get lower while the attenuation values get higher. The difference between the attenuation and NEXT lines is the ACR. Note that for all cables, a theoretical maximum bandwidth exists greater than the specified maximum in the standards. This is appropriate conservative engineering.

Figure 1: Attenuation-to-crosstalk ratio for a Category 5e channel link
Solving problems relating to ACR usually means troubleshooting NEXT because, short of replacing the cable, the only way to reduce attenuation is to use shorter cables. PSACR is similar to ACR but is the difference between insertion loss and NEXT.