One Size Does NOT Fit All
The $7.2 billion of broadband stimulus funding in the American Recovery and Reinvestment Act (ARRA) represents a historic opportunity to grow the availability and penetration of broadband service in rural America. In fact, the rural market has already shown strong demand. According to a recent study by comScore, U.S. rural markets experienced a 16 percent increase in broadband penetration from the second quarter of 2007 to the same period in 2009. This growth actually exceeded that of metropolitan and micropolitan areas.
Whether using broadband stimulus or other funding sources, those who plan to deploy next-generation broadband networks have an opportunity to take advantage of the lessons learned from similar deployments over the past several years. One of the major lessons learned is the importance of developing and implementing a well thought out test strategy as part of the planning and execution of any broadband project.
Why Develop a Test Strategy
A cursory analysis of the nearly 2,200 applications totaling almost $28 billion submitted for the first round of stimulus funds reveals that there is no “one size fits all” approach to broadband stimulus projects. There is a mix of Last Mile, Middle Mile, and broadband adoption project proposals. The Last Mile applications span an array of different broadband technologies including fiber-to-the-home (FTTH), broadband over power line (BPL), DSL, wireless, satellite, and others.
Similarly, there is no “one size fits all” approach to developing and implementing a sound test strategy for the project, but there are some common threads as well as a clear need to develop a plan. The ARRA has clear language around the goal of funding projects that are viable and sustainable in the long-term. Whether using stimulus dollars or not, happy customers and a healthy balance sheet are common goals for any broadband deployment. Having a sound network test strategy for both installation and maintenance is one key component of attaining these goals.
A test strategy should focus on maximizing customer quality of experience (QoE) by ensuring quality of service (QoS), whether it is voice, video, or data, across the network to the customer’s display. It should also enable greater efficiency throughout the network lifecycle from construction through installation to ongoing operation and maintenance.
Developing a Test Strategy: Systematic Approach
The first step in developing a test strategy is to map out exactly what will be deployed and how. What services will be offered? What technologies and architectures will be employed to deliver these services? How will the network and service offerings be deployed and who is responsible for what tasks at each phase of the network lifecycle? Using this information, the next step is to identify the key focal or hand-off points, both from a network (physical and logical) and from a work-process perspective, and anticipating what can go wrong.
If video service is distributed from a head-end across a fiber-based transport network to several serving offices and then delivered across a passive optical network (PON) architecture to the subscriber, it is important to verify the quality of the video content from the content provider at the head-end before distributing it across the network. At the network edge or serving office, it is useful to determine prior to distribution across the PON network to the subscribers if any video problems are being created during transport from the head-end.
If contractors are used for the construction of the PON portion of the network, it is critical to have well-defined, quantitative acceptance criteria to ensure the quality of the fiber plant before the installation of network equipment, provisioning, and beginning customer installs.
Identifying these key points, the associated risks and implementing controls in the form of test processes enhances efficiency during network deployment. In the example above, it is much easier to proactively verify the fiber plant prior to installing equipment and turning up the network than it is to identify, locate, and clear faults or problems in parallel with customer service installations.
Another benefit is that when problems do arise, it is easier to sectionalize or isolate the source, thereby improving efficiency as well as responsiveness to customers. For example, it would be a waste of time and resources to send a technician to the customer’s residence to troubleshoot a video problem that originates in the serving office or the transport segment of the network.
The exact test processes or most appropriate test equipment will vary with the network size, topology, and composition of the workforce responsible for the deployment. The common thread however, is the need for a systematic test strategy that addresses service delivery end-to-end across the network and throughout each phase of the network lifecycle.
Developing a Test Strategy: Multi-Layered Approach
Another common thread in formulating a test strategy is the need to take a multi-layered approach. While it is clearly not necessary to test every layer of the stack at every test point across the entire network to ensure end-to-end QoS, it is critical to recognize the different layers and interrelationships between them at key points across the network.
Figure 1. IP Video Quality Layer Model
The model for IP video shown in Figure 1 provides a good illustration. This model objectively maps QoS parameters and QoE indicators and then organizes them into 4 quality layers. The next step is to apply this model to the different segments of the network as shown in Figure 2. This helps determine which metrics are most critical to test at different points across the network.
Figure 2. IP Video Network Segment Quality Model
For example, Figure 2 maps content quality to the network segment because insertion of local programming and video on demand (VoD) represent effective changes to the content relative to what was present and previously tested upstream at the head-end. Therefore, it makes sense to verify the content layer prior to distribution across the access network.
Conversely, since the access network is just a transport mechanism for the video service, content quality is not mapped to this segment. Using this model, if video stream or transport errors are present on the access segment, the next step is to drill down to the access transport (xDSL, PON, etc.) and physical layers for trouble isolation.
A helpful general rule of thumb for installations is to start at the physical layer and work up. For troubleshooting, however, begin at the service layer and work down towards the physical layer.
For instance, if deploying higher-speed Internet service over asymmetrical digital subscriber line (ADSL2+), a relatively simple series of tests can be used to ensure the copper pair is within limits and free of impairments prior to connecting the pair to the digital subscriber line access multiplexer (DSLAM) and going live with the service to the customer. There are situations where it is logical to skip a layer when there is information readily available at the next layer that provides the required level of insight.
For example, if a customer complains of degraded video with an IP video service provided over DSL and excessive errors and loss of synch are clearly evident on the DSL circuit, it is not necessary to start out by looking for video continuity errors or packet loss. In this scenario, it is better to conduct video QoS test following resolution of the physical layer issue causing the DSL errors. This is because some trouble scenarios can have more than a single root cause. Failure to spend the few extra minutes required for a thorough verification can result in unnecessary additional truck-rolls.
In summary, a well-developed test strategy systematically encompasses an end-to-end view of service delivery across the network and incorporates a multi-layered approach to testing. Testing in each key segment of the network should be appropriate to the type of service provided, the technology and architecture employed for delivery, and how the network is deployed and maintained.
Again, when it comes to the details of what tests and equipment are required, there is no "one size fits all" solution. Furthermore, taking this recommended holistic and systematic approach should equate to more efficient networks as well as both time and cost savings in the long-term.
Inspect Before You Connect
The use of fiber connectors throughout the network is a common and growing practice. Today’s connector design and production techniques have eliminated most of the problems to achieving the core alignment and physical contact required for an efficient connection, but contamination remains a significant problem. In fact, for many service providers, contamination is the number one source of physical layer troubleshooting.
Figure 3. Fiber Contamination and Its Effect on Signal Performance
A single particle mated into the core of a fiber can cause significant back reflection and insertion loss, enough to impair the signal as shown in Figure 3. Additionally, connectors are much easier to clean prior to mating and before embedding debris into the fiber. Mating of contaminated connectors can also lead to damaged equipment. The simple and cost-effective solution to this problem is to use a fiber microscope to simply Inspect Before You Connect and proactively clean any contamination prior to mating connectors.
Finding the Root of the Problem
These quick and easy-to-perform steps are essential so that if problems arise during the installation, issues with the fiber plant have already been ruled out and the root issue resides either inside the home or upstream in the network.
Step 1. Prior to connecting network equipment and beginning to turn-up service to customers, the quality of the fiber plant should be verified. Typically, all fibers coming from the exchange or office are connected to splitters at the fiber distribution hub (FDH). An optical time domain reflectometer (OTDR) measurement from the office or location where the optical line termination (OLT) unit will be placed is recommended to verify splice quality and to ensure all fibers are connected through to FDH.
Step 2. A measurement at 1310/1550 nm is then performed from the hub downstream to check fiber and splice quality.
Step 3. After the OLT has been connected to the fiber plant and during the customer installation process, the proper optical power levels should be verified both at the curb or junction box. This should be done prior to connecting the drop cable at the side of the house and prior to connecting the drop to the optical network terminal (ONT).
Conquering the Last Hundred Feet of the Network
Whether the complaint is related to no data, slow data connection, voice problems, or video impairment, an increasing number of customer QoE-related calls are isolated to the customer premises -- as high as 50 percent in some cases. However, most root causes are issues related to: equipment configuration or settings problems, firmware issues, “pilot error”, and wireless performance or set-up problems.
In fact, about 20 percent of customer complaints involve issues related to in-home wiring or cabling problems. At the same time, the customer premises can be the most costly domain in which to troubleshoot problems. Therefore, regardless of the network size, unnecessarily sending a technician to the customer premises or spending excessive time troubleshooting within the home should be avoided.
One key lesson learned in this area is that the customer premises equipment (residential gateways, ONT, set-top-boxes, etc.) are becoming increasingly intelligent and often provide in-home network statistics to aid with identifying and diagnosing problems. This capability can be leveraged locally to reduce in-home troubleshooting time and, where possible, remotely via the sharing of this information upstream in the network (enabled by standards such as TR-069) for remote diagnostics and more proactive responses.
The second lesson learned is that for the roughly 20 percent of the problems associated with in-home wiring/cabling, a good wiring tool saves time during installation and troubleshooting. During video service installation in residences with separate coaxial cable runs to multiple rooms, it is typical to map which runs connect to which rooms prior to connecting the residential gateway or ONT. While mapping where each run goes, performing a quick check to ensure that each cable is free of impairments such as cheap splitters, nested splitters, broken shield, bad or corroded barrels and connectors (among other issues) saves time and effort later versus a plug-and-pray approach.
Figure 4. In-Home Coaxial Cable Mapping
Figure 4 shows an example of proactively combining the processes of mapping where each cable run goes and checking for common impairments. It is much easier to proactively identify, locate, and clear cable problems prior to connecting equipment than trying to go back and troubleshoot after the fact.
Metropolitan, Micropolitan: ...The term "core based statistical area" (CBSA) became effective in 2000 and refers collectively to metropolitan and micropolitan statistical areas…The 2000 standards provide that each CBSA must contain at least 1 urban area of 10,000 or more population. Each metropolitan statistical area must have at least 1 urbanized area of 50,000 or more inhabitants. Each micropolitan statistical area must have at least 1 urban cluster of at least 10,000 but less than 50,000 population. Source: http://www.census.gov/population/www/metroareas/aboutmetro.html. November 19, 2009.
About the Author
Jon Beckman is JDSU director of strategy for test instruments, Communications Test & Measurement. With more than 10 years in the communications test and measurements industry, Jon has served in product management and marketing roles with a focus on access network and broadband technologies and applications. JDSU enables broadband and optical innovation in communications, commercial, and consumer markets, and is a provider of communications test and measurement solutions and optical products for telecommunications service providers, cable operators, and network equipment manufacturers. For more information, visit www.jdsu.com.