Peter Orne Wireless Government Subscribe to Newsletter Tell a Friend Print this Page 05/09/2006 Broadband-Wireless Networks: 10 Do's & Don't's NetLogix, a broadband-wireless consulting firm in San Diego, leads a daylong professional-development training seminar for local-government professionals at the W2i Digital Cities Convention . The lead presenter at these sessions is Scott Akrie, President and Managing Partner, and a 25-year veteran of the telecommunications and wireless-networking industry. Akrie offers several Do’s and Don’t’s for project planners and implementers of broadband-wireless infrastructure deployments, placing a special emphasis on training and process to avoid mistakes down the road. Our clients quickly learn one important attribute about NetLogix—We are brutally honest. We tell them what they need to hear, which often isn’t what they want to hear. Our blunt, upfront approach may be shortsighted for a firm that earns its keep by talking to people. But we see a greater issue at stake for our clients: Their survival in a new and challenging industry. As an industry, broadband wireless is still in its infancy, and the availability of equipment has outpaced the development of expertise capable of correctly installing these networks. Ironically, there is so little expertise that it’s very difficult to differentiate those who know what they’re doing from those who don’t. Today, several deployed networks don’t work as well as they should, and we’ve seen the same equipment performing very differently from one network to the next. On the basis of our experience and approach at solving problems that arise with these networks, NetLogix has created a list of Do’s and Don’t’s for project planners and overseers: 10. Poor Spectrum Planning and Management Do manage your RF well. Just because a license isn’t required to operate over some radio frequency (RF) spectrum, you still need to do good spectrum management. Do be responsible. Even if your equipment appears to have an automatic channel selection feature, do an initial analysis and ensure that this “channel selector” works the way you expect it to. Develop a channel plan and coordinate with other wireless ISPs in your area. A primary cause of RF interference is “self-interference.” 9. Unrealistic Expectations Don’t overdo it. Many local-government requests for proposals (RFPs) are generated with requirements for services and applications that are unrealistic today. Even in cases where applications and services are realistic, many responses are coming back with the wrong equipment to meet the requirements. If you’re going to generate an RFP, know what you’re talking about, or hire someone who does. 8. Incorrect Equipment Selection Don’t make poor equipment purchasing choices—a common result of lack of expertise and experience, or falling for marketing hype. Be sure you understand what vendors’ equipment offerings have in common and how they differ from one another. All equipment works differently. Some equipment will be superior for you depending on your specific network and application requirements. A common mistake we see is that service providers make technology selections without visiting networks to test those technologies. 7. Poor Processes Do create a good roadmap. So many networks are being installed with no processes—or with poor processes. Process is not rocket science but a function of experience. Don’t let the blind lead the blind. Develop a “Train the Trainer” philosophy. Beginning with the right processes will save you from reverse engineering problems down the road. 6. Overconfidence and Arrogance “I installed a Wi-Fi network at home, so deploying a hot zone should be a walk in the park.” Not quite. Therefore, don’t overestimate yourself. Engineering a three-node hot zone is far less complicated than engineering a one-square-mile hot zone, which is exponentially less complicated than engineering (and managing) a citywide Wi-Fi network. A novice can’t build a carrier-class network. Most networks are architected without carrier class in mind. We’re amazed at the number of networks we come across with no network monitoring in place. Which leads us to…. 5. Lack of Effective Network Management Do monitor your network. A broadband-wireless network is dynamic. While each radio tends to be static, the airwaves are not. We have yet to find anyone that has taken the basic steps to monitor their network appropriately from Day 1. Few even know the right questions to ask. We see RF engineers tasked with network design and IP engineers tasked with NOC design—only you need cross-discipline experience to develop the network right. Also, don’t assume that any vendor’s NMS, OSS, EMS or whatever-they-want-to-call-their-management-software will give you all the tools you will require. Remember, only part of the network is wireless. Additionally, ICMP (a.k.a. PING) is not enough to understand the health of your network. Focus on SNMP management capabilities. Lastly, measure, trend, and threshold network performance—and have someone monitor and review the data. 4. Poor Network Design Don’t fail on design. Remember the concept of TQM, or Total Quality Management? Initially, it was difficult to assess the cost of poor quality. But here’s a way to think about cost. The most challenging of situations that we go into requires us to charge time and materials to reverse-engineer a network just to figure out the best way to leverage what’s been done so that we can redesign it in a way that works. Do it right the first time. Trust us when we say that it will cost you less in the long run. 3. Lack of Clear Objectives Do know why you’re getting into this. So many jurisdictions want a broadband-wireless network, but how many know why, where, and what for? Start small and capture the metrics you need to support your decisions. The pilot projects we’ve engineered have all provided supporting evidence in the form of an increasing take rate, positive feedback, network reliability, and a variety of other metrics. Remember, you didn’t just jump into the deep end of the pool when you first learned to swim. 2. Incomplete Budgets and Business Plans Do think realistically about funding. We believe in developing sound budgets and business plans. But a business model is different from real life. Regardless of what you want, you’re often going to need more (more resources, more radios, more money, etc.). You should also consider the alternative, which is to do less. Remember, you cannot change the laws of physics. Many business plans often do not accurately account for real-world equipment performance. And the No. 1 lesson learned is: 1. Lack of Expertise and Training Do get training ! We’re falling down as an industry on this one. No vendor that we have seen has adequate training. In the vendors’ defense, they’re not carriers, so how would they know? NetLogix has developed training classes to fill the void, but we’re a long way off. As a point of reference, telecommunications just started being taught in colleges within the past decade or so. Previously, you had to be invited to Bell Labs or Bellcore if you wanted decent training—often for months. We need some rigor at the foundation of this industry. Remember, the larger the company, the more people you have working on your project who potentially don’t know what the heck they’re doing. back Related Items: • Groningen • W2i Finalizes Program Agenda for Digital Cities Convention in Washington, DC • City Services Drive Network in St. Louis • New-Technology Rollouts: Avoid Putting Lipstick on the Pig! • Doug Townsend, IT Director, Medford (OR) - Part 2 Comments No records were found. Post new comment: Only registered users can add comments. Please Log-in

• MORE BLOGS