Anyone who attended college in the late 1960s and early to mid-1970s should be familiar with nationwide grassroots campus organization called Students for a Democratic Society (SDS). In some institutions, membership in this group automatically identified you as an anti-establishment radical, while at other colleges it was simply a way to make some anti-war noise and perhaps get a date or two.
Today, a good number of these young and previously naive college kids are graying executives and middle managers who are more concerned with mortgage payments and lackluster IRAs. To any former radicals who are now involved in technology and network security, the term 'SDS' has taken on a whole new meaning. It is no longer associated with on-campus and rebellious rallies and maybe the burning of an effigy or two, but is the abbreviation for "structured distribution systems" -- a less political, but nonetheless important philosophy that is spreading throughout the wired and wireless enterprise networking.
In essence, an SDS approach focuses on bringing more control and management to wireless networks by gathering all of your key resources in a central point or wiring cabinet, rather than expanding them throughout your organization without a specific plan or architectural map. The concepts much more sense if you are dealing with hard-wired coaxial cabling and fiber optics rather than wireless, since you can see the physical layout of the wired system. It is a bit harder when you have to visualize the channels of communications and structure of wireless systems in order to map out the best and most manageable route to take.
In an SDS system (wired or wireless) all of the common network equipment such as routers, hubs, switches, DSL or cable modems, and so on would be located in the wiring closet. This is a comparatively simple task if you are dealing with smaller and more manageable office systems, or maybe a wireless network that is installed within the remote division of a large company. Basically, you label all your wires and connections, limit the number of hub exchanges and connect everything back to a central location. By eliminating physical layers, you can limit the possibility of unauthorized network breaches -- which, by some estimates, mostly happen at the physical layer.
Maintaining an SDS system within larger enterprises is a bit more challenging, however, especially as companies install multiple networks and sub-structures that are designed to handle such things as user access and non-direct security (via the wired backbone rather than directly connected to the wireless system). Most networks rely on a ring or distributed star approach to map out network resources, since these provide much more reliability and control. These systems usually function well when wireless networks are installed on multiple floors or throughout multiple buildings across a corporate campus. Problems can arise, however, as changes occur at the client or end-user site -- especially as the physical characteristics of the sites alter and shift.
For example, office layouts are changed to accommodate new employees or employee downsizing, company departments are shifted around in the name of productivity or executive whims, and wired and wireless networks are physically altered and moved about to accommodate the different changes in the physical look and feel of a typical office.
Also, the design of many of the wireless systems available today definitely makes it more difficult to think in terms of SDS, since these systems are specifically structured to handle user activities within their separate 300-foot bubble of communications, and then hand it off to other wireless APS and network points as an end user travels throughout a typical multi-point network. As a result, control and management becomes more difficult (although not impossible) since it usually involves installing outside systems and technology to watch over and manage total or at least partial network activity. (Such monitoring and control systems are currently available from companies like Newbury Networks, Inc., Ecutel, Inc., Mobile Automation and others -- with some offering higher and more secure degrees of protection and control than others).
Fortunately, as the designers of wireless systems think more physically and take a more wired approach to wireless systems design, a newer generation of products will become available that apply more centralized and wiring cabinet-centric controls to a network. There are a handful of companies now working on SDS-based wireless APs, and the IEEE is also developing a standard that would dictate specific management and control specifications for wireless systems. The push for such systems will also increase as the products become available later this year that make use of emerging IEEE standards that apply wireless encryption protocols to 802.11 wireless systems (presently, most Wi-Fi systems are based on wired equivalent protocols (WEP) and slightly more secure wireless standards such as WPA.
What can you do as a conscientious non-objector to higher levels of security in wireless and a more managed and centralized approach? First of all, think in terms of centralized and controlled management when mapping out your wireless system. It is really unbelievable how many network administrators and IT types think about optimal layout and design after the systems have been installed and are up and running. Take out advice: It is really a mistake to tackle system architectures, wireless AP location and control as an after-market concern. Remember, you'll pay more money to have air conditioning installed in your new car after you drive it off the lot, and you'll pay dearly if you don't have an implementation and control plan before you plug in your first wireless AP.
The second bit of advice we can give is to design and develop systems that are flexile enough to take advantage of emerging new technologies and approaches. We talk to a lot of enterprise executives who complain bitterly about being locked into a specific vendor, or stuck with a proprietary network developed by a systems integrator who was thinking about more long-term self-serving benefits than providing a workable solution for your networking problem.
In wireless networking, the enterprise buyer not only has to be aware (caveat emptor), but should also be a bit of a technology clairvoyant and networking rebel, since the best systems are those that are poised to benefit from the future and do not necessarily take the conventional approach to wireless networking.
About the author
Tim Scannell is the president and chief analyst with Shoreline Research, a Quincy, Mass.-based consulting company specializing in mobile and wireless technology and initiatives. Shoreline works with end users, looking to implement mobile solutions, and vendors, developing new products and seeking business and customer opportunities. The company also specializes in training and strategic planning projects. For more information on Shoreline Research and the company's strategic services please go to http://www.shorelineresearch.com.