Maximizing WLAN Efficiency for the Enterprise

The amount of traffic on wireless LANs (WLANs) has skyrocketed as enterprise dependence on wireless connectivity continues to increase significantly. According to IDC, by 2009, most large corporations expect 75% of their workforce will use WLANs exclusively. The increasing dependence on WLANs is driven by the significant productivity benefits to employees who want continuous connectivity to their primary applications as they move about their office throughout the work day. The ability to stay connected and remain productive while being mobile is becoming much more a necessity than the luxury it had been in the past. However, despite the fact that WLANs improve employee mobility, they do not always deliver the productivity returns that end users expect. WLAN connections are considerably slower and less reliable with varying bandwidth than the dedicated switch connections that are also available to the same users.

Device failures and service problems on the WLAN are more problematic and difficult to solve than on the wired network for a number of reasons. WLAN technology has yet to achieve the maturity and reliability of the wired network and WLANs increasingly rely on the functionality of back-end services, which affect performance. Also, there are many bandwidth-hungry applications that traverse the network infrastructure, including streaming video, storage backup, batch processing, and more that compete for and use up available bandwidth.

These problems are exacerbated by the emergence of new devices that seek to leverage the flexibility of these networks with a class of products that are a blend between a work and personal device. The Apple iPhone is the most notable example of these hybrid devices. The iPhone has many smartphone business productivity features in addition to the usual features found in a personal electronics gadget. In the past, the iPhone was purchased by individuals who would bring them into the workplace and use them for business; however, now that AT&T has recently begun offering the iPhone as part of its business plans, it is now easier to acquire and pay for an iPhone through enterprise accounting departments.

The iPhone has the potential to create significant performance problems for enterprise WLANs and the mission-critical applications that run across them. Overloaded WLANs can be a problem for enterprises of any size in just about any industry. The challenge starts with end users who typically do not understand the potential performance issues affecting networked applications that traverse a shared WLAN with a 54Mbps connection (the majority of networks are 802.11g) vs. the dedicated 100Mbps switch connection that they would experience with a typical wired infrastructure.

The iPhone’s impact on WLANs can best be illustrated by examining a real-world example. The vice president of sales at a medical supply company gave his team iPhones as a reward for a very successful year. The sales team, excited about the new toys, immediately began to use them over the company’s WLAN network. On the afternoon of the same day, a number of these overzealous sales professionals began to download their favorite songs and iTunes videos for the long commute home. The result was a performance problem on the WLAN that flooded the helpdesk with a string of calls.

These calls are reported as problems, such as the “financial application is running slow,” “the wireless is broken”, or “my database queries are slow”, and are immediately escalated to different I.T. teams. This typically is no trivial matter as Yankee Group estimates that finding the problem is 90% of the troubleshooting effort. Several I.T. personnel are now chasing down the source of these perceived individual issues, which ultimately wastes critical time and hinders productivity. Even more troubling, when the users are done using their iPhones for bandwidth-hogging downloads, the issues “disappear” for the affected users, without ever determining what caused the problem.

From an enterprise network manager standpoint, there are limited troubleshooting tools available.  However, these tools fail to provide a broader view of what is happening with the networked applications and other elements on the network. Network managers are constantly faced with solving problems with a complete lack of background information. Enterprises need a solution that empowers I.T. organizations to better understand their networks so that they can proactively detect and control problems such as bandwidth-hogging WLAN applications in real-time.

A new approach, Rapid Problem Identification, or RPI, is a technology that addresses this issue. With this approach, the focus is on the behavior of each and every application, client, server and network within the enterprise ecosystem. This technology analyzes each and every element and looks at traditional performance variables like packet rate and bit rate but also this solution can help I.T. establish a typical profile that can be leveraged for comparison when there is a deviation within the ecosystem. It also discovers hidden problems that often plague I.T.: bad policy, unsupported applications, and excessive application use.

Higher probability symptoms are then examined across the three dimensions of time, location and application. Armed with this “big picture” of the how the various networked applications should be performing, the RPI solution then correlates and analyzes all the WLAN issues with other symptoms experienced on the infrastructure to determine the “root cause” of the situation. Using heuristic analysis, RPI can eliminate groups or symptoms that have little or no probability of pointing to the root cause. This information helps network operations personnel determine exactly who “owns” the problem, so they can escalate the issue to the right group for resolution. The result is actionable information about the source of the core problem that enables WLAN services to be quickly restored to normal quality levels.

In the case of the medical supply company, to detect problems with WLAN, the RPI solution takes a top-down approach to problem identification that provides an understanding of the application experience of the WLAN and how it differs from its normal state at the exact time that the performance issues are being reported. The I.T. department can then correlate and analyze all the WLAN-specific issues with other symptoms experienced on the infrastructure to determine a single specific problem source.  First the technology groups symptoms into logical buckets to determine whether there are application commonalities in the symptoms. 

Generally speaking, problems with the WLAN are more often than not caused by a user or pattern of users – in this case it was multiple users uploading video with their iPhones. The result is actionable information about the source of the core problem that allows for a quick problem resolution to get the company’s WLAN service up and running to normal quality levels.

The issue of degrading WLAN efficiency in the enterprise is not only tied to the iPhone. This is and will be a continuing trend in the rise of new endpoints that leverage the I.T. infrastructure for both professional and personal use.  Given this trend, I.T. needs to be actively searching for emerging solutions that help them understand and manage the behavior of all network elements across the entire enterprise ecosystem.