Tablet computing, smartphones and the current BYOD craze have put a significant emphasis on the corporate wireless LAN network. Even in this current macro where IT spending seems to be in a bit of a lull, one of the few areas of enterprise networking spend that’s growing is Wi-Fi. A recent ZK Research survey shows Wi-Fi being the No. 1 area of spend for network infrastructure over the next 12 months. This should come as no surprise, considering the majority of devices coming into the enterprise are wireless-only.
All of this activity has given credibility to the notion that the wireless network, in many organizations, will become the primary access network, instead of an augmentation to the wired network. The question I pose is whether the Wi-Fi network is really ready to assume this role. As Peter Parker’s uncle said before he died, “with great power comes great responsibility.” I’m not convinced, at this point in time, that wireless LAN is really up to the task of operating as the primary network.
To replace the wired LAN, the wireless network needs to act like a wired LAN. Here are the main areas where I feel wireless LAN needs to improve the most:
- High-density support for devices. I think we’ve all experienced the result of inadequate support for a large number of devices. You’re working, everything is great, and then all of a sudden, your experience goes from great to unusable. You disconnect, then try to reconnect, but there are too many devices so you can’t get back on. Frustrating! Ultimately, the performance of the network should remain the same no matter the number of users or the throughput – just like a wired network. The concept of “airtime fairness” has been bantered about the industry for a while now, but no one has managed to achieve this in scale.
- Infrastructure resiliency. This has been an area of focus for the router and switch vendors for years. Make a network that achieves five-nines or as close to it as possible. Improving controller failover times through features like hitless failovers is a must. I know some vendors let you cluster controllers, but that gets fairly expensive, which is why the wired industry works with N+1 redundancy. The other factor regarding resiliency, which many people might consider more important than infrastructure resiliency, is client resiliency. Clients and applications need to perform uninterrupted in the event of a hardware failure.
- Better support for real-time media such as voice and video. Again, this has been a primary focus area for the wired network vendors because of the high amounts of IP-based multi-media traffic seen on networks today. An increase in the number of tablets, smartphones and laptops means much of this traffic is going to move to the wireless segments. Poor performance will remove much of the return on investment of these collaboration solutions or even their consumer counterparts. Additionally, the wireless network needs to maintain the state of these sessions to enable applications like Facetime and Skype to stay connected, instead of dropping when short outages occur.
- Application-specific support. To most wireless networks, all applications are the same. However, there are some applications, such as Microsoft Lync, Citrix, etc., for which some application-specific tweaks will greatly improve user experience.
Over the next 24 months I expect a large number of organizations of all sizes to consider making their wireless network the primary access network, or even the only network, if the organization is small enough. To capitalize on this, the wireless LAN industry needs to step up and start making their wireless solutions look and act like wired solutions. Vendors that can do this will stand a great chance of capturing some significant share.