Wireless adoption in the real world
Quick - what's the most popular wireless application in the enterprise today? It should come as no surprise that the answer is voice! Someone once joked that the killer app for wireless data would be voice, and I suppose once voice over IP gets established on wireless LANs and eventually in wide-area wireless networks, that will undoubtedly prove correct. But what about real wireless data - what are enterprises doing today?
Because most-wide-area wireless services have traditionally provided less than modem-like throughput, the most popular applications have been those that need little throughput and have a high tolerance for latency (the amount of time it takes to get information from one end of a network connection to the other). Latency is important because many wide-area wireless systems employ store-and-forward techniques in order to manage traffic against available infrastructure capacity. So, paging, packet radio, and other low-speed services ended up supporting messaging, very limited e-mail (no HTML or attachments in most cases), and vertical apps like field sales and service-fleet functions.
Today's 2.5G and 3G networks offer higher throughput, with increasingly-available coverage, albeit at higher prices than the older services. The core problem here is that GPRS and 1XRTT are frequently priced by the byte - which makes absolutely no sense to me. After all, how can we really control the amount of data we send, especially in a Web-services-based model? All-you-can-eat plans exist, but they are quite pricy. A notable exception is T-Mobile's $30/month unlimited GPRS plan using the popular Sidekick Color model - which is worth considering if they have coverage in places you're going to be. Such an approach could cover a huge range of applications at a very reasonable price. You might notice, however, that the Sidekick is largely a consumer-class device. It's worth noting here that the cellular carriers have a profound consumer bias in their marketing and product/service offerings. The reason for this is simple - that's where the demand is today. What I think we'll find over time, however, is that the differences between "consumer" and "business" wireless products and services blur considerably, with only marketing messages and pricing (and to some degree service and support) serving as differentiators.
What else is the enterprise doing with wireless? Consider the following examples:
- Synchronization - Today's "sometimes connected" mobile computing/communicating paradigm is based in the reality that wireless communications just aren't available everywhere we need them. As a consequence, the essential technique for mobility is to keep one's mobile device (typically a PDA, notebook, or smartphone) synchronized with one's office network resources, including e-mail and documents of all forms. For an examination of this process, see Synchrologic's white paper "Mobilizing Enterprise Applications with Synchronization".
- Instant messaging - While e-mail, paging, and related messaging technologies remain important to enterprise wireless users, they really are designed to provide what we call "temporal decoupling" between sender and recipient. That is, a variable amount of time can go by between messages, with no loss of value. We're gradually seeing, however, an increasing role for wireless instant messaging which in many ways parallels the effect that instant messaging is in general having on Internet-based communications. For an example, see http://www.air2web.com/2IM.jsp ">Air2web's 2IM product.
- Building-to-building bridges - Finally, many firms have discovered that it makes sense to use wireless links between buildings in a campus or metropolitan-area setting to connect both voice and data networks. Such point-to-point and point-to-multi-point products have performance that can match that available on wire. The general way to evaluate whether a bridge is for you is to compare the cost to lease the wireless equipment against the cost to obtain capacity from a carrier of some form. You may be surprised to find that bridges are cheaper in many cases, especially over the life-cycle of a given installation. The only other key factor is to make sure that the geometry of the proposed installation works (line of site is important, but not always required), and there may be other considerations if you use a licensed wireless technology. For examples of enterprise-class bridges, see:
And, of course, we've not even mentioned wireless LANs here, which are having a profound impact on the enterprise both in-building and via public-access hot spots. As we've noted before, WLANs and cellular offerings end up merging seamlessly over time. And that's really what we want in enterprise applications, after all - transparency, ease-of-use, and uniformity. With the rise of wireless Web services, the increasing availability of wireless broadband, and subscriber units available at prices that don't break the corporate bank, we're getting ever closer to wireless as an enterprise essential.