January 2004

Wireless VoIP Inside Out

BY RICH WATSON

Have you ever been on a business trip, needed to contact an associate back at the home office and reached for your cell phone to place a call for a quick answer? If you are in a meeting inside a customer or vendor�s facilities, you might be unhappy with the results of your call. While you may not be unhappy with the quality of the answer you got from your colleague, you might be very unsatisfied by the quality of the call made to the associate. Most experienced cell phone users have found this truth -- call reception or signal quality inside of an office building is very often problematic. You may even find yourself looking for an outside window where the signal may be better to complete that important call.

Wide-area network coverage inside a building is never guaranteed. The laws of physics dictate that dense building materials (e.g., steel or brick) may block WAN signals, making access to network services either limited or non-existent. From a consumer�s perspective, the ideal situation regarding WAN services is that they should be totally seamless, equally accessible both inside and outside buildings. So, how can the communications industry address this problem and provide more reliable ubiquitous service?

One option is to bring the �outside� in. That is, extend WAN coverage by installing WAN components (or pico-cells) inside commercial buildings. This is a very straightforward solution that works, but can be very expensive, is carrier-dependent and can only be accomplished with the full cooperation of buildings� owners. This approach may provide some extended services for large public facilities, like shopping malls and school campuses, but is not a universal approach for solving all in-building coverage problems.

So, what is an alternative solution for providing WAN service coverage inside buildings? Voice over WiFi -- in the form of WLAN �hotspots� -- has caught the attention of the industry and appears to be a viable option. Assuming that WiFi hotspots can be deployed to provide connectivity to a carrier�s WAN, such services would provide an acceptable, yet relatively inexpensive, wireless infrastructure solution for support of wireless VoIP. However, this solution still imposes its own requirement: A dual mode (WAN/WiFi) phone.

Having a phone that always works inside and outside will be in high demand, but what will it take to provide such a product?

DUAL MODE PHONES
Is it possible to build a dual-mode phone using today�s technology? Yes. However, some key questions must be addressed in order to create a successful product. Besides unit cost issues, the primary consideration in developing a successful dual-mode phone is �which WAN technology -- GSM, CDMA or other -- should be used?� Another parallel question is �which WiFi technology -- 802.11a, 802.11b or 802.11g -- to marry with the WAN?� All of these technological combinations are possible, but what will drive market penetration and adoption will be defined by the configuration that affords the largest market segment (maybe the solution is to just use the carrier technology that already serves the largest market segment), as well as who is providing such a solution. Some recent product announcements indicate that it looks like wireless carriers will initially offer 802.11b WiFi (11Mbps�2.4 GHz) service in conjunction with their WAN service in support of any dual-mode devices.

Even when the right RF technologies are married, there are questions as to how the system will function in real-time use. If I�m on a call originally connected through the WAN and I walk into a building and lose WAN connectivity, will the call �roll� over to the indoor WiFi network and allow me to continue my call? If such a feature is provided, how will this operation affect my service level agreement? Will it affect my minutes? If I am connected to the in-house WLAN, can I get connectivity to the facility�s PBX? Will the phone adopt some or all of the extended telephony features provided through the PBX (e.g., call transfer, conferencing, hold, park, do not disturb, etc.)?

Another key factor that will dictate how these options may be implemented is �who owns the WLAN?� Providing ubiquitous wireless service may be difficult because of conflicting business requirements from the entity that owns the WLAN infrastructure. If the wireless network carrier owns the WLAN (e.g., commercial hotspots), they will view the WiFi network as a direct extension of their WAN and can impose certain access policies. If a business or enterprise owns the WLAN, they will view the WLAN as an extension of their corporate network and may impose a different set of access policies focused on protecting their corporate data. Simplicity in managing access policies will be the key to the success of any dual-mode product.

DEPLOYMENT MODELS
Assuming that the aforementioned technical and configuration problems can be resolved, how can these products be deployed? The implementation options are dependent on the level of convergence that is employed: converging the indoor and outdoor RF services (WAN and WiFi), convergence of network (voice and data), convergence of telephony services (carrier and PBX), or convergence in the device (voice/data). It is necessary enumerates the various complementary levels of convergence that can be represented in a total solution.

In general, the compounded intricacy of converged functionality will require more complex, yet more functionally rich solutions. The limitations of such converged solutions will be highly dependent on the cost benefit analysis of any particular market solution.

There are two basic implementation models that can be proposed:
� One Number Model: This is the configuration where a phone will always retain a single phone number and consistent feature set. This model is a simple extension of the WAN with pico-cells being deployed within a facility.
� Two Number Model: This configuration is more complex, but would have the phone adopt a second phone number and extended feature set based on a PBX that may be co-configured into the WLAN.

The first deployment model shows the convergence of in-building and outdoor WAN services and infers a dedicated WAN service wireless network is installed within a facility. The RF technology accessible from within a facility is the same as that accessible on the wide area network outside the building. This architectural approach requires no changes to the radio technology in the phone (single-mode). While this simplifies the client requirements, it does require a potentially expensive and more complex in-facility deployment of WAN RF technology. This would be a model more likely to be pursued by a WAN carrier in order to provide services in large public buildings or campuses.

An enhanced version of the one-number architecture might also provide for simple PBX integration. This convergence model of wide area and indoor telephony services highlights a collaborative offering from a carrier and a PBX vendor. In this model, a �mobility� server is required (whether an adjunct device or enhanced PBX software) to coordinate the presence information of a cell-phone within the premises. With this configuration, associations can be created between the internal desktop phone and the cell phone, making it possible for calls to a user�s desktop extension to also be directed to the associated cell phone number. Like the first model, RF connectivity inside the building still requires deployment of carrier RF technology as pico-cells to provide RF coverage.

In general, neither of these approaches is practical for broad commercial deployments because of the costs and logistics of negotiating with each facility�s network and telephony support teams. In addition, such network design configurations are not truly �converged� architectures, as the voice RF network is not converged with the in-building data network.

The second configuration can also demonstrate an architecture that can provide extended features when the cell phone is registered with the PBX. In this model, the wireless phone can adopt a two-number configuration. While retaining the proxied features of the WAN phone, the phone can take on an �internal� phone number and have access to all the PBX provided features, when registered to the WiFi-connected PBX. This �split personality� model has quite an appeal for many end-users who will receive a higher level of telephony connectivity when they are in-building. In this mode, transfers, conferencing, forwarding, and do-not-disturb features are all possible, providing greater flexibility and connectivity to the in-building associates.

A dual-mode phone version of both aforementioned configurations can also be envisioned, where the in-building RF is provided using a WiFi LAN -- a dedicated RF convergence. Seamless telephony access to and from the WAN would be managed by enhanced software from the PBX vendors that would route inbound and outbound calls from the WAN to the phone registered on the WiFi LAN. This �proxy� service would be provided as a collaborative product between PBX vendors and WAN carriers.

A dual-mode, WiFi configuration also introduces another variable: the call control protocol. The protocol used to set up and tear down a call are key to a reliable VoIP implementation. However, there are still too many call control protocols in the market to afford a simple answer to this problem. Most likely, either the integrated PBX vendor or the Wireless Internet Service Provider (WISP) will determine the call control protocol. Today, most PBX vendors support proprietary call control protocols within their VoIP solutions in order to maximize the vendor-specific value added features. However, the growing support for SIP will most likely drive any carrier offerings. Which call control protocol is required may limit the number of options for any WAN/WiFi dual-mode phone solution.

The last convergence option would be the converged use of a WiFi network for both voice and data. This model requires a much higher level of integration and management, including the proper mapping and support of security options. It is possible with today�s WiFi products to support multiple logical VLAN configurations at the RF interface, permitting logical isolation of information at its source for security reasons. Many companies view voice traffic as a lower security risk than application data and are willing to accept a lower (and more bandwidth efficient) security policy for voice. How readily accessible WAN/WLAN voice traffic with this configuration will be is dependent upon the level of security deemed necessary. This is typically dictated by who owns the WLAN. A WiFi �hotspot� solution would represent collaboration between a wireless carrier and the hotspot provider that could simplify voice/data access, whereas enterprise WiFi networks will have more stringent security requirements.

FORM FACTOR OPTIONS
The form factor of the wireless device may also have a lot to do with its commercial availability and wireless access options. A dedicated handset (e.g., wireless phone) is a much simpler device with a more straightforward access configuration requirement. This single use device typically draws its voice/data services from the WAN provider. A PDA product, being both a voice and data device, will have a more complex configuration requirement simply because of its highly converged usage model. Dual-mode radios for PCI cards are starting to enter the market and will enable almost any laptop computer to function as a WAN or WiFi telephony device. Such devices can support more sophisticated data applications and thus may have tighter security requirements, as protecting a corporation�s persistent data is of key concern to CIOs. Because of this security concern, enterprise voice/data solutions may not be easily applied to a generic market solution (e.g., �hotspot�) as the two solutions may impose divergent security policies.

MARKETING CHALLENGES
Simplicity in installation and use is another key element in the success of any product and is particularly critical for a multi-mode phone product. Any WAN/WiFi phone solution must be simple to install and use or the user community will not accept it, no matter how feature-rich it may be. WAN/WiFi product offerings will be multi-vendor by their nature and any successful solution will involve a combination of components from several sources, for example:

Mobile Unit Vendor <--> WLAN Vendor <--> PBX Vendor <--> WAN Carrier

Any successful offering may involve products from four or five different companies. But, how will a consumer identify and select such product solutions? The challenges of launching and supporting products will be a collaborative effort between these cooperating companies. To thrive, WAN/WiFi products will have to be promoted and sold as a solution from a single source (such as a carrier or PBX vendor) with support for all components provided through a single service agreement. Without such simplicity, having a successful WAN/WiFi phone solution will be difficult.

SUMMARY
The idea of having a wireless phone that functions across multiple wireless technologies is a grand idea. The consumer demand for such a product is real and a dual-mode product (WAN/WiFi) fills the bill. The question remains when such a product will be available? Hardware advances, such as chips that support both WiFi (802.11b, 802.11g and 802.11a) and WAN wireless connections, will bring such products one step closer to reality, so look for offerings from major carriers and/or PBX vendors in the next 18 to 24 months.

In evaluating these new products, it will be important to obtain a clear understanding of their feature sets and what constitutes an end-to-end solution for a particular market segment or enterprise deployment. There will be lots of choices. Make sure to completely investigate them, as not all features may be present, nor will ubiquitous seamless roaming be supported in many initial product offerings. However, the combination of technology advances, converged product architectures, and collaborative product development efforts between major vendors will yield wireless telephony products that can be used �inside and out.�

Richard Watson is director of telephony product marketing for Symbol Technologies� Wireless Systems Division in San Jose, CA. Prior to taking on the marketing role for Symbol�s NetVision family of WiFi Telephony products, he managed the software engineering team for three years and was responsible for developing Symbol�s WiFi Telephony products.

[ Return To The January 2004 Table Of Contents ]