Handover Roaming

Handover roaming itself comes in two flavors. Horizontal handover roaming refers to the technical ability to handover either from one access point to another within the same technology (i.e. 802.11a to 802.11a, or 802.11b to 802.11b).. On the other hand, being able to handover between two different types of networks is called “vertical” roaming, e.g., from 802.11a to 802.11b or from any WLAN technology to GPRS/UMTS/CDMA. There are many intricate technical issues that make vertical roaming difficult. These issues include what happens when you roam onto a network whose quality is either higher or lower than the quality of the network you were just in. How do you navigate varying qualities of service? Most difficult of all, how to you navigate different roaming policies and prices? 

Currently, the only way to accomplish multi-access handover in an elegant fashion is to employ Mobile IP v6. Several companies offer a type of node that facilitates the use of Mobile IP in a vertical roaming situation. Ericsson, for example, demonstrated a node called the “Wireless Server Node” in Hong Kong in June of 2001, which precisely accomplishes this. The bigger question for an operator is whether they really need to deploy such a node at the early stage of the WLAN implementation, or just use the existing AAA to create a virtual “one network” feel. Deploying a node that accomplishes vertical roaming is certainly a benefit, and it’s up to the operator as to how and when to integrate such a node, and to decide if they need it.  There are many in the industry who feel that to have a true mobile Internet, you must integrate WLAN into 3G systems. If you have enough customers who want to go from their office to a cab to the airport hotspot, without session interruption, then it definitely is a good idea. The use of Static IP versus Mobile IP will continue to be  debated. But in today’s hotspot landscape, it appears that most hotspot users power down, go to a hotspot, and then power up again, thus making Static IP sufficient.

What is extremely important to the business proposition is that the end-user receives one bill for both their WLAN and GPRS/UMTS/CDMA service. This can be readily accomplished via the AAAnode (authentication, authorization, and accounting) whose main providers Bridgewater Systems (Canada) and Funk Software (U.S.) have already integrated advanced WLAN handling features. Through the AAA, billing can be accomplished as well as QoS, policy management, and many other important items to running and operating a WLAN network, whether or not it is vertically integrated with the WAN.

Contractual Roaming

Contractual roaming is the most important issue facing the hotspot operator today.  Contractual roaming means the ability to use hotspots that do not belong to your service provider while still getting one bill.  Not many people outside the world of telecommunications realize how complex it is to hammer out agreements for roaming. There are clearinghouses, lawyers and many people involved in the process of putting even one agreement together. The Achilles heel of CDMA has been the lack of such clearinghouses, thus creating a burden of “contract by contract” negotiation for roaming agreements. In the hotspot world, this is a also a great difficulty, one that must be solved before the hotspot phenomenon truly becomes a realistic business proposition. There are hotspot operators who have roaming agreements, but only a few.

Some hotspot operators are operating in a “control everything” vacuum and seem to feel they can do it alone. For others, the simple lack of a real movement or sufficient capital is hindering them. There are aggregators such as Boingo and Yellowspots who forge agreements between several operators and provide a form of unified billing to the customer, thus providing virtual roaming landscapes. This is certainly a step in the right direction, but it is vulnerable to the difficulties of negotiating qualities of service and other items that will only become more complex as the networks evolve.