All dual-mode phones are NOT alike

A common misconception is that all dual-mode phones are the same. Many people consider Wi-Fi enabled phones to be expensive and have poor battery life.

I’d like to dispel this myth by making a distinction between UMA and non-UMA enabled devices. I concur that non-UMA dual-mode handsets are often higher end (and therefore more expensive) and have poor battery performance (when in Wi-Fi). However, UMA-enabled devices are often mid-tier feature phones with respectable if not ‘stellar’ battery performance.

Fundamentally, UMA is a technology for mobile operators. It is designed to make Wi-Fi (and IP and broadband) ‘friendly’ to the mobile operator. So, what is UMA’s secret?

When a UMA-enabled dual-mode phone enters a known Wi-Fi coverage area (home, office), the phone automatically and seamlessly switches from the GSM network to the Wi-Fi network. Now the connection between the phone and the mobile network is over the public internet rather than the GSM radio access network. Because it’s not used when the phone is in Wi-Fi coverage, the GSM radio is put into a sleep/hibernate mode.

If/when a call arrives for the handset, the call is routed through the UMA tunnel over the internet to the handset, and the GSM radio is not used.

This is one key element of improving battery performance in UMA-enabled handsets. By putting the GSM radio in a sleep mode, the phone continues to run a single radio. Anyone who has run their Bluetooth radio all day knows that having two radios on in a phone significantly impacts the battery performance.

Contrast this with the operation of a non-UMA dual-mode phone. As the phone enters a known Wi-Fi location, generally the user needs to turn on or enable the Wi-Fi radio. Primarily this is because the battery impact is so great that ‘automatic’ Wi-Fi usage would have the un-intended consequence of sucking the battery dry.

Now with the Wi-Fi radio is operating and associated with an access point. But to make or receive GSM calls, the GSM radio must remain on as well. In a non-UMA device, the only connection between the mobile core network and the handset is over the existing GSM link. In this case, Wi-Fi is simply a bolt-on to the phone’s primary function as a GSM communications device.

There are many implications of this ‘bolt on’ approach.

First, people with non-UMA devices don’t use the Wi-Fi connection much if at all. The performance hit to the handset is too great for regular usage. Certainly with iPhone, the largest selling dual-mode device on the market, most people tend to use the EDGE data connection rather than Wi-Fi.

Second, there is no session mobility for non-UMA devices. Start surfing on the Wi-Fi connection, and if you walk out the door to GSM, that connection is lost. Wi-Fi and GSM are typically completely different subsystems on a non-UMA device. Of course with UMA, there is full session mobility for voice, data and IMS applications between Wi-Fi and the GSM network.

Third, because Wi-Fi is typically bolted onto a non-UMA device for basic data services, the Wi-Fi sub-system is not optimized for a mobile device. This was a common problem for the first UMA-enabled handsets. The Wi-Fi subsystems were simply ported from laptops and behaved like they were running on (relatively) power in-sensitive devices like PCs rather than on handsets.

Companies like NXP have invested tremendous resources in optimizing Wi-Fi (radio, stacks) to be voice-centric rather than laptop/PC centric. This optimization has paid off well, as the performance of their t409 product achieves 8 hours of talk time in Wi-Fi, according to one analyst.

Many see dual-mode devices as the next growth opportunity for Wi-Fi. This is certainly true. But there is a big difference between non-UMA and UMA-enabled devices.