TECHNOLOGY
ASK THE EXPERTS

Have a Question for the UMA Experts? Send it to experts@kineto.com

Mark Powell
Vice President, General Manager - Client Business Unit
Co-Founder

Mark Powell has served as vice president and general manager of Kineto Wireless' client software business unit since co-founding the company in 2001. Mr. Powell is responsible for the company's global business and alliance strategy with handset and technology partners. Mr. Powell is a wireless industry veteran with more than 20 years of experience in business development, marketing, manufacturing and engineering roles. Prior to joining Kineto Wireless, Mr. Powell worked for both Motorola and Nokia Mobile Phones; he was responsible for cellular handset portfolio planning. Mr. Powell holds an EE degree from the Highbury College of Technology in England. view full bio
Battery and Performance for Dual-Mode Phones
There is a lot of talk about battery life in dual-mode phones. What are the factors driving performance? Much has been said about power consumption in dual-mode GSM/Wi-Fi phones. While it is certainly true the initial Wi-Fi protocol was not optimized for voice, it has proven flexible enough to rise to the challenge. Work by the Wi-Fi Alliance to develop the Wi-Fi Multi-Media (WMM) specifications has resulted in some important performance improvements in battery life for dual-mode phones. Kineto certainly recommends support for the WMM profile on devices and Wi-Fi access points.

One primary power draw in dual-mode devices is having multiple radios on simultaneously. It makes sense that any time the GSM and Wi-Fi radios are powered on, there is a higher current draw than if just one radio were running. Minimizing the time the radios are on simultaneously is a key factor in managing power performance for dual-mode handsets.

Additionally, there are two system-level capabilities that have a dramatic effect on battery performance for dual-mode phones. First is how service is delivered when the phone is on Wi-Fi. Second, when on Wi-Fi, how are the VoIP packets processed? The difference in a UMA system has a big effect on the performance of dual-mode phones.
So all dual-mode phones and dual-mode services are not alike?
No, UMA-enabled dual mode phones have been tuned and optimized to support voice services over Wi-Fi, and dramatically outperform devices without the UMA protocol.
How are mobile services delivered to UMA-enabled devices?
In UMA, when the user/device walks into range of the home Wi-Fi access point, the phone automatically establishes a direct link to the mobile core network over the IP network. Once the UMA connection is established, the mobile network routes all voice calls (and SMSs, MMSs,. . . ) to the phone over the Wi-Fi/IP network. The GSM radio connection to the phone is no longer used. Therefore to save power, the GSM radio is put in a sleep mode. In a UMA solution, both radios are only on for a very short period of time during the handover procedure. Users also get a seamless experience in UMA. The Wi-Fi / GSM handover procedure is seamless and the subscriber notices no difference in the services delivered.

Alternatively, non-UMA devices provide something less than a seamless user experience. Because UMA is the only standard approach for mobile/Wi-Fi convergence, any non-UMA approach is ad-hoc, often not linked with the mobile service provider, and consequently does not offer a seamless experience.

Many of the dual-mode phones on the market today suggest that subscribers leave the GSM radio on concurrently with the Wi-Fi radio to get the semblance of a seamless user experience. In this case subscribers can access the Wi-Fi radio for IP-based services while maintaining the GSM connection for voice, SMS, MMS services. The problem is that with both Wi-Fi and GSM radios powered on simultaneously, users receive dismal power performance.

The alternative option, deployed frequently in Asia, is to have the user power off the GSM radio when on Wi-Fi. While this addresses the battery/power issues, it does not offer a compelling user experience. With the GSM radio off, calls to the mobile subscriber are immediately routed to voice mail. Thus, the user ends up with two services on the phone, a GSM service and a Wi-Fi service, which do not operate concurrently.
This approach is often used in enterprise solutions. When the subscriber is in the enterprise, the GSM radio is powered off and the subscriber's PBX direct dial extension becomes the primary phone number for the device.
How is voice over Wi-Fi handled in UMA versus non-UMA systems?
UMA defines the use of the existing GSM AMR codec as the voice coder when in Wi-Fi. The AMR codec is hardware-assisted on the baseband of the device. Because it's used as the primary codec in GSM mode, this implementation has been fully optimized for voice coding with the lower possible power usage.

In non-UMA deployments, often a "VoIP client" is downloaded onto the phone. This client runs its own codec and the compression/calculations are performed in application space on the handset. This is incredibly inefficient and uses a tremendous amount of power on the device.