Speed of light is too slow for mobile comms, says RIM CTO

The problem with the speed of light is that it is just too slow, says David Yach, CTO of Research in Motion's software.

The problem with the speed of light is that it is just too slow, says David Yach, CTO of Research in Motion's software.

In perfect conditions, it takes a signal around 100 milliseconds to reach the other side of the Earth, and another 100 milliseconds to get back to you. "Latency will never be overcome," he says. "You can send huge amounts of data, but you cannot necessarily do it back and forth, especially anything that requires tight time coordination."

That latency is a fundamental constraint now limiting the use of IT in many applications. Clever techniques such as buffering help to fool the hardware, but applications that make use of video streaming and real-time control do not tolerate delays.

The Blackberry maker, in fact the mobile communications industry, faces other tough hurdles. These are the absolute limit on how many bits of data you can shovel into a given chunk of radio spectrum, and how many radios and antennas you can shoe-horn into a mobile phone. Then there is the battery power required to bring all that kit to life.

High-end phones already have 11 or 12 antennas, which is a challenge for space and battery engineers. "We have got the space gremlins under control with 3G phones, but LTE [Long Term Evolution, or 4G mobile] will renew the challenge," says Yach.

Wireless limitations


There are others. "Like all radio technologists, we have had groups working on LTE technologies for a number of years now," says Yach. "The first networks are starting to roll out - Verizon in the US and a couple of others elsewhere. We are also monitoring closely all the unanswered questions, such as 'how are the carriers planning to do voice?', and 'are they all going to agree on doing it the same way?'."

There is also disharmony about what frequencies to use. Common sense says that every country should use the same ones to keep equipment costs low. "Right now there appear to be 21 different LTE bands around the world, because countries are allocating spectrum from whatever spectrum has not yet been allocated," he says.

Yach likes LTE. There are aspects of LTE technology that he sees as superior to 3G, such as potential battery life and latency. But he says, "LTE is the equivalent of taking a 1in water pipe from the street into your home, and putting in a 2in pipe. You get a lot more pressure. But if all your neighbours also move to 2in pipes and the water mains don't change, there is no more capacity to go around."

Wireless will never have the broadband capacity of fibre or even copper. The physics do not allow for that. But it will drive people's perception and need for data and for new behaviours. That will impact the mobile space, Yach says.

Functionality trade-off


But users' demand for capacity is already outpacing suppliers' ability to deliver more. For RIM the biggest issue is the tension between the industry move towards smaller, thinner, sleeker devices and the need to stuff more things into the case.

"We have always believed that efficiency matters, partly because of economics and partly because we do the math and physics. In web browsing, for instance, you can put three Blackberry users on every one of almost every other competitive product. E-mails are actually a better ratio," Yach says.

This is because RIM does "crafty things" in terms of compressing data, and not resending data unnecessarily. That helps improve battery life because it is not using the radios as much. It also pushes data to the phone, using quiet moments on the network, making the information available almost instantly.

RIM also pays a lot of attention to what users see on screen. "It is my personal hypothesis that people do not like touch-screens as much as they like big screens," says Yach. "Our belief is that people like different things. Some like touch-screens, others don't. Others like a small form factor, other like a full qwerty keyboard.

"Our measure for screen density is when young people cannot see the pixels any more. Once you get beyond that you are trading off performance and battery for nothing. What matters more for LCD screens is the size of the frame around each pixel. Same thing with cameras - what matters more is the quality of the lens, not the pixel count. Then there is colour saturation and other things you can do in processing that affect the quality of the picture," he says.

Weighing up demand


In terms of video, everyone wants to get to high definition. "That is a two-megapixel camera, 1920 by 1080," says Yach. "But being able to record it at that rate and transmit it is a challenge. We are close to the point where high-end phones can record HD video, and when that becomes standard, frame rates, colour and lights levels will be more important than pixels."

Yach believes unified communications, including video calls, will, even are, becoming reality. But he is not sure people will use it. "Are you going to use it on a train or a bus and hold it in front of you and speak out loud? It is nice to have when you are travelling and you can see the people back home, but I don't think people want to be seen all the time," he says.

Economics determine its popularity, especially on 3G networks. "Video calls will cost more than audio calls, so people will make an economic decision about the value of that visual image," Yach says.

Skype Video and YouTube are burning up a lot of mobile carriers' bits, which is why they often ban or throttle the traffic. "Initially the carriers had lots of capacity, so they were pushing users to use it all, with visions of sugarplums about how much money they were going to make. It did not work out like that. I think they will have to offer videoconferencing for the same price as audio conferencing. Will it be used? Yes. Will it be ubiquitous? I don't think so," he says.

This raises the net neutrality issue. "There is a human debate to be had," Yach says. "Who has the right to decide whose traffic is more important, or the right to grab all the available bandwidth? There are some technical issues, but fundamentally it is a moral argument. As far as RIM is concerned, we are pragmatists," he says.

Read more on Wireless networking