Your browser is not supported. Please update it.

22 November 2006

MediaFLo re-iterates technical edge over DVB-H in the real world

Talking to Qualcomm about MediaFLO again this week, it continues to re-iterate claims that not only is MediaFLO a far better system than DVB-H, but it can offer double the number of TV stations over the same spectrum.

This all stems from what it now sees as a demonstrable 4dB of difference in radio power requirements which translates into a 40% saving on the cost of building a network, said Omar Javaid, Senior Director of Business Development, at MediaFLO.

‘We have not yet released the data on the trial at Cambridge in the UK which we did with BSkyB, but I can tell you that those results have affirmed our claims with respect to DVB-H, in fact the trial has outperformed our claims,’ he added.

The fundamental differences claimed when using MediaFLO is that it has a simplified air interface, not one based around sending internet packets, a two layer version of the H.264 Codec, each traveling on a different layer of modulation, which falls back to a less ambitious video scheme when radio reach falls away; and a less frequent use of the physical radio, which saves on battery life.

The two tier Codec is apparently already specified within the H.264 standard but has not been implemented by any other company said Javaid.

Qualcomm pulled its presentation of the results of the UK trial at the last moment, which was to coincide with a meeting with analysts, including a visit from the CEO Paul Jacobs. The analyst meeting went ahead, but the delay of the results, set back a month or so, left the MediaFLO team with less ammunition to convince the world that it is the right way to go.

So far Qualcomm has the biggest design win on the planet for mobile TV delivery in Verizon, which has shrouded its launch plans in secrecy since agreeing to an exclusive period, adding the relevant extra radios and demodulators to its phones, and is largely expected to deliver MediaFLO in markets where it rolls out EV-DO in the US, in the 1st quarter of 2007.

‘We are also committed to delivering MediaFLO as a single frequency network. DVB-H offers that as a preferred delivery method, but the 3 Italia service in Italy is based on a variety of frequencies, simply because it couldn’t get contiguous frequency allocations,’ said Javaid.

‘This means that some transmitters they use are simply not powerful enough for in-building penetration. There are different power limitations for transmitters in different territories across Europe, there is no international agreement on the power allowed for transmission in UHF. Which is why we are hearing about problems with the Italian service achieving in-building penetration,’ he added.

‘Some countries in Europe have a 10 kW limit, others 20 kW, and in some small areas in Italy we know 3 Italia is using gap fillers of 2 to 5 kW because they are co-broadcasting.’

Javaid also points out that most DVB-H trials around the world are being installed with transmitters much closer together than they would be in a real life situation. ‘They obviously want to see what the effect of a trial is without penetration being an issue,’ he said, ‘but you cannot assume that because they are putting transmitters 2 miles apart in the trials that DVB-H will be deployed that way.’

‘In MediaFLO we are achieving between a 20 and a 25 kilometer transmission radius and achieving in-building coverage, using 50 kW effective radiated power.’

If Javaid is right about the 3 Italia in-building penetration being an issue, it will be up against it in the coming months. The world’s mobile TV eyes are pointed directly at Italy to see how DVB-H performs in a real world commercial environment, and 3 Italia is hoping to sign up half a million subscribers by Christmas and already claims to have half that number. Rival Telecom Italia Mobile has launched with roughly the same coverage, and around 8 TV channels compared to 13, but it is charging just €5 a month, compared with 3 Italia’s €29 a month. Vodafone which will use the same Mediaset provided services as TIM has yet to show its arm commercially.

The delayed results of the BSkyB trial, when they emerge will also have to back up Javaid’s other claims about penetration and cost. He claims that in a 8 MHz UHF channel MediaFLO can deliver 40 TV channels, plus audio plus clip cast services. In the US the claim was originally less aggressive stating that a US 6MHz TV channel was allowing 20 TV channels in full QVGA (320 x 240 pixels) at 30 frames per second. The basic layer of the codec delivers 15 frames per second and Javaid claims that the secondary codec data reaches 90% of the cell, bringing this up to 30 frames a second. Just how a TV signal is sent with some I-Frames in one service and some in another, we have yet to find out, and Qualcomm has more convincing to do in the real world, the same as DVB-H has.

But for now DVB-H remains the dominant mobile TV standard globally with more trials already live and more planned this week in the middle east (See separate story), fast approaching 40 trials against MediaFLO’s five.

In the meantime the UK’s only mobile TV service Movio, run by BT and sold only by Virgin Mobile, sent us a phone to view the service in an attempt to convince us that it is a viable TV service. Well in our view it remains non-viable. It is jerky, there is no lip synch and the sound is bitty. It delivers over a 96 kbps DAB service and recently companies like Orange have raised the bar on the cellular unicast TV services, taking them up from 70 kbps to around 120 kbps using HSDPA for improved delivery. BT is going to have quite a problem selling Movio to Orange, Vodafone and O2 if their 3G network is already offering better TV resolution.