10 May 2011
Wireless communication protocol promises high speed data transfer between portable devices
The advent of devices such as the iPod, tablets and smartphones has changed the way we deal with data. Rather than storing it in particular locations – the cd rack, for example – people are now more likely to carry all their data around with them.
But that's not a particular problem; memory is cheap enough that the latest consumer electronics devices have more than enough storage capacity.
The problem today is moving data to and from mobile devices; not only downloading new music or movies to your portable device, but also transferring data from one device to another. If you're taking a lot of photographs or videos, you want to upload them directly to your pc or maybe to YouTube. And you might want to transfer images to another portable device. Beyond that is the need to develop secure and workable payment systems.
A number of methods are available today, including Bluetooth, Near Field Communication and Wi-Fi. But what is required is a method which offers high performance, low cost, high security and, importantly, ease of use. The solution, at least in the view of a consortium of consumer electronic product developers, is TransferJet (www.transferjet.org).
TransferJet is a wireless technology that is said to combine the speed of Ultra Wide Band with the ease of Near Field Communications. And the members of the TransferJet Consortium claim the approach allows data to be transferred at rates of up to 560Mbit/s, simply by 'touching' another device. Established by Sony, the TransferJet Consortium now boasts more than 50 member companies
TransferJet is thought to be useful in many applications. With a peak data transfer rate of 560Mbit/s and an effective throughput up to 375Mbit/s, communication is always point to point between two active devices, simplifying the system.
Because the technique is intended to operate over very short range – a few centimetres – it is possible to operate in the radio signal's near field and to therefore use very little power to transmit data. In fact, the Consortium suggests a figure of -70dBm/MHz.
The point to point topology also simplifies network setup and management procedures. And, since the near field is non polarised, the communicating devices do not have to be lined up exactly in order to make a good connection.
Version 1.0 of TransferJet uses a centre frequency of 4.48GHz and a bandwidth of 560MHz. This, coupled with the low transmit power, enables unlicensed operation in many countries. In addition, the TransferJet protocol includes error detection and correction, packet acknowledgement and packet resend features, all of which minimise complexity and interference.
TransferJet enabled devices can also detect the presence of a similar device as it comes within range, so power is saved by transmitting data only when another device is detected.
While communicating over a range of no more than a few centimetres may seem to be a disadvantage, the TransferJet Consortium claims the opposite. One particular benefit is the elimination of multipath fading, or interference, experienced with Bluetooth and Wi-Fi. Because of this, designs do not need to include equalisers or signal processing systems.
Short range communication also means the protocol can assume that, because two devices have come into range of each other, their users have authorised the connection.
The Consortium is equating this with plugging in a USB cable and says TransferJet provides the ease of a USB cable without the cable.
Nevertheless, devices must proceed through search, discovery, selection, authentication, connection and transfer in order to transfer data. But with TransferJet, these steps are collapsed into one process – the 'touch'.
Despite the short range over which TransferJet operates, designing the antenna is seen to be a challenge. 'If two TransferJet devices are separated by more than a few centimetres, they should do nothing – not even detect each other. That is a very difficult task for a conventional antenna. A typical antenna is designed to radiate a signal as far as possible', the Consortium says.
According to the field equations for an ideal dipole excited by a sinusoidal current, the field strength of the far field parts of the signal varies inversely with range. However, the near field parts vary inversely with the square of range. So the near field intensity drops off more quickly with range than does the far field. The near field also stores power, which is dissipated only if another TransferJet enabled device appears in the near field.
Enabling such communication requires a different kind of antenna. Instead of a conventional antenna, the TransferJet Coupler is designed to suppress the far field component and to emphasise the near field signal.
The approach creates a virtual 'bubble' of signal energy that drops off very quickly beyond a few centimetres. Once established, the link will not break unless the devices are separated beyond the bubble distance – described by the Consortium as a 'soft' engage feature.
While the protocol has been developed and members of the Consortium are working on the TransferJet Coupler, there is still the need for silicon. And Consortium member Toshiba Electronics is one of those looking to develop suitable devices.
Armin Derpmanns, general manager of Toshiba's asic and SoC group, explained the attraction of the technology. "Toshiba makes a lot of storage devices, but users have a lot of trouble transferring data from one device to another – they have to use wires or USB or something similar. The advantage of TransferJet is that it provides high speed intuitive communications with low power consumption."
The advantage of TransferJet over competing technologies is illustrated by comparing the download time for a 4.7Gbyte movie. According to Derpmanns, downloading that file using Bluetooth 2.0+EDR would take 209minutes. Downloading the same file using IEEE802.11g takes 12min. TransferJet, however, can transfer the data in just 100s.
TransferJet is also being seen as a potential enabler for payment applications, working in concert with NFC. Using a smartphone as an example, Derpmanns suggested that data – a video, for example – could be downloaded from a terminal at high speed using TransferJet, while payment would be made using NFC.
Toshiba's TransferJet solution, currently under development, integrates an rf section and a digital baseband on one chip manufactured using the company's 65nm rfcmos technology. The chip, which features an rf switch, a matching circuit and a low noise amplifier, also supports UHS-I communications within the SD definition. This chip is likely to consume 180mW when transmitting data at 375Mbit/s.
Derpmanns expects this device to launch sometime in 2012. This will be followed in early 2013 by a similar part with a USB2.0 interface. Derpmanns also envisages the chip and coupler being packaged in an SD card to widen the potential range of applications.
However, a planned development to TransferJet will see data transfer rates increase substantially. In V2 of the protocol, the will move to a centre frequency of 60GHz, increasing the data transfer rate to 2Gbit/s. Expect this chip in mid 2014.