Data transfer speeds are important to most, although the cost in terms of power consumed must also be considered. Communications systems designers normally choose a single technology to transfer and control the flow of data. For example, ultrawideband (UWB) technology offers high data rates, but has some trade-offs. For example, the time needed for connection to another UWB device can be considerably longer than other wireless communications formats, such as Bluetooth. UWB systems use very low power-per-bit in transmission mode, but if a UWB communications device has to search to find its communications partner, that savings is lost. A hybrid system, using Bluetooth technology as the device control manager, makes UWB more efficient and enables its use in a broader range of consumer electronic devices.
Information transfer, in the form of documents, music, news, and other data, is vital to the modern workplace and lifestyle. Time is often critical in making a transfer, but the power expended to make that transfer is also important. To understand the advantages of a combination of Bluetooth and UWB technologies, it is useful to go back to a basic principle: why use wireless communications at all? Wireless, as a technology, is usually two orders of magnitude more difficult than using wires. Wireless technology is better suited for devices that move around or are too far apart to make wires practical.
Five classes of wireless-communications devices offer decreasing areas of coverage:
- Wireless Global Area Networks (WGANs)
- Wireless Regional Area Networks (WRANs)
- Wireless Metropolitan Area Networks (WMANs)
- Wireless Local Area Networks (WLANs)
- Wireless Personal Area Networks (WPANs)
Figure 1 presents a simple graphical representation of these wireless-communications classes. In approximate terms, WGAN and WRAN are about two orders of magnitude greater distance, with the WMAN and WLAN about one order of magnitude—arriving at about a 10 meter range for WPANs. WGAN technology uses satellite communications and has the potential for covering vast areas, for example in the manner of satellite-television and Global Positioning System (GPS) systems. WRANs are terrestrially based systems and include traditional broadcast radio and television, plus the communications technology being defined by IEEE Project 802.22. WMANs are the domain of mobile telephony, but also have an IEEE activity, Project 802.16. Although other, less successful, examples of WLANs exist, IEEE 802.11a/b/g/h is perhaps the best-known example.
The IEEE also defines WPANs with various subtypes of Project 802.15, including a version of the Bluetooth technology (802.15.1), a multi-media-oriented type (802.15.3), and a low data rate (802.15.4). Two UWB technologies fought to a stalemate to become the physical layer (PHY) for 802.15.3: a direct-sequence spread-spectrum (DSSS) version promoted by the UWB Forum (www.uwbforum.org), and a multiband OFDM version promoted by WiMedia (www.wimedia.org)1. Many of these technologies share an overall structure, which, for example, makes it possible for synergy between Bluetooth and UWB technologies.
With more that a billion units in the field at the end of 2006, Bluetooth technology is far and away the most successful WPAN technology. The Bluetooth Special Interest Group selected WiMedia’s version of UWB to boost performance. This boost will enable Highly Mobile Devices (HMDs) to benefit from the UWB’s better than 400 Mb/s data rate, which in turn will enable new classes of Bluetooth applications.
To encourage mobility, an effective HMD must be portable and easy to carry or transport. This also implies low power consumption to support a small battery. The critical issue for an HMD that utilizes a WPAN is minimizing power consumption. Mobility is both an advantage and a potential problem in the WPAN environment. Data transfers without wires is convenient but, with an effective range of 10 m, a wireless connection can be easily lost with just a little movement. Would it make more sense to increase the range of an HMD device?
Farther and faster are not always better for wireless applications. For some applications, the reach of the system is critical. For example, early UWB radios were used to communicate with submerged submarines. And in some applications, such as point-to-point microwave radio systems, high speed is critical without critical dependence on power consumption. But for most applications, trade-offs must be considered.
Digital-communication throughputs are traditionally measured in the number of bits transmitted per second. With wireless communications, especially consumer-oriented versions, there is literally another dimension to be considered. Data transmissions through a wire occur on a point-to-point basis and essentially represent a one-dimensional (1D) operation, a vector. Without the wire medium, air serves as the transport medium in a wireless system in three-dimensional (3D) space. Figure 2 is a representation of this 3D attribute. The black line represents the wired data path and the red and blue spheres represent nodes. The green portion is the volume where wireless radios can communicate.
When the population of wireless-communication devices is sparse, increasing the reach of the devices only has an effect on power consumption. In a consumer device, a larger problem occurs when there are many devices in range of each other. Unlike most wired systems, the presence of collocated communication paths greatly impacts data throughput. The larger the spheres of communications, the more coexistence problems exist, which result in retransmissions, causing an even greater impact on the battery life for a mobile device.
Coexistence is defined in the IEEE 802.19 Technical Activity Group as “the ability of one system to perform a task in a given shared environment where other systems have an ability to perform their tasks and may or may not be using the same set of rules.”
If minimizing power is a key to the success of an HMD, then using the technology with the lowest instantaneous power budget would make the most sense. Unfortunately, a system’s power requirements change under various conditions. Bluetooth technology’s approximate average of 11 mA of consumption far surpasses the performance of the 500 mA UWB radio. However, there are many more factors involved in the total power cost of moving data. The table approximates the most important of those factors. (It should be noted that the actual power consumption of the UWB radio is not currently available without non-disclosure agreements in place. The power-consumption figure cited above is approximate and based on first-generation UWB systems. The power consumption cited for Bluetooth is from a commercially available single-chip radio.)
Consider the impact on a HMD of transferring a 2-Gb file from a server. The amount of power (P) consumed (in milliampere hours) to transfer a number of megabits (B) would be:
P = cPc)/3600> + (B/D)(Pr/3600)
Tc = the control time consumed (in seconds),
Pc = the control power consumed in mAh,
Pr = the data-transfer power consumed (in mAh),
B = the amount of data transferred (Mb),
D = the data rate (Mb/s), and
P = the total amount of power consumed (mAh).
The first grouping of terms in the equation represents the setup mAh and the second one describes the power cost of transferring the data in terms of mAh.
For a Bluetooth data transfer, the power consumption is:
P = + (11/3600) = 16.3 mAh
For a UWB solution, the power consumption is:
P = + (500/3600) = 6.11 mAh
An UWB radio consumes less power because of its 400 Mb/s data rate, even with a connection time of 4 s. But this advantage in speed and power is lost the longer radio remains powered on. HMD users do not want to have to shut their radios on and off to save power. Most Bluetooth consumers expect that the radios are always on and available. It dramatically changes the results of the analysis if the equation reflects the power required for the radio to remain ready to transfer, but not yet active for 12 h.
But when data rate is considered in the calculations along with the time to make a connection, the power consumption picture changes:
P = (Tc 3 Pc) + r> + (tw 3 Pw)
tw = the time waiting and
Pw= the power level to wait to receive data in low-power mode.
For Bluetooth, the power consumption is:
P = /3600 + + 12(0.18) = 18.5 mAh
For a UWB solution, the power consumption is:
P = /3600 + + 12(20) = 246 mAh
When the time spent waiting for a connection stretches into a 12-h day, the advantage of a hybrid radio system is clear. If the connection and wait periods are performed with Bluetooth technology and UWB is used for the actual data transfer, the power savings is significant.
For a Bluetooth setup and subsequent waiting period, and UWB performing the data transfer, the solution becomes:
P = /3600 + + 12(0.18) = 7.72 mAh
The result is a data transfer that is two orders of magnitude faster than Bluetooth alone and consumes about 3 percent of the power of UWB alone. The combination of technologies actually consumes less power than the Bluetooth-only file transfer while remaining on and active for 12 h.
Bluetooth is a preeminent example for a WPAN technology. Although its initial use was for audio applications, such as mobile phone headsets, the communications protocol was designed for a much broader application space and currently supports over 30 different application profiles.
Figure 3 shows an architectural view of the Bluetooth stack. Unlike the IEEE 802 stack, the Bluetooth stack includes many features that are included in layer 6 of the ISO OSI structure (Fig. 4). However, the Bluetooth stack does not closely follow the traditional ISO layering, although there is correspondence at the lower layers. An ISO PHY would encompass the Bluetooth PHY plus a bit of the Baseband portion of the protocol. An ISO Logical Link Control (LLC) would take in the rest of the Baseband, the Link Manager (LM), and bits of the Host Controller Interface (HCI) control.
Bluetooth technology was designed to be a low-power, low-complexity means of wirelessly connecting devices. The technology allows devices to be able to create an ad hoc network, which allows highly mobile users to share information readily.
Current applications run comfortably at present Bluetooth data rates from 1 to 3 Mb/s with very low power consumption rates. To satisfy the expanding capabilities of HMDs, with new features like automatic synchronization of large amounts of information, the data rates must be much faster. Blending any two technologies together requires care. Welding two technologies that essentially do similar functions and were developed for different types of applications is even more of a challenge.
The right side of Fig. 4 shows the WiMedia architecture that is being deployed. The grey areas denote the WiMedia purview. The structure of the WiMedia stack reflects its IEEE 802.15.3 origins (although now standardized as Ecma-369) and has a traditional ISO OSI structure (shown on the left-hand side). The small block labeled SAP represents Service Access Point definitions, an abstract interface description required by the IEEE 802 project to ensure conformity of MAC/PHY combinations to the IEEE 802.2 Logical Link Control (LLC) definition.
There are many ways a weld of the two technologies can happen. The leading candidate detaches the WiMedia stack from its WiNet or certified Wireless USB upper stack and attaches it to the L2CAP (Logical Link Control Adaptation Protocol) / Device Manager block. It becomes a peer of the current 2.4 GHz radio/baseband as shown in Fig. 5.
Bluetooth is the most successful WPAN technology in history; the combination of Bluetooth and UWB will have a ready market, making possible new applications such as streaming personal video, TV broadcasting and wireless portable mass storage. The first consumer devices will have limited requirements for the UWB bands and likely be hosted in advanced high-end mobile phones. Along with these, there will be optional Bluetooth UWB add-ons for PCs. Very early non-mobile-oriented UWB-only products may be seen in early 2007. Subsequent devices, integrated with Bluetooth technology, will supply the ease of use, power saving, and interoperability expected for consumer devices.
1. By mid-2006, the WiMedia version dominated the UWB space and has been standardized by Eema.