UWB-Based Wireless Sensor Network

The wireless sensor network is an emerging pioneering and interdisciplinary research hotspot in the world. It integrates the sensor, embedded computing, wireless communication network, distributed information processing and micro electromechanical technologies. It is able to do real-time monitoring, perceiving and collection of information about environments or monitored objects by integrated micro sensors. It can also use the embedded system to process the information, and send the information to end users via multi-hop relay in the randomly self-organized wireless communication network. Accordingly, it realizes the concept of "ubiquitous computing". When Technology Review  of USA talked about ten emerging technologies in 2003, the wireless sensor network technology was on the top of the list. Defense Today  of USA even thinks application and development of the wireless sensor network would make a new epoch of military technology revolution. Therefore, it is predictable that the development and broad application of the wireless sensor network would greatly influence and impel people´s social life and industrial revolution.

    Nodes of the wireless sensor network feature low power consumption, low cost and small size. However, conventional sinusoidal carrier-based wireless transmission technologies cannot meet the requirements of the wireless sensor network because of the limitation of intermediate and Radio Frequency (RF) circuits and some inherent parts. Ultra Wide Band (UWB) communication technology is an innovative wireless transmission technology. It generally adopts an extremely narrow pulse (pulse width between the order of ns and that of ps) or an extremely wide frequency spectrum (relative bandwidth that is over 20% or absolute bandwidth over 500 MHz).
Compared with the conventional sinusoidal carrier-based communication system, the UWB system has the following advantages:

• High transmission data rate
• High efficiency of space-frequency spectrum
• High precision of distance measurement
• Low probability of interception
• Multipath interference immunity
• Frequency spectrum sharing with the existing systems
• Low power consumption
• Low cost
• Easy implementation of all digitalization
  

    These advantages help the UWB wireless transmission technology naturally combine with the wireless sensor network. Therefore, the research and development of the UWB-based wireless sensor network has been getting increasing attention.

1 Wireless Sensor Network
The wireless sensor network is a special Ad hoc network. It is applied in the areas with difficulty in wire laying and power supply, and the regions where working staff cannot stay (polluted area, environment-protected area or territory of hostile forces). Furthermore, it is applicable in the places where certain temporary affairs took place (such as natural disasters and destruction of fixed communication network). It saves from the support of fixed network and features rapid build-up and strong resistance to destruction. Therefore, it may be widely applied in such fields as military, industry, transportation and environment protection.

    Typically, a wireless sensor network uses micro air vehicles to scatter a large amount of sensor nodes in target areas, and these nodes will quickly form a wireless network by self-organization. The nodes are the collectors and transmitters of information, as well as the routers. Collected data reach the network gateway through multi-hop routing. The network gateway (also called the Sink node) is a special node. It may not only use the Internet, mobile communication network or satellites to communicate with the monitoring and control center, but also collect data by using unmanned air vehicles to fly over the network.

1.1 Architecture of Wireless Sensor Network
Figure 1 illustrates a typical wireless sensor network, which includes distributed sensor nodes (groups), Sinks, Internet (or satellites) and the task manager [1].

    The sensor network node consists of four basic units: the sensor unit (including the sensor and analog/digital conversion module), the processing unit (including the CPU, memory and embedded operation system), the wireless communication unit and the power unit. In addition, optional functional units include the automatic power supply system, positioning system, and more. It is shown in Figure 2.

    The architecture of the wireless sensor network is different from that of the conventional computer networks and communication networks. Figure 3 shows the architecture, which consists of layered network communication protocols and sensor network management modules. For the layered network communication protocols, there are five layers: the physical layer, the data link layer, the network layer, the transmission layer and the application layer. Network management includes the modules of energy management, topology management, Quality of Service (QoS) control, mobility management and network security.

1.2 R&D Status Quo and Application Prospect of  Wireless Sensor Network
The United States has initiated the research and development of wireless sensor networks since the beginning of 1990s. The study projects include the Wireless Integrated Network Sensors (WINS) research program at UCLA, funded by Defense Advanced Research Project Agency (DARPA) in 1993-1999, the Smart Dust project at UC Berkeley, funded by DARPA in 1999-2001, the DARPA-sponsored Sensor Information Technology (SensIT) program at UC Berkeley and other 24 research institutes in 1998-2002, and the SeaWeb program of US Navy in 1999-2004. So far, the United States has developed diversified wireless sensor network nodes, including Berkeley Motes, Sensoria WINS, Berkeley Piconodes, MIT uAMPs, Smart Mesh Dust Mote, Intel iMote and Intel Xscale Nodes. The nodes are designed for different applications. They are dissimilar in the hardware size, power consumption and design cost. However, most of them support TinyOS, an open-source operating system designed for wireless embedded sensor networks.

    In recent years, some Chinese research institutes have also started their study in the field of wireless sensor network, funded by National Natural Science Foundation of China and the Hi-tech Research and Development Program of China (generally called the "863" Program). These funded institutes include University of Science and Technology of China, Tsinghua University, Institute of Computing Technology of Chinese Academy of Sciences, Shanghai Institute of Microsystem and Information Technology of Chinese Academy of Sciences, Shenyang Institute of Automation of Chinese Academy of Sciences, and Institute of Intelligent Machines of Chinese Academy of Sciences.

    More research institutes home and abroad are joining the study on wireless sensor networks. This is totally owed to their broad applications and tremendous impacts on the social life. Table 1 lists the potential applications of the wireless sensor network.

2 Strengths of UWB-based Wireless Sensor Network
The conventional sinusoidal carrier-based communication network cannot meet the requirements of the sensor node, such as low cost, low power consumption, simple design and interference immunity. The UWB pulse radio technology has rapid development in recent years and attracts much attention from the industrial and academic worlds. It has numerous advantages over the sinusoidal carrier-based communication technology, and enables effective and reasonable transmission in the wireless sensor network.

2.1 Low Cost, Low Power Consumption and  Simple Design of Transceiver and Hardware Circuits
UWB is a new and unconventional technology for wireless transmission. The UWB system adopts extremely narrow pulses or extremely wide frequency spectrum to transmit information. The transceiver has no conventional circuits of intermediate frequency and RF in it. Therefore, the design of the system is simple and its cost and power consumption is much lower than that of the sinusoidal carrier-based communication system. The UWB based wireless sensor network may well solve the problems on size, cost and power consumption that exist in the conventional wireless sensor network, therefore, it well meets the demands on the micro sensor node. Moreover, its capability for high data transmission in a short distance facilitates the gateway nodes to transmit a mass of data and offer real-time multimedia services. For transmitting information of one bit, UWB consumes significantly lower power than the traditional wireless communication models with low power consumption (such as bluetooth, Zigbee and TR series)[2].

2.2 Large-capacity Spatial Transmission
The capacity of spatial transmission plays an important role in the wireless sensor network with high-density nodes. The UWB technology enables a significantly higher transmission capacity per unit area than other short-distance wireless technologies. Therefore, It is more application for transmission in the wireless sensor network with high-density nodes.

2.3 Strong Capability of Multipath Separation
Since most RF signals for conventional wireless communication are continuous or maintain longer than the multipath transmission time, the multipath transmission effect limits communication quality and the rate of data transmission. What UWB radio transmits are single-period pulses with extremely short maintaining time and extremely low duty cycle, therefore, multipath signals may be separated with regard to time. Without time overlap, the pulse multipath signals easily enable the separation of multipath components and the full usage of the energy of transmitting signals. Numerous experimental results show that the maximum fading of UWB radio signals is just 5 dB in the multipath environment where the multipath fading of conventional radio signals reaches 10-30 dB. The strong capability of multipath separation not only enables the UWB based wireless sensor network applicable in a bad and complex multipath environment, but also saves power consumption of data transmission of the sensor network.

2.4 Excellent Interference Immunity and High Security
With a low duty cycle of pulses and multi-pulse transmission of information of a bit, the UWB wireless communication technology supports high processing gain and improves interference immunity. Therefore, it is a good choice for information transmission in a bad electromagnetic environment. Moreover, the UWB signal generally disperses its energy in a broad frequency band. For a regular communication system, this signal is equal to the white noise. Its power spectral density is even lower than that of the natural environmental noise. Therefore, it has a low probability of being intercepted and detected. After the pseudo-random pulse is generated by coding, the detection will be more difficult. This strength of UWB enables excellent secrecy of the wireless sensor network for military. In addition, UWB has the characteristics of low power spectral density and high processing gain, which ensures the excellent capability of same frequency band sharing of the UWB based wireless sensor network. This well solves the problem of electromagnetic compatibility in a complex environment.

2.5 High Precision Distance Measurement and Positioning
The correct positioning of the sensor node plays an important role in application of the wireless sensor network. A direct way to gather node location information is to use Global Positioning System (GPS). However, the usage of GPS in the wireless sensor network is restricted by several factors such as cost, size and power consumption. Therefore, it is difficult to come true. In addition, it is difficult to apply GPS in indoor networks. UWB radio has a pulse width of the order of ns (and even that of ps), and occupies more than 1 GHz bandwidth. Therefore, it has a capability of centimeter-order relative positioning. Some foreign corporations are developing UWB based communication/positioning systems for the military. The IEEE 802.15.4a working group is making standards for the physical layer of the low-speed wireless personal area network. Its study focuses on the solution of communication and high precision distance measurement/positioning with extremely low power consumption. The UWB technology is just its main candidate solution.

    In a word, the UWB based wireless sensor network has a number of unique strengths. Therefore, it is becoming a research hotspot and  a development trend of the next-generation wireless sensor network.

3 Key Technologies for UWB Based Wireless Sensor Network
With regard to the study of the UWB based wireless sensor network, several hotspot factors, besides the general key technologies for the wireless sensor network, should get attention. These factors include the followings:

• UWB wireless transmission technology applicable in the wireless sensor network
• Media Access Control (MAC) protocols for the wireless sensor network that is combined with UWB
• Routing technology based on positioning information
• Distance measurement methods with high precision
• Schemes of cross-layer design

3.1 UWB Transmission Technology
When the application environments of the wireless sensor network is taken into consideration, low power consumption and low cost become the important requirements of the design of the wireless sensor network. Therefore, the development of appropriate UWB transmission technology should consider the requirements of the wireless sensor network. The development focuses on simple modulation/demodulation technology with low power consumption, small transceiver with low cost and communication/positioning integration with reasonable structure. Therefore, it can be concluded that noncoherent UWB wireless transmission technology is a good solution.

3.2 MAC Protocols
The power consumption management is a core part of the study of wireless sensor network. The RF module of a node consumes most of the energy, therefore, it becomes a main object of optimization. The MAC protocols directly control the RF module, and plays an important role in power consumption control of the node. The invalid consumption of a sensor node mainly comes from idle detection, data confliction, crosstalk (i.e. receiving and processing the data to other nodes), and message overhead control.

    The MAC protocols reduce power consumption mostly by decreasing the data flows, extending the dormancy time of the RF module and avoiding data confliction. Presently, the representative MAC protocols for the wireless sensor network are Sensor-MAC, Timeout-MAC, Wise-MAC, Berkley-MAC and Data-gathering MAC. Moreover, IEEE 802.15.4 that features low cost, low power consumption and low data rate has detailed specifications for the MAC protocol. The study of the MAC protocol for the UWB based wireless sensor network can start with the basic UWB channel division. Then, the study makes use of multiple access of UWB radio (such as hop-time multiple access), combines the available MAC protocols with the positioning function of the UWB technology, and finally obtains a distributed MAC protocol with low power consumption and balanced capabilities.

3.3 Routing Protocols
The task of routing protocols is to set up routing between the sensor node and the sink node for reliable data transmission. Since the resources of the wireless sensor network are limited, the design principle of the routing protocol asks for the simple algorithm. In addition, much storage of status information should be avoided at the node, and great routing information switching is not permitted between nodes. Presently, the representative routing protocols include Flooding/Gossiping Protocol, SPIN Protocol, Directed Diffusion Protocol, LEACH Protocol and TEEN Protocol. The routing protocol integrates the high-precision positioning capability of the UWB technology, and uses location information to transfer data to the target area. This avoids the data from broadcasting in the entire network for finding the target node. Information about the relative distance between nodes also helps choose more energy-saving routes for data packets.

3.4 Techniques for High-precision Distance Measurement/Positioning
According to positioning mechanism, available self-positioning algorithms for the wireless sensor network can be divided into the range-based and the range-free.

    The range-based positioning algorithms measure the point-to-point distance or angle between nodes, and then use the trilateration, triangulation or maximum likelihood methods to estimate the node location. The range-free positioning algorithms estimate the node location according to the information about such as network interconnection, consequently, information about the distance and angle is unnecessary. The latter has advantages in cost and power consumption, but its precision is relatively low. The range-free algorithms include the centroid algorithm, convex algorithm, DV-Hop, Amorphous, MDS-MAP and APIT. The UWB pulse has less width than 1 ns, occupies more than 1 GHz of bandwidth, and makes Time of Arrival (TOA) estimation to measure distance. Therefore, it can theoretically realize the distance measurement with centimeter-order precision[3]. However, with the influence from complex multipath and Near Line of Sight (NLOS), its precision of distance measurement and positioning hardly reaches the theoretical value in practice. Currently, an urgent problem is to choose positioning mechanism with high performance-cost ratio and energy-saving algorithm for positioning and tracing.

High-precision distance measurement, positioning and tracing of the wireless sensor network can be implemented by the design that combines MAC and routing protocols.

3.5 Optimization of Protocol Stack and  Cross-layer Design
The optimized protocol stack can directly support the optimization of network energy management on the premise that certain communication performance (such as the transmission rate, delay and loss tolerance) is guaranteed. The protocol stack optimization of the wireless sensor network requests overall cross-layer design that combines such important indexes as fault tolerance, interference immunity and power consumption for different application environments. The real application environment, characteristics of UWB radio and requirements of data collection, data convergence, object positioning/tracing, inquiry and management are the necessary factors for the study of architecture of the UWB wireless sensor network. The goal of the study is a UWB wireless sensor network with self-organizing, distributed and cross-layer optimizable architecture.

4 Conclusion
The UWB technology and the wireless sensor network can be integrated well. With great advantages over conventional sensor networks, the UWB based wireless sensor network has potential application, and is becoming the trend of next-generation wireless sensor network.

References
[1] Akyildiz I F. A Survey on Sensor Networks[J]. IEEE Communication Magazine, 2002, (8): 102-114.
[2] Oppermann I. UWB Wireless Sensor Networks: UWEN ─A Practical Example[J]. IEEE. Communication Magazine, 2004,42(12):27-32.
[3] Chung W C, Ha D S. An Accurate Ultra-Wideband (UWB) Ranging for Precision Asset Location [C]. IEEE Conference Ultrawideband System Technology (UWBST), Reston, 2003, 11: 389-393.

Manuscript received: 2005-06-16