An Investigation of UWB-Based Wireless Networks in Industrial Automation

Transcription

1 IJCSNS International Journal o Coputer Science and Network Security, VOL.8 No.2, February An Investigation o UWB-Based Wireless Networks in Industrial Autoation Abdellah Chehri, Paul Fortier, Pierre-Martin Tardi LRCS Laboratory, 45, 3 rd Ave, Val-d Or, Québec, J9P S2, Canada. Departent o Electrical and Coputer Engineering, Université Laval, Sainte-Foy, Québec, GK 7P4, Canada. Suary Ultra-wideband (UWB) counication has been the subject o extensive research due to its unique capabilities and potential applications. Soe see UWB as an enabling technology or new wireless applications that span ro high-data-rate transission o raw ultiedia video to new location-aware, low-data-rate, and low-power counication o sensor and actuator data. Underground ines, which are characterized by their tough working conditions and hazardous environents, require oolproo ine-wide counication systes or sooth unctioning o ine workings and ensuring better saety. During the last years, a nuber o Canadian ining copanies have started to deploy odern counications systes in order to autoate processes, reduce operational costs and with the objective o increasing saety and productivity. However, any iportant aspects o UWB counication systes have not yet been thoroughly investigated in this kind o environents. In this paper we give an overview on UWB counication. Also we investigate via easureents and siulations the applications and design challenges or UWB counication and ranging in underground ines. Key words: UWB, Tie Hopping, Rake receiver, IEEE a, IEEE a, Underground ines, UWB Ranging.. Introduction Mining industry is one o the ost diicult industries, concerning production and technological processes. The surroundings in which we are providing these works in ines is inluenced by actors like the possibility o gas explosion, harul atosphere, high teperature, huidity, dust, noise, vibration, etc. The harsh physical environent and distinct topology that ake ining dangerous act as a hindrance or constraint to the very techniques and technologies that could iprove saety and productivity. Up until the late 98 s, instant, reliable radio counication was alost non existent in underground ines. The diiculty and high cost and lack o reliable technologies being widely available ade counications underground diicult. With industry deanding better counications underground as a ajor part o draatic saety and productivity iproveents in underground ining. In the interests o ine saety and productivity, it is vital that operators are continuously aware o underground conditions and risk proiles. They ust be able to locate and counicate with ine workers at all ties - particularly in the event o ires, roo alls or other lie-threatening situations. Soe copanies start to develop new technology that would provide robust wireless notiication and coordination under noral operations and during disasters. The world o ultra-wideband has changed draatically in very recent history. In the past two decade, UWB was used or radar, sensing, ilitary counications and any other applications. A substantial change occurred in February 22, when the FCC issued a ruling that UWB could be used or data counications as well as or radar and saety applications []. Considering the iportance o this technology, uture deployent o UWB wireless networks in underground ine will be an interesting application. UWB systes have potentially low coplexity and low cost, have a very good tie doain resolution, which acilitates location and tracking applications, extreely siple design (and thus cost) o radio, large processing gain in the presence o intererence, extreely low power spectral density or covert operations. UWB can provide an excellent cobination o high perorance and low coplexity or wireless counication and networking. This paper deals with the potential applications and the challenges o using UWB based-wireless networks as uture solution or autoation in underground ines.

2 8 IJCSNS International Journal o Coputer Science and Network Security, VOL.8 No.2, February Application o UWB Wireless Counication in underground ines Traditionally, various hard wired and wireless ine counication systes are available on the arket. These systes dier in the use o their physical counication inrastructure as well as in the counication protocols used. Data counication is perored using e.g. phone lines, obile radio systes, leaky eeder cables and iber optic backbones together with wireless counication basing on ine radio and obile phone syste technology. Dierent philosophies exist regarding the use o analog or digital counication and big dierences are visible in the available bandwidth. In ining, inrastructure cost is essential. Thus, a counication syste should be integrated as a ulti purpose syste capable o transerring all types o inoration as data, voice and video on an identical inrastructure. An obvious advantage o wireless transission is a signiicant reduction and sipliication in wiring and harness. It has been estiated that typical wiring cost in industrial installations is US $ 3-65 per eter and adopting wireless technology would eliinate 2-8 % o this cost [2]. Another advantage o wireless terinal is their obility (these terinals can be placed in transporting vehicles, rotating equipent,...). In act, the otivation or using wireless counication in industrial ining autoation is generally twoold: econoy and saety. Wireless counication technologies in ining will have a signiicant ipact on ine operations in the coing decades, giving ine anagers and sta uch greater understanding and control over ining processes in under to onitor and optiize ining operations. Increases in wireless counication capabilities also are establishing the technology base necessary to support reote and autonoous ining operations. In a ine gallery there is a requireent or any types o counications. Aong the voice counication aong ine workers is very critical. Video surveillance through inrequent snapshots in ine gallery is another application o interest and is used or data analyses. On top o applications or iproved public saety through the use o vehicular radar systes or collision avoidance, reote control applications are also o interest to the ine operators so that achinery can operate in extree conditions. Wireless sensor onitoring is another application that is very crucial or the saety o ine workers [3]. In the proposed applications or IEEE , several copanies ention sensor networks, which stand to derive huge beneits ro the low power and location aware properties o UWB. Since UWB has excellent spatial resolution it can be advantageously applied in the ield o localization and tracking [4]. There are a nuber o applications that would take advantage ro precise positioning and navigation such as autoatic storage and tracking o various targets [5]. All these types o counications can use UWB technology. An exaple or a odern underground UWB counication syste integrating both counication and autoation deands are given in Fig.. An exaple o sipliied diagra representing global UWB wireless counication in underground ines is shown in Figure 2. Three ajor coponents are shown: an application layer syste, an integration layer and a counication layer. The application layer includes, or exaple, the graphical user interace that is displayed on the end user s coputer. The display will provide an overall picture o the current risk proile o the ine and show a speciic sensor data or security onitoring, agents (drit, vehicle, ines) location and process data (production anageent, working place ) when requested. All the data displayed can use UWBcounication (PHY layer). Fig. : Integrated underground counication syste. Fig. 2: Sipliied diagra o the onitoring application or ining industry.

3 IJCSNS International Journal o Coputer Science and Network Security, VOL.8 No.2, February UWB Counication Advantage o UWB Copared to narrowband systes, UWB has several advantages. Because o the cobination o wide bandwidth and low power, UWB signals have a low probability o detection and intercept. Additionally, the wide bandwidth gives UWB excellent iunity to intererence ro narrowband systes and ro ultipath eects. Another signiicant advantage o UWB is its high data rate. Also, the carrierless nature o UWB gives it potential or siple circuit ipleentation without interediate oscillators and ixers. UWB devices ay have a nearly all digital ipleentation in CMOS without inial analog RF electronic []. This siple architecture can translate to low power dissipation and low cost, which opens a variety o possible obile applications. In general, UWB technology has any beneits due to its Ultrawideband nature, which include the ollowing: Coexistence with current narrowband and wideband radio services, Large channel capacity: it can support real-tie high-deinition video streaing, Ability to work with low SNR: oers high perorance in noisy environents. High perorance in ultipath channels: delivers higher signal strengths in adverse conditions. Siple transceiver architecture: witch can enables ultra-low power, saller or actor, and better ean tie between ailures, all at a reduced cost. Challenges UWB technology or wireless networks is not all about advantages. In act, there are any challenges involved in using nanosecond-duration pulses or counications. Soe o the ain diiculties o UWB counications are suarized in ollowing Table []. Challenge Pulse-shape distortion Channel estiation High-requency synchronization Low transission power Proble Low perorance using classical atched ilter receivers. Diiculty predicting the teplate signals. Very ast ADCs required. Inoration can travel only short distances. Table : Soe challenges and probles associated with UWB systes. 4. TH-IR-UWB Counication A tie hopping odulation orat eploying ipulse signal technology has several eatures which ay ake it attractive or ultiple-access counications. A typical hopping orat with pulse-position data odulation (PPM) is given by: s ( t) = E w ( t T c T δ b ) () tx = ( ) s i ( x c / N s where tx t) is rando process describing the transitted signal. w(t) is the basic pulse onocycle that noinally begins at tie zero on the transitter s clock, and the quantities with superscript (k) indicate transitter-dependent quantities. th Hence the signal eitted by the k transitter consists o a large nuber o onocycle waveors shited to th dierent ties, the onocycle noinally beginning at tie t T c Tc δb. The rae tie T is typically hundred ties the pulses durationt. The value β = T / T p called spreading ratio. A pseudo-rando (i) tie hopping code ( c ), c ) N h ; provides an additional shit in order to avoid catastrophic collision in (i) ultiple-access. The sequence ( b ) is the binary stea generated by the th source and δ is the additional tie shit utilized by binary pulse position odulation. I N > a repetition code is introduced. s Multiple Access Intererence When N u transitters are active in UWB networks, the received signal can odel as: r( t) = ru ( t) + rui ( t) + n( t) (2) Where r u (t) and r ni (t) are the useul signal and MUI contributions at the receiver input. With N s = () () () r ( t) = E w( t T c T δ b ) (3) u and r MUI N = u N s n= 2 = rx () ( n) ( n) ( ( t) E w( t T c T δ b τ rx The average sybol error rate coincides with the average bit error rate P(e) since odulation is binary and corresponds to the probability o isdirecting a reerence bit transitted by the useul transitter. c c p n) )

4 82 IJCSNS International Journal o Coputer Science and Network Security, VOL.8 No.2, February 28 In the case o PPM orthogonal case and under the hypothesis or perect power control, bit error probability ( P e ) is given by [6]: P e = Q( / 2A) (4) 2 T p () 2 A = ( N + ) /( Eb ) (2Rb ( N u ) R ( τ ) dτ ) / N st Tp Where: - R ( t) is the autocorrelation unction o pulse w (t). - Rb is the bit rate. - Q(x) is copleentary error unction. The asyptotic theatrical ( P e ) is represented or high ulti-user intererence (Fig. 3) and low ulti-user intererence (Fig. 4) or dierent bit rate. Pe Kbits/s -7 Mbits/s Mbits/s 2 Mbits/s E b /N (db) Fig. 3: PEB or high ulti-user intererences (Nu=3). Capacity o Multi-user UWB With ore than one user active in the networks, ultipleaccess intererence (MAI) is a actor liiting perorance and capacity, especially or a large nuber o users. The capacity o a UWB ultiple access syste that is dependent on a speciic pulse shape is coputed by Zhao and Haiovich in [7] and suarized in [8]. According to [7], the single user capacity CM PPM (easured in bit by sybol) as a unction o the channel sybol SNR, is given by: CM PPM ( SNR) = log 2 M M (5) E log exp SNR ( v v ) v x 2 = ( ) where v are rando variables with M, and they have the ollowing distribution conditional on the transitted signal x : v : N( SNR,) v : N(,), For ulti-user capacity, each user contributes a raction o the traic in the channel. The aggregate capacity in the channel can be assued as the su o all singer user capacity. Fig. 5 presents the aggregate ulti-user capacity in bits per PPM sybol o UWB as a unction o the nubers o users or various nubers o odulation levels M, and or ixed β T / T = 5. 2 = p 6-PPM 8-PPM 4-PPM 2-PPM Pe Aggregate Capacity Kbits/s -7 Mbits/s Mbits/s 2 Mbits/s E b /N (db) Fig. 4: PEB or low ulti-user intererences (Nu=5) N u Fig. 5: Nuber o Users vs. Aggregate Capacity (bits/sybol) β=5.

5 IJCSNS International Journal o Coputer Science and Network Security, VOL.8 No.2, February UWB Channel Measureent Propagation odeling is vital or the design and developent o wireless counications systes. Channel characterization reers to extracting the channel paraeters ro easured data. The extracted paraeters are used to quantiy the eect o the channel on counication UWB systes using this channel as signal transission ediu e.g., in deriving optial reception ethods, estiating the syste perorance, peroring design tradeos, etc. Wireless channel quality causes undaental liitations on the perorance o wireless networks in underground ines. The propagation in underground channel is coplicated due to scattering and rough suraces diraction. The quality o a channel is a coplex cobination o eects due to path loss, ultipath ading. Measureent procedure and locations The easureents are carried out in an real underground ine gallery i.e., in a orer gold ine now operated by the Canadian Center or Minerals and Energy Technology (CANMET) as a ine-technology laboratory in 53k northern o Montreal, Qc, Canada. To our knowledge, our previous UWB channel easureents and odeling in the ining environent were the irst work in these kinds o environents [9]. The easureents were perored at spatially distributed locations throughout the test area by ixing the transitter in central location and oving the receiver. These easureents are adopted or wireless access point scenarios, short range peer-to-peer systes, and other applications with obile terinals should take the variability o the channel over space into account. Large scale paraeters The average path loss is, in general, expressed as a distance-power law unction. Path loss or every easureent position is obtained by averaging all power proiles. The epirical odel o power attenuation is given by the orula: PL( db) = Plo + nlog ( d / d) + S σ (6) where d is the distance in eters between the transitter and receiver, and S σ is the shadow ading. The path loss exponent n obtained ro peroring a least-square it is.47 and 2.45 or environent LOS and NLOS respectively (Fig. 7). Path Loss [db] NLOS LOS log (d)[] Fig. 7: Scatter plot o path loss vs. Tx/Rx separation. Measureent data Noral CDF it (a) F(x) Shadow ading (db) (b) Measureent data Noral CDF it F(x) Shadow ading (db) Fig. 8: CDF o shadow ading it to Noral distribution- (a) LOS, (b) NLOS. Fig. 6: Exaple o underground ine gallery (LRCS). The shadow ading is considered to have a zero-ean Gaussian distribution with standard deviation σ. The S

6 84 IJCSNS International Journal o Coputer Science and Network Security, VOL.8 No.2, February 28 statistical analysis o S gives σ S =. db or LOS and σ S = 2.94 db or NLOS respectively (see Table. 2). Scenario n Pl d σ S LOS NLOS Table 2: Large scale channel s paraeters. Analysis o RMS Delay Spread The RMS delay spread is a good easure o ultipath spread, it is deined: 2 τ rs = ( τ τ ) p( τ ) dτ / p( τ ) dτ (7) where p(τ ) is the power delay proile and τ is the ean excess delay deined as: τ = ( τ ) dτ / p( τ ) dτ (8) The axiu variations in the RMS delay spread or LOS environent were about.8 ns or 5 db and 23.6 ns, or 2 db. For NLOS environent, they were about 29.7 ns (5 db) and ns (2 db). This indicates that NLOS channels are subject to high spatial variations o the RMS delay spread. This could be attributed to the presence o ore scatterers, and hence ore paths in NLOS environents than in LOS environents. The variation o the ean excess delay also indicates a siilar pattern. The variations o RMS and ean excess were suarized in Tab. 3. Table 3: Average values o RMS and ean excess delay in underground ines (in nsec). Sall Scale Analysis The arrival tie o the irst path can be ound by applying a threshold to the peak o the received signal. For our study, we used two dierent thresholds (e.g., 5 db and 2 db). In order to evaluate the sall-scale aplitude ading statistics, the relative aplitudes over sall-scale areas were calculated. Epirical data ro bins at speciic excess delays were atched to soe typical theoretical distributions or aplitude ading statistics such as lognoral [], Rice [], Rayleigh [2], Nakagai [3], and Weibull [4]. Fro our easureent capaigns we have ound that the Nakagai distribution can be a good it or sall-scale aplitude ading statistics in underground ines. The advantage o Nakagai distribution is that, or special cases such as = and or very large values o, it can be generalized to becoe the Rayleigh and lognoral distributions, respectively. 6. TH-IR-UWB Deodulation/Detection Coherent Rake Receiver As the channel exhibits requency-selective ading due to the extreely wideband nature o the transitted signal, the received signal is inherent with path diversity [6]. A RAKE receiver can be used to exploit the diversity by constructively cobining the separable onocycles ro distinguishable propagation paths or iproving transission perorance. Consider a RAKE receiver with L ingers to collect received signal energy ro the L strongest paths. This can be characterized as a type o tie diversity. The cobination o dierent signal coponents will increase the SNR, which will iprove link perorance. In theory, the receiver structure consists o a correlator ilter that is atched to the transitted waveor that represents one sybol, and a tapped delay line that atches the channel ipulse response. It is also possible to ipleent this structure as a nuber o correlators that are sapled at the delays related to speciic nuber o ultipath coponents; each o those correlators can be called RAKE inger. In UWB systes, requency dependency is taking into consideration [5], the receiver uses several RAKE ingers or each ultipath coponent (MPC) spaced at the Nyquist sapling distance in order to collect the energy in the MPC. The nuber o rake ingers in this case becoes very large [6]. Due to this proble o energy capture, several sipliied RAKE structures have been proposed. Selective Rake (Srake) and Partial rake (Prake). The Srake receiver collect energy ro L strongest MPCs while the Prake collects energy ro the L irst MPCs. The Srake structure has been adopted in this section. Srake outperors Prake because Srake collects ore channel energy than the Prake [6]. The BER siulation results obtained using siulated UWB underground ines channel data is shown in ig. 9.

7 IJCSNS International Journal o Coputer Science and Network Security, VOL.8 No.2, February Regardless o LOS or NLOS, the BER perorance o the SRAKE receivers iproves with the escalating nuber o ingers, since increasing the nuber o ingers enlarges the aount o energy captured by the receiver. In act, the perorances o Prb o Fig 9 depend on the particular realization o each channel ipulse response. Hence, to derive the average perorance o the Selective RAKE, on should average Prb values over several realization o the channel. Pr b - -2 correlation receiver syste; thus a TR syste does not require channel estiation and has weak dependence on distortion. The second one is the Dierential Transitted Reerence. DTR based on the concept o autocorrelation deodulation and dierential encoding [2]. A DTR syste is obtained by a siple odiication to the TR syste in which a reerence is not transitted separately but instead the pulse previously sent is used as a reerence pulse. The third receiver is called MDTR. Chao and R. Scholtz [2] recoended a receiver coniguration which generates a reerence by cobining M consecutively generated sybol waveors in a DTR receiver. An overview o perorance o non-coherent, IR-UWB receivers speciically ocussed transitted reerence schees is given in [7], [2], [22], [23]. The perorance (BER) o dierent receiver structures, versus bit energy per noise power ratio (E b /N ) is presented in igure. -3 Srake (3 branches-los) Srake (5 branches-los) Srake ( branches-nlos) Srake (5 branches-nlos) E/N (db) Fig. 9: Perorance o Selective RAKE receiver under LOS and NLOS (underground ines environent) Non Coherent receiver BER TR DTR MDTR The coherent receiver structures such as RAKE receivers provide good BER perorance at the expense o high coputational and hardware coplexity [7]. For optial coherent reception, several paraeters need to be estiated including ultipath delays, channel coeicients or each delayed ultipath coponents. In UWB systes, the nuber o ultipath coponents is very large, while the power in each o the ultipath coponents is very low [6]. Thereore, the estiation o delays and coeicients o the received ultipath coponents is not a trivial task. Non coherent receivers do not require channel estiation or received pulse estiation, and exploit the rich ultipath channel characteristics o the UWB channel. Aong the non-coherent UWB detectors, we consider three types which are interesting because o their perorance, their robustness to channel ading, and their relative siplicity. The irst one is called transitted reerence (TR). TR schee is not a new technique but had been proposed or spread spectru counication [8]. It has regained popularity with UWB counication systes ater Hoctor and Tolinson [9] proposed a UWB TR syste with a siple receiver structure which captures all o the energy available in a UWB ultipath channel or deodulation at the receiver. TR is a E b /N (db) Fig. : Closed or curves o BEP o the three receiver structures in dense ultipath environent. BER TR (T d =) DTR (T d =) MDTR (T d =) TD (T d =5) -7 DTR (T d =5) MDTR (T d =5) Np Fig. : Siulated bit error rates as a unction o nuber o pulses or dierent non-coherent receivers in AWGN channel Bs Np and or dierent T ( E b /N = 5dB). d

8 86 IJCSNS International Journal o Coputer Science and Network Security, VOL.8 No.2, February 28 In igure, BER perorance is presented versus nuber o pulse and or dierent a ultipath channel with ultipath delay spread tie T. d jointly energy detection and atch iltering or ToA estiation. Here, both detectors use energy detection; hence, coarse and ine search could be executed in the sae bank o integrators. 7. UWB Ranging in underground ines Wireless UWB positioning techniques can provide ideal solution or locate a iner and equipent in real-tie. Soe potential uses include locator UWB-beacons or eergency services, traic anageent, iners and asset tracking in order to increased saety and security in underground ines. The characteristics o UWB signals provide the potential o highly accurate position and location estiation. The ain issue o positioning is how to achieve good estiation ranging accuracy. UWB radio sees to be a prone solution or achieving such high accuracy. Statistical characterization o the UWB ranging In UWB tie-based range estiation, the distance calculation is corrupted by error sources such as easureent error, tie synchronization, inaccuracies in the processing algoriths and the transitter/receiver structure, ultipath ading. All o these types o errors can degrade the positioning accuracy. However, the ajor sources o error in tie-based localization are easureent noise and NLOS propagation error. Measureent noise is usually odeled as a zero-ean Gaussian rando variable, while NLOS error usually has an unknown distribution. Recently it has been shown that the NLOS error in TOA easureents can be odeled by the cobination o zero ean Gaussian and Exponential distribution [24]. In order to odeling the distance easureent error in underground ines, we dierentiate the sall errors caused by ultipath ro the large errors produced by the occurrence o UDP (undetected direct path) conditions. First, aassuing the actual distance between the Tx and the Rx is d; the estiated distance ( d ~ ) in a TOA positioning syste evaluated using two step energy detectors. This estiator is used to increase the perorances o the estiator. First step is a coarse search; the goal is interesting to detect the block where there is a large presence o the cluster. While the second step (ine search) is used detect the irst path with ore accuracy (Fig 2). In act, the idea o two-step ToA estiation ethod can be traced back to [25]. However, the authors eploy Fig. 2: Illustration o two steps ToA estiation. In our siulation, we consider the noralized ranging error estiated using the proposed algorith and ignore the other types o error (i.e., synchronization, theral noise, receiver sensitivity...). Without considering these errors, the noralized ranging error between the transitter and receiver can odeled as: ~ d = d + d * ψ ( d) (9) Fig. 3 shows the Pd o the noralized ranging error. It can be observed that Noral distribution its well the siulated data in LOS case. However, the Weibull distribution its well with the siulated data in NLOS case. The general orula or the Weibull distribution is given by: γ x μ ( x) = α α ( γ ) x μ exp α γ () where α, γ, μ R, α, γ > and x μ, α is scale paraeter, γ is the shape paraeter, and μ is the location paraeter (or zero shit). The values o both distributions paraeters or LOS (i.e., Noral) and NLOS (i.e., Weibull) cases were presented in Fig. 3.

9 IJCSNS International Journal o Coputer Science and Network Security, VOL.8 No.2, February Probability Density Probability Density Noralized Ranging Error Distribution: Noral Siga:.69 Mu: -.92 Distribution: Weibull Zero Shit: Scale: 6.57 Shape: Noralized Ranging Error Fig. 3: Pd characteristic o the noralized ranging error (siulated over UWB underground ines channel realizations) or both LOS (a) and NLOS (b) environents, (Eb/N= db). 8. Conclusion Ultra-wideband technology is revolutionizing the wireless industry by opening doors or new applications as well as copleenting existing wireless systes. With UWB s distinctive properties due to its ability to siultaneously provide both counications and positioning, ake it particularly attractive or industrial ining applications. In the next ew years, UWB will becoe a viable, copetitive wireless technology or transitting inoration rates with high rate date over short distances and low. In Fact, UWB-wireless counication can potentially iprove process operations, product quality, and productivity, boost the saety o ining and iprove rescue operations during disasters. However, UWB wirelesses or underground ining autoation have unortunately been poorly investigated. This paper we explored the undaentals o UWB technology with particular ephasis on ipulse radio (IR) techniques. The goal is to provide an investigation o using UWB technology in industrial ining applications or counication and ranging (localization). More generally, the analysis discussed in this paper can be considered to provide an opening or uture coputer and counications research in UWB counication in underground ining environents. The results presented herein (with other works) are currently exploited in the design o wireless local area networks counication and or localization applications in underground ining environents. Acknowledgent This work was done with support o Université du Québec en Abitibi Téiscaingue (UQAT). The author would like to acknowledge the continuing guidance o LRCS laboratory. The authors also wish to thank the personnel o MMSL-CANMET (Mining and Mineral Sciences Laboratories Canadian Center or Minerals and Energy Technology) experiental ine. Reerences [] F. Nekoogar, Ultra-Wideband Counications: Fundaentals and Applications, Prentice Hall, Hardcover, Published August 25, 24 pages, [2] Sensors Magazine, Wireless or Industry, June 24. [3] A. Chehri, P. Fortier, P.-M. Tardi, Security Monitoring Using Wireless Sensor Networks, 27 Counication Networks and Services Research Conerence, Fredericton, May, 27. [4] R. Zetik, J. Sachs, R. Thoa, UWB localization - active and passive approach, Instruentation and Measureent Technology Conerence, Vol. 2, pp. 5-9, 8-2 May, 24. [5] R. J. Fontana, E. A. Richley, A. J. Marzullo, L. C. Beard, R. W. T. Mulloy, E. J. Knight, An ultra wideband radar or icro air vehicle applications, IEEE Conerence on Ultra Wideband Systes and Technologies, pp. 87-9, 2-23 May, 22. [6] M.-G. DiBenedetto, G. Gianloca, Understanding ultra wideband radio undaentals, Prentive Hall Counications Engineering and Eerging Technologies Series, 24. [7] L. Zhao, A. M. Haiovich, The capacity o a UWB ultiple access counication syste, International Conerence on Counications, ICC 2, pp , May 22. [8] M. Ghavai, Ultra Wideband Signals and Systes in Counication Engineering, John Wiley, 24. [9] A. Chehri, P. Fortier, P.-M. Tardi, Characterization o the Ultra-Wideband channel in conined environents with diracting rough suraces, Subitted to IEEE Transactions on Antennas and Propagation. [] J. R. Foerster, Q. Li, UWB channel odeling contribution ro Intel, IEEE P82.5-2/279r- SG3a, June 22. [] J. Kunisch, J. Pap, Measureent results and odeling aspects or the UWB radio channel, in Proc. IEEE Con. UWB Systes and Technologies (UWBST2), Baltiore, MD, pp. 9 23, May 22,

10 88 IJCSNS International Journal o Coputer Science and Network Security, VOL.8 No.2, February 28 [2] R. J. -M. Craer, R. A. Scholtz, M. Z. Win, Evaluation o an UWB propagation channel, IEEE Trans. Ant. Propagation, Vol. 5, No. 5, pp , May 22. [3] D. Cassioli, M. Z. Win, and A. F. Molisch, The ultra-wide bandwidth indoor channel: ro statistical odel to siulations, IEEE J. Select. Areas Coun., vol. 2, no. 6, pp , August 22. [4] P. Pagani, P. Pajusco, Experiental assessent o the UWB channel variability in a dynaic indoor environent, PIMRC4, Barcelona, pp , Septeber 24. [5] R. C. Qiu, A generalized tie doain ultipath channel and its applications in UWB wireless optial receiver design: Wave-based syste analysis, IEEE Transactions on Wireless Counications, pp , 24. [6] A. F. Molisch, "Ultra wideband propagation channels-theory, easureent, and odeling," IEEE Transactions on Vehicular Technology, vol. 54, no. 5, pp , 25. [7] H. Arslan, A.N. Chen, M. G. Benedetto, Ultra- Wideband Wireless Counication, Wiley Interscience, Hoboken, New Jersey, 26. [8] C. K. Rushorth, Transitted-reerence techniques or rando or unknown channels, IEEE Trans. Inoration Theory, Vol., No., pp , Jan [9] R. Hoctor and H. Tolinson, Delay-hopped, transitted reerence RF counications, IEEE Ultra Wideband Systes and Technologies, Baltiore, pp , May 22. [2] Y. L. Chao and R. Scholtz, Optial and suboptial receivers or ultra-wideband transitted reerence systes, Proceedings GLOBECOM 3, pp , San Francisco, CA, -5 Dec. 23. [2] S. Zhao, H. Liu, Z. Tian, A decision eedback autocorrelation receiver or pulsed ultra-wideband systes, Proc. IEEE Rawcon 4, Atlanta, GA, Sep. 24. [22] L. Stoica L, A. Rabbachin, I. Opperann, Ipulse radio based non-coherent UWB transceiver architectures - an exaple, IEEE International Conerence on Ultra Wideband Systes and Technologies, Waltha, USA: , 26. [23] A. Rabbachin, I. Opperann, Coparison o UWB auto-correlation and transitted reerence schees, IEE International Conerence on Ultra- Wideband, 5-8 Sept. 25 [24].B. Denis, J. Keignart, N. Daniele, Ipact o NLOS propagation upon ranging precision in UWB systes, UWBST, pp , Reston, USA, Noveber 23. [25]. S. Gezici, Z. Sahinoglu, A. Molisch, H. Kobayashi, H. Poor, A Two-Step Tie o Arrival Estation Algorith or Ipulse Radio Ultra Wideband Systes, Proc. 3th European Signal Processing Conerence (EUSIPCO 25), Antalya, Turkey, Sept. 4-8, 25.