Appendix AAbstract –The s11 parameter is defined as the reflected voltage divided by the incident voltageand is the same as the return loss which is equal to – RL = −20 log( s11 )It is desirable to have as little reflected voltage as possible and hence a return loss of-∞. Practically this is not possible however, if the Voltage Standing Wave Ratio(VSWR) is less than 2, the antenna performs acceptably. In terms of return loss, aVSWR of less than 2 compares to a return loss less than -9.54dB (common practice isto use -10dB for simplicity).The results obtained from research papers and displayed in this report are useddetermine which antenna will be used for experimentation. However, there are alsoother factors which are taken into consideration such as ease of fabrication and also asnot all data was able to be obtained for all antennas, some data may be extrapolatedfrom certain antennas for other antennas.It is also worth noting that the group delay and phase response are related by – ∂θ (ω ) Groupdelay = − ∂ωwhere θ (ω ) is the phase angle dependent on frequency. Hence a linear phaseresponse gives a constant group delay.
Bicone, Monocone, LPDA, F-probe, Simulated HornFigure 1 – Images of simulated antenna from Figure 2 – s11 parameter results and radiation pattern of antennas in Figure 1 from Discussion –Figure 2 shows the return loss of the different antennas from figure 1 betweenfrequencies of DC and 10GHz. It can be seen from figure 2 that the best performersare the bicone (with a bandwidth of >7GHz) and the monocone (with a bandwidth ofapproximately 7GHz). The horn antenna performed the worst with a bandwidth ofapproximately 500MHz and the LPDA and f-probe antennas performed rathersimilarly with the LPDA having a rather poor return loss over certain frequencies anda good return loss over others due to the design of the antenna.The radiation patterns depicted in figure 2 are for the monocone antenna (top) and thef-probe antenna (bottom). Because the f-probe antenna and the horn antenna are bothdirectional antennas, it is assumed that the horn antenna with have a similar radiationpattern to that of the f-probe antenna. The radiation pattern for the monocone showsan omnidirectional pattern which is expected due to the symmetry of the antenna’sdesign. It is assumed that the radiation pattern for the bicone will be similar to that ofthe monocone due to the similarities of design. The radiation pattern of the LPDAhowever is unknown from figure 2 although figure 15 does show the radiation patternfor a different LPDA and hence this can be used as a comparison.Although the cone antennas have the best performance, these are harder to fabricatethan say the LPDA antenna and hence must be taken into consideration.
Section 2 - Helix AntennasFigure 3 – Helix antenna and s11 parameter results from Figure 4 – Helix antenna with cone feed and s11 parameter results from Discussion –
The helix antenna shown in figure 3 performs particularly well at certain frequenciesbut poorly over the range of 2 – 6GHz. The helix antenna shown in figure 4 (with acapacitive coupling) performs the best out of all the helix antennas with a bandwidthof slightly less than 4GHz. All other antennas have relatively poor bandwidthcompared to the capacitive coupled helix antenna. Although the radiation pattern isnot given, it can be assumed that due to the symmetrical design of the antenna that theradiation pattern will be omnidirectional. Also, due to the complex shape of theantenna, it will be harder to fabricate than an average planar antenna.
Square Planar MonopoleFigure 5 – Square Monopole antenna and s11 parameter results from Figure 6 – Radiation pattern of square monopole antenna at 5GHz from Discussion –Figure 5 shows the square planar monopole antenna having a large bandwidth ofapproximately 10GHz. The radiation pattern is also shown at 5GHz and can be seento be omnidirectional. Figure 1 however does show that this large bandwidth isdependent on the notches in the design however this does not make the fabricationmuch more difficult.
Square Monopole with various feeding stripsFigure 7 – Square monopole antenna with various feeding strips and results of s11parameter from Figure 8 – Radiation pattern of square monopole antenna with various feeds from Discussion –Figure 7 shows that a square monopole antenna with a trident shape or 2 branchfeeding strip performs better than that of a simple feeding strip. The trident anddouble feeding strip antennas perform very similarly however the trident feeding stripdoes have a slightly larger bandwidth of approximately 10GHz. Once again theradiation pattern is omnidirectional and is rather simple to fabricate.
Step shaped Planar MonopoleFigure 9 – Step shaped planar monopole antenna and results of s11 parameter from Figure 10 – Radiation pattern of step shaped planar monopole antenna from Discussion –The step shaped planar monopole antenna performs quite poorly over the UWBfrequency range with a bandwidth of only 4.4GHz. The radiation pattern is similar tothat of the previous monopole antennas and it is simple to build however thebandwidth is rather poor.
U shaped Planar MonopoleFigure 11 – U shaped planar monopole antenna and results for s11 parameter from Discussion –The U shaped planar monopole antenna performed slightly better than the step planarmonopole antenna with a bandwidth of approximately 5GHz however this is stillrather poor over the UWB frequency range.
Cylindrical planar monopoleFigure 12 – Dimensions of cylindrical planar monopole antenna from Figure 13 – s11 parameter results for cylindrical planar monopole antenna from Discussion –The cylindrical planar monopole antenna performs very similarly to the U shapedplanar monopole antenna which has a rather poor bandwidth over the UWB frequencyrange.
Cross plate MonopoleFigure 14 – Cross plate monopole antenna and s11 parameter results from Discussion –The cross plate monopole antenna can be seen to have a bandwidth of approximately10GHz which is good for a UWB application. The antenna performs better with abent cross plate rather than a planar cross plate.
LPDAFigure 15 – Dimensions for LPDA and s11 parameter results from Figure 16 – Radiation pattern for LPDA from Figure 17 – Plot showing the group delay for LPDA from Monopole Antenna
Discussion –Figure 15 shows that the LPDA has a bandwidth of approximately 10GHz which issuitable for a UWB application however the return loss does vary significantly withfrequency as shown by all the downward spikes in the plot. Figure 15 also shows thatthe original design (shown in figure 15) performs better than the band-notched design(not shown). The radiation pattern of the LPDA can be seen to be omnidirectional,although it is not the best radiation pattern of antenna previously looked at. Figure 17does show the LPDA to have a relatively constant group delay over the UWBfrequency range which is a desirable characteristic of an antenna.
Monopole AntennaFigure 17 – Dimensions of monopole antenna from Figure 18 – s11 parameter results for monopole antenna with various dimensions fromDiscussion –It can be seen in figure 18 that the monopole antenna has a poor return loss betweenapproximately 5 and 7 GHz which is rather poor compared to some of the previousantennas. It can also be assumed that the radiation pattern of this antenna will berather similar to that of previous planar monopole antennas and hence has no standout characteristics above other antennas.
Planar DipoleFigure 19 – Dimensions of planar dipole antenna from Figure 20 – s11 parameter results and radiation pattern of the planar dipole antennafrom Discussion -Figure 20 shows a poor return loss over the UWB frequency range (particularly above6GHz). Its radiation pattern is very similar to that of other planar monopole antennas.
Planar MonopoleFigure 21 – Dimensions of planar monopole antenna from Figure 22 – s11 parameter results of planar monopole antenna from Discussion –Figure 22 shows that the monopole antenna shown in figure 21 has a bandwidth ofapproximately 10GHz; this is an acceptable bandwidth for an UWB application.
Vivaldi AntennaFigure 23 – Dimensions of Vivaldi antenna from Figure 24 – VSWR plot and group delay plot of Vivaldi antenna from Figure 25 – Dimensions of a second Vivaldi antenna from 
Figure 26 – VSWR plot and group delay plot of second Vivaldi antenna from 
Discussion –Unlike previous antenna data shown, the data for the Vivaldi antenna gives a plot ofthe VSWR rather than the return loss. As stated in the abstract, having a return loss ofless than 10dB is approximately the same as having a VSWR of less than 2. Figure24 shows that the first type of Vivaldi antenna has a simulated bandwidth of >9GHzand a measured bandwidth of 7.5GHz and a relatively constant group delay (except at9GHz where strange things happen for reasons unknown). Figure 26 shows that thesecond type of Vivaldi antenna has a simulated bandwidth of >9GHz and a measuredbandwidth of approximately 8GHz and a relatively constant group delay.
Planar Dipole AntennasFigure 27 – Circular, bow-tie and elliptical dipole antennas from Figure 28 – VSWR plot and frequency response of dipole antennas from Figure 29 – Radiation patterns of dipole antennas from Discussion –Figure 28 shows that both the circular and elliptical dipole antennas have largebandwidths while that of the bow-tie dipole is slightly worse. Figure 28 also showsthe phase response of the antennas and shows that the elliptical antenna has anonlinear phase response (and hence not a constant group delay) whereas the circular
and bow-tie antenna have rather linear phase responses with that of the bow-tie beingslightly better than that of the circular. The radiation patterns appear to be similar tothose of previous planar antennas.
Circular Monopole AntennaFigure 30 – Circular monopole antenna from Figure 31 – s11 parameter results of circular monopole antenna from DiscussionFigure 31 shows that the bandwidth of the circular monopole antenna to be less than7.5GHz over the UWB frequency range. The impedance bandwidth is from 3.1GHzto approximately 10GHz and hence does not cover the entire UWB frequency range.
Microstrip AntennaFigure 32 – Microstrip antenna from Figure 33 – s11 parameter results of microstrip antenna from DiscussionThe microstrip antenna shows a relatively poor impedance bandwidth beingapproximately 6GHz between 4.5GHz and 10.5GHz.
Conclusion –Firstly, because there were many antennas which had a suitable bandwidth for UWB,all of the others can be disregarded. This leaves the bicone, monocone, helix withcapacitive coupling, square planar monopole, cross plate monopole, LPDA, planarmonopole, Vivaldi and circular and elliptical dipole antennas. All of these antennashave very similar (omnidirectional) radiation patterns so this does not have an effecton the choice of antennas.Secondly, the bicone, monocone and helix antennas are the most difficult antennas tofabricate and do not provide much, if any, difference compared to other, easier tofabricate antenna designs. This leaves the square planar monopole, cross platemonopole, LPDA, planar monopole, Vivaldi, circular monopole, microstrip and thecircular and elliptical antennas.From this point, the major determining factor has become the phase response or groupdelay of the antenna. Because all of the antennas have very similar radiation patternsand bandwidths, the antennas that phase responses were unable to be obtained for willbe disregarded hence leaving the LPDA, Vivaldi, circular monopole, microstrip andthe circular and elliptical dipole antennas. Also, because the elliptical dipole has anon-linear phase response, it can also be disregarded. This leaves a choice betweenfive antennas, the LPDA, Vivaldi and the circular dipole. It was determined that theLPDA and circular dipole were too difficult to simulate hence the Vivaldi, circularmonopole and microstrip antennas will be simulated.
Bibliography – Sibbile, A 2005, ‘Modulation Scheme and Channel Dependence of Ultra-Wideband Antenna Performance’, IEEE Antennas and Wireless Propagation Letters,vol. 4 – Yang, Y, et al. ‘The Design of Ultra-wideband Antennas with PerformanceClose to the Fundamental Limit’, Virginia Tech Antenna Group, Blacksburg, VA,USA – Wong, K.L. ‘High-Performance Ultra-Wideband Planar Antenna Design’, Dept.of Electrical Engineering National sun Yat-Sen University Kaohsiung, Taiwan. – Chen, S.Y., et al. 2006, ‘Unipolar Log-Periodic Slot Antenna Fed by a CPW forUWB Applications’, IEEE Antennas and Wireless Propagation, vol. 5 – Xiao-Xiang, HE, 2009, ‘New band-notched UWB antenna’, College ofInformation Science and technology, Nanjing University of Aeronautics andAstronautics, Nanjing, P.R. China – Zhao, CD, 2004, ‘Analysis on the Properties of a Coupled Planar Dipole UWBAntenna’, IEEE Antennas and Wireless Propagation Letters, vol. 3 – Choi, SH, 2003, ‘A new Ultra-Wideband Antenna for UWB Applications’,Microwave and Optical Technology Letters, vol. 40, no. 5, Mar 2004 – Mehdipour, A. 2007, ‘Complete Dispersion Analysis of Vivaldi Antenna forUltra Wideband Applications’, Progress in Electromagnetics Research, pp 85-96. – Hecimovic, N. ‘The Improvements of the Antenna Parameters in Ultra-Wideband Communications’, Ericsson Nikola Tesla, d.d., Croatia. – Liang, J. 2005, ‘Study of Printed Circular Disc Monopole Antenna for UWBSystems’, IEEE Transactions on Antennas and Propagation, vol. 53, no. 11. – Lim, K.-S. 2008, ‘Design and Construction of Microstrip UWB Antenna withTime Domain Analysis’, Progress in Electromagnetics Research, vol.3 pp 153 – 164.
Appendix B 9m Coaxial Cabling Magnitude 9m Coaxial Cabling Phase 0 0 -500 -5 -1000 -10 -1500 Magnitude, dB Phase, rad -15 -2000 -2500 -20 -3000 -25 -3500 -30 -4000 2 4 6 8 10 12 2 4 6 8 10 12 Frequency, GHz Frequency, GHz (a) (b)Figure 1. Cable magnitude and Phase response for 9m of cablesFigure 2. Cable magnitude and Phase response for 9m of cables
Project Management Plan – Ultra Wideband Antennas OFFCDT Andrew Spear School of Engineering and Information Technology UNSW@ADFAPurposeThe project will focus on antennas suitable for a typical Ultra Wideband (UWB) of3.1GHz to 10.6GHz. The purpose is to research existing UWB antenna technology,identify key features of UWB antennas, analysing, characterising and constructingparticular antenna designs and to give a comparison of the relative performance andtrade-offs between fabrication, cost and performance.ObjectivesThe aim of the project was to develop a diagnostic tool in order to produce an UltraWideband (UWB) antenna with a linear phase response in order to give a constantgroup delay using a passive equalisation network.MilestonesThe milestones for the project are –1. To obtain one or two UWB antenna designs which can be simulated and constructed for testing. 1.1. Draft a document on UWB antenna technology, highlighting the particular requirements for UWB systems and applications. 1.2. Have construction details and published performance characteristics/ targets for antenna designs.2. To simulate the antenna designs and obtain data which can be manipulated using MATLAB. 2.1. Draft a document detailing the process to be used for simulation of the antenna designs and also the desired data which is to be extracted. 2.2. Draft a report detailing the data obtained from simulation and a comparison of this data to what theory predicts.3. To characterise the permittivity of the antenna substrate over the frequency range of 3.1GHz – 10.6GHz. 3.1. Have constructed test boards. 3.2. Obtain an accurate simulation model for the antenna substrate.4. To construct the antenna design and conduct real life testing of the system. 4.1. Have constructed the antenna. 4.2. Draft a report detailing the data obtained from the real life testing and a comparison of this to theory/data obtained from simulation.5. To develop a diagnostic tool in order to determine whether a particular antenna can be equalised to have a constant group delay over the UWB frequency spectrum.
5.1. Draft a document detailing the requirements for a system to give a linear phase response. 5.2. To produce the diagnostic tool to be operated in MATLAB. 5.3. To adapt the diagnostic tool in order to simulate equalisation networks of varying complexity. 5.4. To draft a report detailing the data obtained and a comparison to what is predicted by theory/simulation (if simulation is possible).Key Dates • Week 25/26 – VIVA • Week 40 – 15 minute seminar • Week 44 – Thesis due
Projected TimelineTask Task Start (Week No.) Finish Total time RemarksNo. (Week spent No.) (Weeks)1 Research antenna theory and UWB antenna 11 13 3 Completed and systems2 Milestone 1.1 11 13 3 Completed3 Milestone 1.2 11 13 3 Completed4 Practice using simulation software 13 13 1 Completed5 Conduct simulations 14 20 5 Completed6 Milestone 2.1 13 13 1 Completed7 Milestone 2.2 16 16 1 Completed8 Milestone 3.1 20 20 1 Completed9 Conduct testing on test boards 21 21 1 Completed10 Milestone 3.2 22 24 3 Completed11 Milestone 4.1 20/29 20/30 3 One antenna built12 Conduct real life testing of antenna 23 32 6 Completed 1-port testing for the 1 antenna13 Milestone 4.2 33 33 114 Milestone 5.1 29 30 215 Milestone 5.2 31 32 216 Simulate system to give linear phase response 32 33 217 Milestone 5.3 34 35 218 Test system to give linear phase response 36 37 219 Milestone 5.4 37 37 120 Collaborate work/write thesis 40 44 4
The above matrix works as follows – • The numbers across the top represent the week number in the year • The numbers down the left represent the task number given in the projected timeline -Anticipated Obstacles –The construction time of the antenna is unknown and hence in the event that the fabrication takes significantly longer than expected the systemto give a linear phase response can begin to be designed based on the results of simulations.Obstacles Encountered –It took a lot longer than expected to get an accurate simulation result from the simulation software (CST Microwave Studio) and hence thatpushed back the construction of the antennas. However, the time taken to construct the antennas was not as long as expected and hence helpedme catch up some time.The construction time of the antennas was approximately 1 week hence it was not a significant issue in terms of time constraints.The biggest obstacle encountered was obtaining the 2-port measurements of the antenna network. This process took much longer than expecteddue to the several iterations of testing which had to be done in order to achieve accurate results. The testing was first conducted outside theanechoic chamber and it was found that there was too much backscatter in the transmission results to do any manipulation of the data inMATLAB. The test was then moved inside the anechoic chamber but in order to do this 9m of cable had to be used to connect the antennas tothe network analyser. These cables introduced a significant amount of attenuation in the results which was reducing the magnitude of thetransmission down to the threshold of the network analyser. The next iteration involved a setup similar to the first (outside of the anechoicchamber) but using the anechoic chamber foam tiles in an attempt to reduce the amount of backscatter. It was found that this did not workparticularly well and finally the network analyser was moved closer to the anechoic chamber to reduce the amount of cable used to connect theantennas to the network analyser which gave accurate results.
Scope Changes –The scope of the project began as investigating UWB antennas and technologies and developing a system to linearise the group delay of theantennas hence giving it a constant group delay. This shifted to producing a tool to take in the 2-port parameters of an antenna and determine theeffect of equalisation on the antennas using a passive network of varying complexities.