Edwin Jan Klein - Universiteit Twente

More documents

Recommendations

Info

Chapter 7 7.1 Introduction In Chapter 1 the WDM-PON was introduced as a possible successor of traditional passive optical networks. In a traditional PON, as shown in Figure 7.1, only two wavelengths are used to transport data. This severely limits the bandwidth available to each user because the information going to- and coming from multiple users has to be multiplexed onto a single wavelength. Another problem is that when the network is upgraded, it is likely that all connected users have to upgrade as well because they all share the same wavelength channels. Head Office OLT Figure 7.1. Traditional PON architecture. A WDM-PON uses multiple wavelengths, as shown in Figure 7.2. Since the network is still based on a passive splitter all users receive the downstream wavelengths. However, the ONUs at the user’s premises can now filter out a specific wavelength with a service for which they have a subscription. This allows the provider to run many services, for instance providing different bandwidths, side by side. In addition, if an upgrade is required, this can be run on a different wavelength to which the users can then migrate gradually. Head Office OLT λ1 λ2 λ1, λ2, λ3 λ1, λ2, λ3 λx ,λy ,λz λ1 λ2 Figure 7.2. WDM PON architecture using a passive splitter (for use with the NAIS transceiver). λ2 λz The WDM-PON therefore has the potential to be more cost effective than the traditional PON. Currently, however, the cost of the required optical components is one of the factors limiting the adoption of the WDM-PON. This is especially true for the ONUs for which the cost cannot be shared amongst multiple users. The WDM- PON ONUs can be very expensive because they have to be able to receive and send data in multiple wavelengths which requires optical filtering and a tunable laser 160 λ1 λx λ1 λ2 λ1, λ2, λ3 λ1, λ2, λ3 λy
Densely integrated devices for WDM-PON source. The NAIS and Broadband Photonics projects, within which the work presented in this thesis was carried out, have in common that they both approach the problem of cost (amongst other issues) through the miniaturization and integration of the required network components. The approach to the actual implementation of the ONUs, however, is quite different, even though both are intended for the WDM-PON, and has implications for the rest of the components in the network. The transceiver at the ONU proposed in the NAIS project takes an active approach. Each user receives all downstream channels as shown in Figure 7.2. From these channels a single channel is then selected. For the upstream signal a single wavelength channel is used which has to be set explicitly (by the user or the Head Office H.O.), through the hardware configuration of the transceiver. The light of the required wavelength for this channel is then generated on chip, by spectral slicing of a broadband source. This has the advantage that old PON networks are easily upgraded because the passive splitters can still be used. The Broadband Photonics project on the other hand has a semi-passive approach using colorless transceivers. From the head-office each ONU transceiver receives two paired wavelength channels. One channel contains the downstream data signal while the other is a continuous wave (CW) signal which is amplified (using a SOA), modulated and returned as the up stream data signal as is shown in Figure 7.3. Head Office OLT λ1, λ3, λ5 λ1, λ2 λ2 ,λ4 ,λ6 Figure 7.3. WDM PON architecture using a wavelength filter (AWG, OADM etc.) in combination with reflective SOA transceivers. λ6 The main advantage of this is that the wavelength can be changed transparently, without the need to reconfigure the transceiver because the transceiver will modulate and reflect any wavelength that is sent down stream (with the exception of the designated downstream wavelengths). This also poses a problem, however, since each transceiver should receive no more than two wavelengths which is not possible when passive splitters are used. In order to send only specific channels to a transceiver the Broadband Photonics network therefore uses reconfigurable optical λ-routers instead of the passive splitters (in the strictest sense of the word this network is no longer a PON network but is nonetheless regarded as such in the following text). These optical routers are useful for several reasons. First of all only those channels that are used by a specific user are sent to that user. This is not only a requirement for the use of the colorless transceiver but it also improves the power budget since no power is wasted in sending a signal to a user that does not require it. It is also beneficial to the network security as a whole since it is more difficult if not impossible for users to “eavesdrop” on signals destined for other users. Another important reason is the fact that, since the device can be reconfigured, the network can be better balanced. It is for instance possible to allocate 161 λ2 λ5 , λ6 λ3 ,λ4 λ4
Page 1 and 2:
DENSELY INTEGRATED MICRORING- RESON
Page 3 and 4:
DENSELY INTEGRATED MICRORING- RESON
Page 5:
To my parents…
Page 8 and 9:
esonators is limited by the spacing
Page 10 and 11:
iv Samenvatting Dit proefschrift be
Page 12 and 13:
schakelaar “aan” is, dan is de
Page 14 and 15:
3.6.2 Overlap loss reduction.......
Page 17 and 18:
Chapter 1 Introduction The devices
Page 19 and 20:
1.2 Broadband to the home 3 Introdu
Page 21 and 22:
1.3 Bringing Fiber to the Home 5 In
Page 23 and 24:
Figure 1.5. The technological scope
Page 25:
9 Introduction these basic paramete
Page 28 and 29:
Chapter 2 2.1 Introduction Integrat
Page 30 and 31:
Chapter 2 light Icav2 in the cavity
Page 32 and 33:
Chapter 2 ∆ = ⎛ β r − β g
Page 34 and 35:
Chapter 2 2.3.3 Combined model The
Page 36 and 37:
Chapter 2 FC 4µ 1µ 2 ⋅ χr = 2
Page 38 and 39:
Chapter 2 P Through /P In (dB) 0 -1
Page 40 and 41:
Chapter 2 The figure shows that it
Page 42 and 43:
Chapter 2 In − jϕr1 2 Figure 2.1
Page 44 and 45:
Chapter 2 differences between the f
Page 46 and 47:
Chapter 2 M = ⋅ FSR = N ⋅ FSR F
Page 48 and 49:
Chapter 2 the heat will therefore f
Page 50 and 51:
Chapter 2 index. Although this can
Page 52 and 53:
Chapter 2 resonators for instance h
Page 54 and 55:
Chapter 2 a rather complex MEMS bas
Page 56 and 57:
Chapter 3 3.1 Introduction The most
Page 58 and 59:
Chapter 3 3.2.1 Drop port on-resona
Page 60 and 61:
Chapter 3 3.2.3 Filter bandwidth In
Page 62 and 63:
Chapter 3 (resulting in a higher ba
Page 64 and 65:
Chapter 3 Figure 3.11. Residual cha
Page 66 and 67:
Chapter 3 3.3 Geometrical design ch
Page 68 and 69:
Chapter 3 resonator and the port wa
Page 70 and 71:
Chapter 3 Another important fact wh
Page 72 and 73:
Chapter 3 The flowchart of the desi
Page 74 and 75:
Chapter 3 While the resonator in th
Page 76 and 77:
Chapter 3 3.6.1 Port waveguide and
Page 78 and 79:
Chapter 3 overlap losses are only 0
Page 80 and 81:
Chapter 3 3.7 Component design cons
Page 82 and 83:
Chapter 3 By maximizing the power t
Page 84 and 85:
Chapter 3 important, due to the fac
Page 86 and 87:
Chapter 4 4.1 Introduction As was s
Page 88 and 89:
Chapter 4 4.2.1 Transient response
Page 90 and 91:
Chapter 4 Under the condition that
Page 92 and 93:
Chapter 4 Figure 4.4. Screenshot of
Page 94 and 95:
Chapter 4 The equations required to
Page 96 and 97:
Chapter 4 using an approach based o
Page 98 and 99:
Chapter 4 4.5.1 Architecture The Au
Page 100 and 101:
Chapter 4 4.5.2 Simulation method A
Page 102 and 103:
Chapter 4 Figure 4.9. A Primitive w
Page 104 and 105:
Chapter 4 Listing 4.3. Pseudo code
Page 106 and 107:
Chapter 4 Listing 4.4. Pseudo code
Page 108 and 109:
Chapter 4 simulated output spectra
Page 110 and 111:
Chapter 4 Figure 4.21. The reflecto
Page 112 and 113:
Chapter 4 P Drop /P In (dB) 0 -10 -
Page 114 and 115:
Chapter 4 Figure 4.27a. OADM with r
Page 116 and 117:
Chapter 4 will propagate through th
Page 118 and 119:
Chapter 4 As demonstrated by these
Page 120 and 121:
Chapter 4 What is more interesting
Page 122 and 123:
Chapter 4 method it does allow time
Page 124 and 125:
Chapter 4 complex than the interact
Page 126 and 127: Chapter 5 5.1 Introduction The mate
Page 128 and 129: Chapter 5 wafer is used (step 1). T
Page 130 and 131: Chapter 5 range of 0.9 to 1.1 µm.
Page 132 and 133: Chapter 5 The process flow with the
Page 134 and 135: Chapter 5 5.3.3 Mask layout The big
Page 136 and 137: Chapter 5 5.4. Fabrication and mask
Page 138 and 139: Chapter 5 1. Definition of the vert
Page 140 and 141: Chapter 5 5.4.2 Fabrication The fab
Page 142 and 143: Chapter 5 Figure 5.21. Process flow
Page 144 and 145: Chapter 5 removed using lift-off of
Page 146 and 147: Chapter 6 6.1 Introduction About ha
Page 148 and 149: Chapter 6 In early, comparatively s
Page 150 and 151: Chapter 6 that the TEOS has a very
Page 152 and 153: Chapter 6 The most important aspect
Page 154 and 155: Chapter 6 6.2.3 Chromium heater des
Page 156 and 157: Chapter 6 Therefore, even if the si
Page 158 and 159: Chapter 6 Next, the heater was heat
Page 160 and 161: Chapter 6 6.3 Characterization of s
Page 162 and 163: Chapter 6 polarization maintaining
Page 164 and 165: Chapter 6 the resonators in this gr
Page 166 and 167: Chapter 6 Although the fitted coupl
Page 168 and 169: Chapter 6 a cladding thickness of 3
Page 170 and 171: Chapter 6 6.4 A wavelength selectiv
Page 172 and 173: Chapter 6 A possible explanation fo
Page 175: Chapter 7 Densely integrated device
Page 179 and 180: Densely integrated devices for WDM-
Page 183 and 184: 7.2.2 Spectral measurements Densely
Page 189 and 190: Power Densely integrated devices fo
Page 191 and 192: 7.3.4 Characterization Densely inte
Page 195 and 196: Figure 7.31a. Eye pattern of the re
Page 197: Densely integrated devices for WDM-
Page 200 and 201: Chapter 8 8.1 Introduction In most
Page 202 and 203: Chapter 8 amount of power will be d
Page 204 and 205: Chapter 8 Another problem that may
Page 206 and 207: Chapter 8 halves that are each comp
Page 209 and 210: Chapter 9 Discussion and Conclusion
Page 211 and 212: Appendix A. Mason’s rule Several
Page 213 and 214: Appendix B. Crosstalk Derivation Th
Page 215 and 216: List of Acronyms ADSL - Asymmetric
Page 217 and 218: List of Symbols Q - Quality Factor
Page 219 and 220: Bibliography [1] Dan Schiller, “E
Page 221 and 222: Regions,” IEEE Photonics Technolo
Page 223 and 224: [66] B. E. Little, S. T. Chu, J. V.
Page 225 and 226: [97] H. Haeiwa, T. Naganawa and Y.
Page 227 and 228:
[135] A. Driessen, D. H. Geuzebroek
Page 229 and 230:
[165] T. Barwicz, M. R. Watts, M. A
Page 231 and 232:
Publication List Patents (pending)
Page 233 and 234:
R. Dekker, L.T.H.Hilderink, M.B.J.
Page 235 and 236:
Networks (BB Photonics),” Interna
Page 237 and 238:
Dankwoord/Acknowledgements Zo, dit
Page 239:
223
show all

Edwin Jan Klein - Universiteit Twente

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?