method combined with the modified matrix pencil (MMP) method. The computational time of the FDTD method can be reduced to 30% by the MMP method. Our numerical simulation shows that the intrinsic passband at the harmonic band is suppressed for this low-pass microstrip filter with the UC-EBG structure in the ground plane. ACKNOWLEDGMENTS This work is supported partially by Natural Science Foundation of China under contract nos. 60271005 and 60371002, and the Ex- cellent Youth Natural Science Foundation of China under contract no. 60325103. REFERENCES 1. E.R. Brown, C.D. Parker, and E. Yablonovitch, Radiation properties of a planar antenna on a photonic-crystal substrate, J Opt Soc Amer B Opt Phys 10 (1993), 404 – 407. 2. M.M. Sigalas, R. Biswas, and K.M. Ho, Theoretical study of dipole antennas on photonic band-gap materials, Microwave Opt Technol Lett 13 (1996), 205–209. 3. R.D. Meade, K.D. Brommer, A.M. Rappe, and J.D. Joannopoulos, Photonic bound states in periodic dielectric materials, Phys Rev B Condens Matter 44 (1991), 13772–13774. 4. F. Yang, K. Ma, Y. Qian, and T. Itoh, A uniplanar compact photonic band-gap (UC-PBG) structure and its application for microwave cir- cuits, IEEE Trans Microwave Theory Tech 47 (1999), 1509 –1514. 5. V. Radisic, Y. Qian, R. Cocciili, and T. Itoh, Novel 2D photonic band-gap structures for microstrip lines, IEEE Microwave Guided Wave Lett 8 (1998), 67–71. 6. I. Rumsey, M. Piket-May, and P.K. Kelly, Photonic band-gap struc- tures used as filters in microstrip circuits, IEEE Microwave Guided Wave Lett 8 (1998), 336 –338. 7. E.N.R.Q. Fernandes, P.H.F. Silva, M.A.B. Melo, and A.G. d’Assunc ¸a ˜o, A new neural network model for accurate analysis of microstrip filters on PBG structure, Euro Microwave Conf Dig, Milan, Italy, 2002. 8. A. Taflove, Advances in computational electrodynamics: The finite- difference time-domain method, Artech House, Boston, 1998. 9. K.S. Yee, Numerical solution of initial boundary value problems involving Maxwell’s equations in isotropic media, IEEE Trans Anten- nas Propagat 14 (1966), 302–307. 10. B. Lu, D. Wei, B.L. Evans, and A.C. Bovik, Improved matrix pencil methods, 1998. 32 nd Asilomar Conf, Signals, Syst Computers, 1998, pp. 1433–1437. 11. J. Razavilar, Y. Li, and K.J.R. Liu, Spectral estimation based on structured low rank matrix pencil, Proc IEEE Int Conf Acoust Speech Signal Processing, Atlanta, Ga, 1996, pp. 2503–2506. 12. Y. Hua and T.K. Sarkar, Matrix pencil method for estimating param- eters of exponentially damped/undamped sinusoids in noise, IEEE Trans Acoustics Speech Signal Processing 38 (1990), 814 – 824. 13. R.S. Chen, D.X. Wang, E.K.N. Yung, and J.M. Jin, Application of the multifrontal method to the vector FEM for analysis of microwave filters, Microwave Opt Technol Lett 31 (2001), 465– 470. © 2004 Wiley Periodicals, Inc. DRIVING HIGH-SPEED VCSELs K. Minoglou, 1 E. D. Kyriakis-Bitzaros, 1,2 S. G. Katsafouros, 1 D. Syvridis, 3 G. Halkias 1 1 Institute of Microelectronics NCSR “Demokritos” 15310 Ag. Paraskevi, Greece 2 Department of Electronics TEI of Piraeus 12244 Egaleo, Greece 3 Department of Informatics and Telecommunications University of Athens 15784 Athens, Greece Received 8 June 2004 ABSTRACT: A comparison of the small-signal and transient responses of high-speed VCSELs using ideal as well as realistic current and volt- age drivers is performed. A nonlinear VCSEL equivalent model is imple- mented and the associated parameters are extracted from the dc and ac measurements of a commercially available packaged device. The simula- tions, using the extracted realistic VCSEL models, show that the employ- ment of voltage drivers results in an improvement of 74% of the 3-dB cutoff frequency of the optical-current signal. Moreover, an 80% reduc- tion of the rise and fall time and a 66% reduction of the signal delay are observed. © 2004 Wiley Periodicals, Inc. Microwave Opt Technol Lett 44: 41– 45, 2005; Published online in Wiley InterScience (www. interscience.wiley.com). DOI 10.1002/mop.20541 Key words: optical interconnections; laser drivers; vertical-cavity sur- face-emitting lasers (VCSELs); circuit model 1. INTRODUCTION The design of a laser diode driver (LDD) is a challenging task, since clear eye diagrams at high speeds require abrupt signal edges and low jitter. Traditionally, LDDs were designed to drive edge- emitting lasers (EELs) and were based on differential-pair topol- ogies acting as current-pulse sources for the laser diodes [1]. This choice of circuit topology is straightforward for EELs exhibiting low series resistance, which is very often combined with consid- erable series parasitic inductance. At first glance, the design task of LDDs for vertical cavity surface emitting lasers (VCSELs) seems relaxed, due to the lower threshold and modulation currents re- quired by these devices. However, high-speed VCSELs present significantly larger series resistance in combination with relatively high shunt capacitance than EELs, thus leading to distinct require- ments and consequently different design approaches for the LDD. The particular aspects of the VCSEL drivers become crucial, especially at multi-Gb/s data rates, because the packaging parasit- ics further aggravate the shunt-capacitance effect. In the present work, the current-pulse as well as the voltage- pulse modes of operation of VCSELs are investigated. Using an accurate nonlinear circuit VCSEL model [2] fed with the param- eters of a commercially available VCSEL, a consistent comparison between the current- and voltage-pulse configurations at high- speed operation is performed. In addition, circuits for driving VCSELs with either voltage- or current-pulses, designed utilizing a commercially available 0.35-m SiGe HBT technology, were used for realistic simulation of the VCSEL response in either configurations. In section 2, we briefly describe the VCSEL model and present the effects on VCSEL’s ac and transient response when driven by ideal current and voltage sources. Section 3 deals with the VCSEL response using realistic driver circuits either in current- or in voltage-pulse mode. Finally, the conclusions are presented in section 4. MICROWAVE AND OPTICAL TECHNOLOGY LETTERS / Vol. 44, No. 1, January 5 2005 41