Measurement Methods Based On Self-mixing Effect Of Semiconductor Laser Linewidth Width Factor

Linewidth Enhancement Factor(LEF)α is an important parameter that describes many characteristics of semiconductor lasers (SL), such as spectral linewidth, lineshapes, frequency modulation/amplitude modulation (FM/AM), optical injection-locking range. Consequently, it is very significant to know the exact value of LEF for applications of SLs.At first, the paper overviews and compares the different measurement methods on LEF, and then an measurement scheme based on the optical feedback self-mixing effects is chosen for estimating LEF in SLs by this paper. The chosen scheme belongs to optical feedback method. In contrast to other methods, this one has very simple set-up, fewer measured parameters. Then the basic theory model of self-mixing effects which is based on the Long and Kobayashi equations and the three-mirror cavity approach respectively is described. The self-mixing signals modulated by injection current and external cavity of SL are simulated and analyzed. The factors affecting on the waveforms such as LEF and the optical feedback factor C as well as the specific characteristics of self-mixing signals are found, which are helpful for studying estimation algorithm of LEF.Secondly, a new algorithm based on the nonlinear least squares method for estimating LEF and factor C in SLs is proposed. This method is suitable for a weak optical feedback case in SLs. Computer simulation and experimental data processing show that the proposed algorithm can provide a good accuracy even in noisy environment. The parameter values gained from the algorithm are good agreement with ones reported in the published literature. Addtionally, how to select reasonably intinal values is presented. The guides for selecting intinal values are summarized, which are very useful for applying the proposed algorithm to estimate LEF and C.Thirdly, zero-maximum-minimum-dots method and newton interpolation algorithm are also studied in this paper. The effectiveness of these two methods have been verified by computer simulation and experiment. The advantage of zero-maximum-minimum-dots method is only measuring simply and needing fewer measuring dots, however, its accuracy is affected strongly by noise. Therefore, newton interpolation algorithm is employed for increasing the accuracy. A better estimation result can be achieved.At last, conclusions are drawn from the whole paper and the further research issues are pointed out.