We propose and demonstrate a novel WDM-CAP-PON based on optical single-side band (OSSB) multi-level multi-band carrier-less amplitude and phase modulation (MM-CAP). To enable high-speed transmission with simplified optical network unit (ONU)-side digital signal processing, 4-level 5 sub-bands CAP-16 is used here, which is generated by the digital to analogue converter (DAC). Optical single-side band (OSSB) technology is applied to extend the transmission distance against the spectrum fading effect. As a proof of concept, the experiment successfully demonstrates 11 WDM channels, 55 sub-bands, for 55 users with 9.3-Gb/s per user (after removing 7% overhead for forward error correction (FEC)) in the downstream over 40-km SMF.
©2013 Optical Society of America
The wide variety of data and services drive the demand of higher bit-rate per user in access networks, which requires systems that can support higher capacity. To meet this capacity needs, access networks are moving from the classic spectral inefficient non-return to zero (NRZ) time-division multiplexing (TDM), to wavelength-division multiplexing (WDM) and more advanced modulation formats [1–15]. However, the complexity and cost raise have to be kept to the minimum in order to make the system feasible for access networks [8,9]. Different techniques have been used for the advanced modulation in the short range communication, such as the quadrature amplitude modulation (QAM) or the phase shift keying (PSK) based the sub-carrier modulation (SCM) [3–6], the orthogonal frequency division multiplexing (OFDM) [7,8], and the carrier-less amplitude/phase modulation (CAP) [9–15]. Overall, the CAP architecture based on the intensity modulation and direct detection has been demonstrated to be less complex and with good performance, which allows relatively high data using optical and electrical components of limited bandwidth [9–15]. No electrical complex-to-real-value conversion, complex mixer, radio frequency (RF) source or optical IQ modulator are required for CAP, compared with QAM-SCM [3–6] and OFDM [7–11]. Higher spectrum efficiency can be obtained for CAP compared with ordinary QAM-SCM in previous works [3–6]. Moreover, high-speed multilevel CAP implementation has been demonstrated by using transversal filters developed for equalization of NRZ data  or direct modulated vertical cavity surface emitting lasers (VCSELs) for access network [11,12]. Especially in  and , they have demonstrated the feasibility of CAP-based access network. However, the data rate is limited (only 1.25-Gb/s for  and less than 1-Gb/s for ) and the multi-band CAP is not investigated for multiple users. Most recently, multi-level multi-band CAP (MM-CAP) has been proposed to extend the bandwidth of each channel for high speed short range data transmission . However, the transmission distance for high speed CAP is limited due to the chromatic dispersion (CD) and fiber loss. The dynamic sub-band operation for WDM-PON with multiple users or multi-services is also not discussed.
From the view of sub-bands in each MM-CAP channel, the shaping filter pair in transmitter and matched filter pair in receiver are related and matched, which gives a possibility for application in multiple users or multiple services access networks. Different data in different sub-band can be only recovered by the related matched-filter at the receiver side, which can be assigned to different users. On the other hand, optical single-side band (OSSB) technology is a good solution against spectrum fading effect as analyzed in .
In this paper, we propose and experimentally demonstrate a novel WDM-CAP-PON based on OSSB multi-level multi-band CAP modulation (MM-CAP) for multiple users access network. 4-level 5 sub-bands CAP-16 is used here, which is generated by the digital to analogue converter (DAC). As a proof of concept, the experiment successfully demonstrates 11 WDM channels, 55 sub-bands, for 55 users with 9.3-Gb/s per user (after removing 7% overhead for forward error correction (FEC)) in the downstream over 40-km SMF. To the best of our knowledge,this is the first demonstration for MM-CAP used for WDM-CAP-PON with such a high speed.
2. The principle of WDM-CAP-PON
Figure 1 shows the principle of downstream for WDM-CAP-PON based on MM-CAP. In the central office, each optical line terminal (OLT) transmitter (Tx) at the ith wavelength λi carries the MM-CAP signals for multiple users. CAP is a multilevel and multi-dimensional modulation format proposed by Bell Labs for short range communication [9–16], which similar to QAM signal, but does not require a RF carrier source. Different from QAM modulation, CAP does not use a sinusoidal carrier to generate two orthogonal components I and Q. The two dimensional CAP can be generated by using two orthogonal filters as shown in Fig. 1 as the filter pair. For multi-band CAP, more than one filter pairs are used, which are located on different frequencies.
At the optical network unit (ONU) side each user can recover the data in each sub-band using a pair of matched filters corresponding to the pulse shaping filters in the transmitter. Only one sub-band signal can be recovered by one matched filter pair. In this way, for each channel, N sub-bands can be assigned to N users without any interference. For WDM-CAP-PON, K wavelengths with N sub-bands can totally be assigned to K × N users. On the other hand, when considering long distance transmission, OSSB can be used against the spectrum fading effect caused by CD and direct detection and enables linear equalizations as in . Here, only intensity modulation and direct detection are needed.
In this paper, as a proof of concept, we study the structure for the downstream with WDM MM-CAP. It is worth noting that our scheme with WDM-CAP-PON for downstream is compatible with the other PON structures for upstream. There are mainly two schemes for upstream. Since the bandwidth requirement for upstream is much less than the downstream, OOK can be directly used as proposed in  where downstream is OSSB OFDM and the upstream is OOK modulated on the same carrier with lower data rate. Alternative schemes is using different wavelengths such as , which using another wavelength for upstream with same modulation formats.
Figure 2 shows the schematic diagram of transmitter and receiver based on multi-band CAP for one data stream. The original bit sequence of a data stream n is first mapped into complex symbols of m-QAM (m is the order of QAM), and then the mapped symbols are up-sampled to match the sample rate of shaping filters. The sample rate of shaping filters is determined by the data baud rate and DAC sample rate. For CAP generation, the in-phase and quadrature components of the up-sampled sequence are separated and sent into the digital shaping filters respectively. The outputs of the filters are subtracted. For multiple users, the transmitter data for each data steam after the orthogonal filter pairs can be added together before the digital-to-analog (D/A) converter. At the ONU side, the received signal after analog-to-digital (A/D) converter is fed into two different matched filters to separate the in-phase and quadrature components. After down-sampling, an equalizer is employed for the complex signal and a decoder is utilized to obtain the original bit sequence. Different and unique matched filter pairs are used for multiple users.
The orthogonal and matched filter pairs fIn(t), fQn(t), mfIn(t) and mfQn(t) are the corresponding shaping filters and form a so-called Hilbert pair in transmitter and receiver, as described in . Different from the single band CAP in , multi-band CAP uses multi-filter pairs which located in different frequency sub-bands . For each sub-band, the two orthogonal filters are constructed by multiplying a square root raised cosine pulse with a sine and cosine function respectively, as shown in Eq. (1) and (2). For MM-CAP with N sub-bands, the orthogonal filter pair of the nth (1~N) sub-band in time domain can be expressed as
At the receiver, generally, we have the matched filters with relations as mfIn(t) = fIn(-t), and mfQn(t) = fQn(-t).
In this way, for the nth sub-band, the I and Q data after matched filter pair can be expressed as
Take 4 sub-bands CAP filters as an example, Fig. 3 shows the time impulse response and frequency response of these filter pairs of fIn(t), fQn(t), for different sub-bands. Here, the up-sample rate is 10, and the roll off fact is 0.2. We can see that, the filter pair for different sub-band has a different waveform in time domain. In frequency domain, these filter pairs are located in different sub-bands, which can be assigned to different users.
3. Experimental setup and results
As a proof of concept, the experimental setup of 11x5x9.3-Gb/s WDM-CAP-PON downstream with 55 users over 40-km SMF is shown in Fig. 4. 11 carriers as the source of 11 channels are generated by an optical comb generator . Odd and even channels are separated by a two port 28GHz-gird WSS and then intensity-modulated by two MZMs. The 5 bands MM-CAP-16 signals are generated by the high-speed DAC, which works at 30GSa/s. 5 different data sequences are generated and first mapped to the 4 level 16-QAM I and Q signals with 10x211 symbols. Then, the ten sets of 4 level sequences are up-sampled to 12 Sa/symbol and filtered by 5 pairs of shaping MM-CAP filters to generate MM-CAP signal with 5 sub-bands. The filters are finite impulse response (FIR) filters with length of 8 symbols each. The roll-off coefficient is 0.2 and the excess bandwidth is set to 15%. The 50-Gb/s MM-CAP-16 signal is generated by simply adding the outputs of the 5 filter pairs. It is worth noting that, the frequency response of output of the DAC is non-flat, with a 3-dB bandwidth less than 13-GHz. Inset (a) and (b) in Fig. 4 show the spectrum of the output of DAC with and without frequency domain pre-equalization, respectively. We can see that, by simply adding weights of each pairs of filters, we can get a more identical frequency response for each sub-band CAP signal. The odd and even channels after MZMs are then filtered to produce the OSSB signals and combined together by a 28GHz-spaced WSS with 14GHz frequency offset and 20GHz bandwidth. The optical spectrum of single channel and 11 channels WDM OSSB MM-CAP signals are shown as (a) and (b) in Fig. 5. The 11 WDM channels signals are transmitted over 40-km SMF with loss of 10-dB.
At the receiver, the channel is first filtered out by a tunable filter (TOF) with 0.3-nm bandwidth. After direct detection, the signals are sampled by a digital scope at 50-GSa/s for offline processing. A 0.9nm TOF is used before PD front-end to remove excess ASE noise. The signals are first resampled to 30GSa/s and then demodulated by the assigned matched-filters pair. Each sub-band CAP is processed by different matched-filters for data recovery. After that, the signals are down-sampled to 2Sa/Symbol before the linear equalizations, including CMMA and phase recovery. Each sub-band is with 2.5GHz bandwidth and 10-Gb/s data rate. Figure 6 shows the back to back (BTB) bit error ratio (BER) as a function of he received optical power for each sub-band of channel 4. We can see that, with frequency domain pre-equalization by adding weights on each sub-band, identical BER performance for sub-band 1~3 and negligible power penalty for sub-band 4 and 5 at BER below the FEC limit of 3.8x10−3 for 7% hard-decision FEC are obtained. We also measure the BER of sub-band 5 without pre-equalization, which shows very poor performance.
Figures 7(a) and 7(b) show the spectrum of received MM-CAP signal after 40-km SMF for double sideband (DSB) and OSSB signals, respectively. We can see that, without OSSB, sub-band 3 and 4 are destroyed due to the power fading effect caused by CD and direct detection. Almost 40% data transmission is cut-off. However, the power fading can be avoid by OSSB as shown in Fig. 7(b).
Figure 8(a) shows the measured BER of each sub-band in channel 4 versus received optical power after 40-km SMF transmission. Inset (i) and (ii) show the constellation of sub-band 1 and 5 at the received power of −15dBm. Sub-bands 1 and 2 have identical performance with about 4.5-dB power penalty compared with BTB case. However, sub-bands 3~5 have larger power penalty about 5.5-dB, 6.5-dB and 6 dB, respectively. It is due to the residual power fading effect under the imperfect OSSB filtering. Figure 8(b) shows the required optical power for the total 55 sub-bands in 11 channels at the BER of 3.8x10−3 for the 7% hard-decision FEC limit after 40-km SMF transmission. The downstream data rate is 50-Gb/s/ch, and totally 11x5x9.3Gb/s for 55 users after removing the FEC overhead. Inset shows the optical spectrum of the received 11 channels. Considering the typical WDM-PON, the input power of each channel is around 10dBm, the fiber loss can be optimized to be about 8dB, and the insertion loss of the commercial wavelength demultiplexer (such as AWG) is less than 4dB. In our case, the required received power of the worst WDM channel for MM-CAP is about −12-dBm. Thus, there is a 10dB power margin for each wavelength. Since only 5 users are considered here, this power margin is enough for each user. On the other hand, the wavelength carrier spacing and user number per wavelength should be optimized and designed for practice use. However, if more users or longer distance are required, higher sensitivity receiver (such as coherent receiver) or optical amplifiers are needed in the local exchanger or remote note as described in , which will be carried out in future investigation.
We propose and experimentally demonstrate a novel WDM-CAP-PON based on OSSB MM-CAP for multiple users access network. 4-level 5 sub-bands CAP-16 is used here, which is generated by the DAC with 30-GSa/s. As a proof of concept, the experiment successfully demonstrates 11 channels, 55 sub-bands, for 55 users with net data rate of 9.3-Gb/s per user in the downstream over 40-km SMF. To the best of our knowledge, this is the first demonstration for MM-CAP used for WDM-CAP-PON with such a high speed.
This work was partially supported by the NHTRDP (973 Program) of China (Grant No. 2010CB328300), and NNSF of China (No. 61177071, No.61250018), International Cooperation Program of Shanghai Science and Technology Association (12510705600), and the China Scholarship Council (201206100076).
References and links
1. D. Breuer, F. Geilhardt, R. Hülsermann, M. Kind, C. Lange, T. Monath, and E. Weis, “Opportunities for next-generation optical access,” IEEE Commun. Mag. 49(2), s16– s24 (2011). [CrossRef]
2. G. K. Chang, A. Chowdhury, Z. Jia, H. C. Chien, M. F. Huang, J. Yu, and G. Ellinas, “Key technologies of WDM-PON for future converged optical broadband access networks,” J. Opt. Commun. Netw. 1(4), C35–C50 (2009). [CrossRef]
3. J. M. Kang and S. K. Han, “A novel hybrid WDM/SCM-PON sharing wavelength for up- and down-link using reflective semiconductor optical amplifier,” IEEE Photon. Technol. Lett. 18(3), 502–504 (2006). [CrossRef]
4. A. Wiberg, B. Olsson, and P. Andrekson, “Single cycle subcarrier modulation,” in Proc. OFC2009, paper OTuE1. [CrossRef]
6. J. Karout, E. Agrell, K. Szczerba, and M. Karlsson, “Optimizing constellations for single-subcarrier intensity-modulated optical systems,” IEEE Trans. Inf. Theory 58(7), 4645–4659 (2012). [CrossRef]
7. B. Liu, X. Xin, L. Zhang, J. Yu, Q. Zhang, and C. Yu, “A WDM-OFDM-PON architecture with centralized lightwave and PolSK-modulated multicast overlay,” Opt. Express 18(3), 2137–2143 (2010). [CrossRef] [PubMed]
8. N. Cvijetic, M. Cvijetic, M. F. Huang, E. Ip, Y. K. Huang, and T. Wang, “Terabit optical access networks based on WDM-OFDMA-PON,” J. Lightwave Technol. 30(4), 493–503 (2012). [CrossRef]
9. J. L. Wei, D. G. Cunningham, R. V. Penty, and I. H. White, “Study of 100 Gigabit Ethernet using carrierless amplitude/phase modulation and optical OFDM,” J. Lightwave Technol. 31(9), 1367–1373 (2013). [CrossRef]
10. J. D. Ingham, R. Penty, I. White, and D. Cunningham, “40 Gb/s carrierless amplitude and phase modulation for low-cost optical data communication links,” in Proc. OFC2011, paper OThZ3.
11. R. Rodes, M. Wieckowski, T. T. Pham, J. B. Jensen, J. Turkiewicz, J. Siuzdak, and I. T. Monroy, “Carrierless amplitude phase modulation of VCSEL with 4 bit/s/Hz spectral efficiency for use in WDM-PON,” Opt. Express 19(27), 26551–26556 (2011). [CrossRef] [PubMed]
12. M. B. Othman, X. Zhang, L. Deng, M. Wieckowski, J. Jensen, and I. T. Monroy, “Experimental investigations of 3D/4D-CAP modulation with DM-VCSELs,” IEEE Photon. Technol. Lett. 24(22), 2009–2012 (2012). [CrossRef]
13. M. I. Olmedo, T. Zuo, J. B. Jensen, Q. Zhong, X. Xu, and I. T. Monroy, “Towards 400GBASE 4-lane solution using direct detection of multiCAP signal in 14 GHz bandwidth per lane,” in Proc. of OFC2013, paper PDP5C.10.
14. J. Wei, L. Geng, D. G. Cunningham, R. V. Penty, and I. White, “100 Gigabit Ethernet transmission enabled by carrierless amplitude and phase modulation using QAM receivers,” in Proc. of OFC2013, paper OW4A.5. [CrossRef]
15. L. Tao, Y. Wang, Y. Gao, A. P. Lau, N. Chi, and C. Lu, “Experimental demonstration of 10 Gb/s multi-level carrier-less amplitude and phase modulation for short range optical communication systems,” Opt. Express 21(5), 6459–6465 (2013). [CrossRef] [PubMed]
16. G. H. Im, D. D. Harman, G. Huang, A. V. Mandzik, M. H. Nguyen, and J. J. Werner, “51.84 Mb/s 16-CAP ATM LAN standard,” IEEE J. Sel. Areas Comm. 13(4), 620–632 (1995). [CrossRef]
17. J. Yu, M. F. Huang, D. Qian, L. Chen, and G. K. Chang, “Centralized lightwave WDM-PON employing 16-QAM intensity modulated OFDM downstream and OOK modulated upstream signals,” IEEE Photon. Technol. Lett. 20(18), 1545–1547 (2008). [CrossRef]
18. J. Zhang, J. Yu, L. Tao, Y. Fang, Y. Wang, Y. Shao, and N. Chi, “Generation of coherent and frequency-lock optical subcarriers by cascading phase modulators driven by sinusoidal sources,” J. Lightwave Technol. 30(24), 3911–3917 (2012). [CrossRef]