Optical Code Division Multiple Access Communication Networks

Optical fiber communication is a communication approach to transport information from one point to another using light as a carrier and optical fibers as transmission media. In ancient times, in order to speed up information transmission, people learned how to use optical signals, such as smoke signals, semaphores, etc., to communicate. However, the utility of these methods was very limited. In the early 1960s, American physicists invented the ruby laser[1], and the proposals for optical communication via dielectric waveguides or glass optical fibers to avoid degradation of the optical signal by the atmosphere were made almost simultaneously in 1966 by Kao and Hockham[2] and Werts[3]. Initially the optical fibers exhibited very high attenuation (i.e., 1000dB/km) and were therefore not competitive with the coaxial cables which they were to replace (i.e., 5 to 10dB/km). In 1970, the Corning Company in America manufactured a fiber-optic with attenuation of 17dB/km, and the optical fiber losses at 1310 nm wavelength were reduced to 0.3dB/km[4] in 1974. In 1977, the field trial of the first commercial use of the multimode fibers between two telephone offices in Chicago 7000 meters distant was made successfully[5].

In an OCDMA network, the transmission signal over a fiber-optic channel is formed by the superimposing of pseudorandom OCDMA signals encoded from multiple channels. The signal is broadcast to each node (subscriber) in the network and a receiver in each node decodes the signal. If the output of the decoder in this receiver is an autocorrelation, the node can detect the information sent to it from the aforementioned pseudorandom signals. Alternatively, if the output of the decoder is a cross-correlation function (no apparent peak value), then the node cannot receive the information. Therefore, in order to implement OCDMA communication and networking, address codes with sufficient performance are required. When a set of code parameters is chosen, a code can be constructed that has as many codewords (corresponding to the number of nodes in the network) as necessary and good enough auto- and cross-correlation so that accurate synchronization can be implemented and the interference (called multiple access interference, MAI) from other nodes can be suppressed effectively by decoding the signals. This requires that the address codes satisfy two conditions[1, 2] :

The number of users in a OCDMA network using one-dimensional (1-D) incoherent time-domain encoding is very limited. This is because the number of subscribers is proportional to the length of frequency-spreading, whereas the data rate of a single user is inversely proportional to the length of frequency-spreading. Because the length of frequency-spreading is constrained by the current state-of-the-art, the larger the number of subscribers in a network, the more serious is the multiple access interference (MAI), or called multiple user interference (MUI). The bit error rate of the system increases and the number of subscribers that can communicate with each other simultaneously is small in practice. Aimed at the shortcoming of one-dimensional incoherent OCDMA codes, two-dimensional (2-D) wavelength-hopping/time-spreading incoherent OCDMA was proposed. It has been shown[1] that 2-D wavelength-hopping/time-spreading (WH/TS) encoding incoherent OCDMA systems, in general, perform better than OCDMA+WDMA (wavelength-division multiple-access) hybrid systems[32]. Even under the best scenario that a central controller is deployed to uniformly distribute all available wavelength to simultaneous subscribers in the hybrid scheme, the 2-D WH/TS scheme performs better if the traffic load is medium. Furthermore, when the traffic is heavy, the performance is much better. Therefore, in comparison with 1-D incoherent OCDMA systems, the incoherent OCDMA systems using 2-D WH/TS encoding not only allow increased number of users and simultaneous communication subscribers in a network, improving the performance of the network, but also simplify the network control and management, reducing processing time and alleviating the complexity and cost of hardware implementation.

The typical network architecture for OCDMA with broadcast star is shown in Fig. 4.1. It can be seen that one of the key issues to implement OCDMA networking and communication is how to encode and decode the user’s data such that the optical channel can be shared, that is, we need to develop the practical encoding and decoding techniques that can be exploited to generate and recognize appropriate code sequences reliably. Therefore, The OCDMA encoders and decoders are the key components to implement OCDMA systems. In order to actualize the data communications among multiple users based on OCDMA communication technology, one unique codeword-waveform is assigned to each subscriber in an OCDMA network, which is chosen from specific OCDMA address codes, and therefore, different users employ different address codeword- waveforms. In an OCDMA network using on-off keying pattern, the user’s data is transmitted by each information bit “1” which is encoded into desired address codeword. However, the transmitter does not produce any optical pulses when the information bit “0” is sent. Figure 4.2 gives the schematic diagrams of waveforms employed to transmit data of three subscribers in an OCDMA network.

From the previous chapters we have learned that a typical OCDMA communication system can be shown as in Fig. 5.1. We assume that there are totally M subscribers implementing full duplex communication in an OCDMA network and N users are active and share the common channels at the same time. (In a practical network, all subscribers connected to a network are not always activated, especially in a subscriber access network and, as a matter of fact, the number of subscribers activated at the same time accounts for about 10% of the total number[1].) We suppose that if the j ^th subscriber wants to send data information to the k ^th user, the address code for receiver k is impressed upon the data by the encoder at the j ^th node. One of the primary goals of OCDMA is to extract data with the desired optical pulse sequence in the presence of all other users’ optical pulse sequences.

As various services in telecommunication network increase rapidly, especially the explosively mushrooming of Internet protocol (IP) services, the global telecommunication network infrastructure is required to be upgraded in order to meet the present demands and the future services for the network. The ultrahigh- speed photonic network will play a crucial role in the upgrade of network. Photonic network can provide large capacity, and at the same time, it has the functions of implementing dynamic reconfiguration of network, flexible management of bandwidth, efficient utilization of network resources, securities of network infrastructure and information transmission, etc., so that it can adapt well to the dynamic service variations in the network and the specific demands from different subscribers. These are implemented through the network architectures and protocols, the manners of signal multiplexing and switching, and the formats of signal encoding and decoding. At present, the multiplexing and switching technologies widely used are the electronic time division and optical wavelength division technologies.