Design and performance analysis of enhanced network coded cooperative communication systems

A wireless communication system consists of multiple wireless nodes that can move around and communicate with each other. Fading is one of major degrading factors that can limit the performance of wireless communication systems. In order to achieve reliable communication in fading environments, channel coding and diversity techniques were proposed and implemented. Multiple input multiple output (MIMO) is one of the diversity techniques which uses multiple transmission and reception nodes per terminal to obtain diversity gains for reduced errors in the transmission of data. But cooperative diversity techniques can realize this purpose without installing multiple transmission antennas per terminal. The broadcasting nature of wireless medium allows neighboring relay nodes to cooperate in communication by forwarding information from source nodes to destination nodes for fading mitigation. Therefore, cooperative communication system utilizes multiple communication routes created by relay nodes and exploits the inherent spatial diversity of the channel for information transmission. While classical single hop communication systems use direct transmission in which a receiver recovers the information using the direct signal only and regarding the reflected signal as interference, the cooperative communication systems consider the other signal as contribution. Therefore, cooperative diversity retrieves information from the combination of two or more signals. In other words, cooperative diversity is a virtual antenna diversity technique that uses distributed antennas belonging to each node in a wireless network for communication. Cooperative diversity is particularly considered useful in wireless ad hoc and sensor networks, where power/bandwidth/size restrictions of the mobile nodes may prevent the use of other diversity techniques to combat channel fading. In all scenarios, cooperation among mobile users of a wireless system has the potential to provide an increased capacity in comparison with the systems without using cooperation. However, in most cases, cooperative communication attains this improvement in error performance by sacrificing the throughput of wireless network. In such cases, network coding technique can substantially improve the data rate of cooperative wireless networks by intelligently combining the forwarded packets of information at the relay nodes. Network coding was originally proposed for lossless communication systems to increase the throughput of routing networks. But the integration of network coding with cooperative wireless networks has the potential to ensure more efficient usage of resources with improved error performance in fading environments. Network coding can ensure the capability to address the ever increasing number of users and devices in wireless networks, in beyond 5G standards. The broadcast nature of wireless systems allows the joint implementation of cooperative communication and network coding, exploiting the benefits of both techniques. In this context, several types of network coding have been discussed in literature in the recent years but each type involves the concept of combining and transmitting the functions of information packets at the relay nodes. In this thesis, random linear network coding, complex field network coding and exclusive-OR (XOR) network coding are simulated and analyzed in different perspectives. In the first part of this thesis, the performance of random linear network coding systems is investigated for Rayleigh and Rician fading channels in terms of decoding failure probabilities at destination nodes. In random linear network coding systems, the information bits at the intermediate nodes are received as packet vectors and the intermediate nodes encode the information data by linearly combining the received packets, with coefficients randomly extracted from the Galois field of a particular size, with equal probability. The setup considered in this section consists of multiple source, relay and destination nodes, with no direct links between sources and destinations. The communication channel outage and hence the packet loss in the given system occurs due to fading. The simulation results for the supposed system model show that the packet decoding failure probabilities are not only determined by the fading environment but also by the coefficients used in the network coding at the relay nodes. In fading channels, decoding failure probabilities can be reduced by increasing the size of the Galois field that contains random coefficients for linear combination of packets at relay nodes. In the second part of this thesis, the bit error rate performance of complex field network coding is analyzed with orthogonal frequency division multiplexing (OFDM) for frequency selective Rayleigh fading channels. In previous literatures, this analysis is performed for flat fading channels. The system model used in the performance evaluation contains multiple source nodes, single or multiple relay nodes and a single destination node. Both amplify and forward as well as decode and forward types of relays are assumed to obtain error performance results and for multi-relay system, relay selection is also taken into account for both amplify and forward and decode and forward relay types. Convolutional channel codes are also optionally integrated in the given system to boost the system performance and hard decision Viterbi decoding is used to decode source bits. In addition to this, asymmetric link environments, with relay nodes at different distances from sources and destination, are also considered. Multiuser detection rules with OFDM are employed at destination node and/or relay nodes to retrieve each source bits at the same time, for the above mentioned scenarios. The simulation results indicate that amplify and forward relay nodes outperform decode and forward relay nodes in all cases but are not very suitable from the implementation perspective since they require the storage of analog waveform at relays. Moreover, a considerable improvement is error performance can be observed when relay selection is provided in the system model with both amplify and forward and decode and forward types of relays. Finally, in asymmetric network, better error performance can be achieved by placing relay nodes closer to the source nodes. The third part of this thesis proposes the channel coded complex field network coding with OFDM, combining the mutual benefits of all the techniques involved and the performance evaluation is made in terms of bit error rate. This work is the extension of the Convolutional channel coded complex field network coding with OFDM and hard decision Viterbi decoding, presented in the second part. The system model is same as the second part with slow and frequency selective Rayleigh fading channel. Again, both amplify and forward and decode and forward types of relays are considered in this context and relay selection is employed in scenarios with multiple relays. However, this part uses soft decision Viterbi and Max-Log-MAP decoding techniques for obtaining coded source bits and therefore, in this case, the multiuser detection rules provide log likelihood ratios for coded source bits. Moreover, the performance analysis is provided for Convolutional codes with different code rates and number of states in the trellis as well as for asymmetric link environments. From simulation results, it is seen that both soft decision Viterbi and Max-Log-MAP channel decoding techniques provide almost same bit error performance in the given system model. Amplify and forward performs better than decode and forward in all cases and relay selection techniques results in a noticeable performance gain. Convolutional codes with lower code rate and more number of trellis states can enhance the system performance considerably and asymmetric links also serve the similar purpose of performance improvement. The final part of thesis is related to an enhanced version of XOR network coding scheme for two-way relay networks, combined with OFDM. Compared to other conventional strategies for two-way relay networks, this strategy performs Galois field addition or bit-wise XOR coding of multiple information packets at sources. The XOR coded packet is transmitted to the relay nodes along with the information packets which facilitates redundancy to acquire transmit diversity gain for improving error performance of the two-way relay network. The system model consists of two source and different number of relay nodes with Convolutional channel encoding in the system to obtain desired error performance. Relay selection is also provided when there are multiple relays in the system and this provision not only improves the error performance but also gives better throughput. The results show that the redundancy provided at the transmission end of two-way relay network enhances bit error performance of the system.