Network coding for cooperative multi-user wireless communication systems

In the first chapter, Space Time Trellis Codes (STTCs), Space Time Block Codes (STBCs) and Sphere-Packing-Space-Time Block Codes (SP-STBC) are reviewed. These schemes belong to the specific family of Multi-Input Multi-Output (MIMO) systems designed for achieving a diversity gain. The performance of the SP-STBC scheme is compared to other coded conventional modulation systems, namely to that of STBC-Phase Shift Keying or Quadrature Amplitude Modulation (STBC-PSK/QAM) and to that of STTC-Phase Shift Keying or Quadrature Amplitude Modulation (STTC-PSK/QAM). The rest of this chapter reviews other preliminaries pertaining to the context of cooperative communications and network coding. In Chapter 2, an in-depth study of the capacity and outage probability of the Continuous-input Continuous-output Memoryless Channel (CCMC), Discrete-input Continuous-output Memoryless Channel (DCMC) and of Differential Discrete-input Continuous-output Memoryless Channel (DDCMC) is presented. The study also considers various propagation phenomena, namely the smallscale fading and the large-scale fading. The frame-length is also taken into consideration when calculating the achievable throughput and outage probability, which serve as useful benchmarks for our near-capacity coding schemes. Extrinsic Information Transfer (EXIT) charts are used for designing Irregular Convolutional Coded Unity Rate Coded M-ary Phase Shift Keying (IrCC-URCMPSK), Irregular Convolutional Coded Unity Rate Coded Differential M-ary Shift Keying (IrCCURC-DMPSK) and Irregular Convolutional Coded Unity Rate Coded Space Time Trellis Coded M-ary Phase Shift Keying (IrCC-URC-STTC-MPSK) schemes. In Chapter 3, a novel Distributed Concatenated IrCC-URC-STTC (DC-IrCC-URC-STTC) scheme is proposed for cooperative single-user systems relying on single-antenna aided relays, based on the studies conducted in Chapter 1 and Chapter 2. In this contribution, each coding arrangement of the entire DC-IrCC-URC-STTC scheme is designed for achieving decoding convergence to a vanishingly low Bit Error Ratio (BER) by employing non-binary EXIT-charts. Additionally, the EXIT charts are employed for calculating the most appropriate positions of the relays by ensuring that decoding convergence to a vanishingly low BER occurs at a similar Signal-to-Noise Ratio (SNR) both at the relays and at the destination. In Chapter 4, Multi-User Cooperative Communications is designed for supporting M users with the aid of near-capacity network coding. We first derive the upper and lower Frame Error Ratio (FER) performance bounds of cooperative multi-user communications systems using network coding. Then, we investigate Near-Capacity Multi-user Network-coding (NCMN) based systems using the IrCC-URC-MPSK scheme of Chapter 2. In parallel to the investigation of coherent NCMN systems, we also explored Near-capacity Non-coherent Cooperative Network-coding aided Multi-user (NNCNM) based systems using the IrCC-URC-DMPSK, which do not require channel estimation at the receiver’s side. This reduces the complexity imposed, albeit this is achieved at a 3 dB SNR-loss. Moreover, a new technique referred to as the Pragmatic Algebraic Linear Equation Method (PALEM) was proposed for exactly determining the number of information sources that may be recovered from the composite NCMN stream, which results in a more accurate evaluation of the attainable FER performance of the NCMN and NNCNM based systems. The design principles presented in this contribution can be extended to a vast range of NCMN and NNCNM based systems using arbitrary channel coding schemes. In Chapter 5, the NCMN and NNCNM based systems of Chapter 4 are generalised for introducing the Generalised NCMN (GNCMN) system, which has a multi-layer architecture and it is capable of operating in multiple modes. More specifically, the GNCMN system may operate upon employing either individually or in a combined fashion using a single Channel Coding (CC) layer plus two network coding layers, namely Network Coding 1 (NC1) and Network Coding 2 (NC2). Additionally, the GNCMN system is capable of simultaneously exploiting the advantages of all the modes available in each layer of the system as well as appropriately combining the advantageous modes across all the three layers. Finally, in Chapter 6, the summary of our findings are presented in order to facilitate our discussions on future research.