Study On Cooperative Communication Systems Based On Superposition Coding And Network Coding

With the rapid development of wireless communications, cooperative relaying has been shown to be a promising solution to boost the communications range of wireless networks and to exploit the diversity by allowing user terminals to share their antennas and to transmit cooperatively. It is one of the most effective ways to mitigate the fading effect of wireless channels. Today, cooperative relaying has become a set of wireless communications technique combining network coding (NC), opportunistic relaying, cooperative sense, and etc. At the same time, superposition coding (SC) has also been considered as an effective method to improve the throughput of wireless network by using the layered coding and modulation. With SC, the multiple signals are created, supposed in modulation domain, and broadcasted by sources. Different signals are modulated on different constellations. At the receiver, the detection strategy for the SC-ed signals is the successive inference cancellation (SIC). Therefore, the combination of the SC, NC, and cooperative relaying, SC-relaying, not only preserves the benefits inherent in the two techniques, but also possesses a few more remarkable advantages. This dissertation contributes to the investigation on the outage performance and application of two-layer SC-relaying based on the superposition coding and network coding. The main contributions are summarized as follows.1) With the high signal-to-noise ratio (SNR) approximation, we derive the closed-form solutions to the outage probabilities of the basic signal and additional signal for the two-layer SC-relaying systems by using the strict mathematical manipulations. Based on the results, the impact of power and channel allocation fractions on system performance is carried out. Then, a set of the optimal solutions to system parameters are obtained by employing difference evolution arithmetic aiming to maximize the throughput. The comparative numerical analyses show that the approximation results are feasible and correct.2) Based on the above SC-relaying model, we investigate the effect of the relay-source quantized channel state feedback (RS-QCSF) on the diversity-multiplexing tradeoff (DMT). For the RS-QCSF SC-relaying systems, we present a novel two-layer multi-level iteration rule for constructing the RS-QCSF index. With the proposed RS-QCSF rule, we obtain the update methods of the power SNR exponents for the basic signal and additional signal, and the initial value range of the power SNR exponent of the basic signal by studying the DMT of the basic signal. At the same, to observe the interaction between the multiplexing gains of the basic signal and additional signal, we discuss the DMT of the additional signal, through which we obtain the constraint relationship of the multiplexing gains. 3) To improve the performance of the two-way asymmetric SC-relaying channels, we present a novel network coding scheme by jointing the superposition coding and denoise-and-forward network coding (SC-DNF-NC). The SC-DNF-NC is presented by the comparison investigation of the two traditional NC schemes for asymmetric relaying channels, which are partitioning segment DNF-NC (P-DNF-NC) and padding zero DNF-NC (Z-DNF-NC). For the three schemes, we analyze the rate regions and the normalized throughputs. The numerical results show that the presented SC-DNF-NC scheme outperforms the other two schemes by allocating the reasonable power factor between the basic signal and addition signal. Moreover, in large SNR regime, the normalized throughputs of the three asymmetric schemes are close to the one of the symmetric DNF-NC scheme who has the optimal throughput, and superior to the one of amplify-and-forward NC.4) We present the network coding cognitive receiver as well as the corresponding rate region by jointing superposition coding and XOR network coding. Then, based on the linear four-node wireless network, we present three different jointing superposition coding and XOR schemes as well as the corresponding rate regions, which are superposition network coding (SC-NC), jointing superposition coding and XOR network coding (SC-XOR-NC), and jointing superposition coding and XOR network coding with cognitive receiver (CR-SC-XOR-NC). The comparison analysis shows that the CR-SC-XOR-NC scheme outperforms the other two schemes.