This dissertation focuses on performance analysis and power allocation for a BICM-coded cooperative relaying network. Since the performance analysis of AF has been discussed in [66][67], the performance analyses in this dissertation mainly focus on S-DF. For the perfor-mance analysis part, unlike the existing works which considered uncoded S-DF with sym-bol-by-symbol forwarding strategy, this dissertation is the first work that studies the BICM-coded cooperative relaying network with packet-by-packet forwarding strategy. The target is to provide the analysis of BER performance at the destination and the diversity order of the network for both S-DF/RT and S-DF/Idle over both fast-fading and block-fading Nakagami-m channels. For the power allocation part, a comprehensively investigation is proposed for not only the AF, S-DF/RT, S-DF/Idle and S-DF/AF relaying (based on a packet-by-packet forwarding) but also DRF relays.
Based on the knowledge of perfect CSI, the target is to allocate power to minimize the BER at the destination, while taking into account the probability of decoding failure at relays. To avoid the cumbersome (if not impossible) evaluation of the exact BER and an inefficient exhaustive search of the optimal power, simplified cost functions which can be optimized efficiently through exist-ing algorithms are proposed for individual modes. The objective is to determine the transmit power allocation that minimizes the BER at the destination. Outline and contributions of this dis-sertation are as follows.
Chapter 2 first describes the network model and the channel statistics. The BICM modulation at the source and the corresponding decoding at relays at the phase-I are given. Then, the model of signals forwarded at phase-II are introduced for AF, S-DF/RT, S-DF/Idle and S-DF/AF, fol-lowing by a general formulation that is used to describe the decoding at the destination and will be further employed in the derivations of all the following chapters.
Chapter 3 investigates the BER performance and diversity order for S-DF/RT and S-DF/Idle over fast-fading Nakagami-m channels with a packet-by-packet forwarding strategy. Given a set
of active relays, this Chapter first derives the approximation of BER at the destination by extend-ing the expurgated bound proposed in [58]. But, unlike [58] which employed the Chernoff bound to obtain the expurgated bound, a close-form evaluation is proposed. Such an evaluation can be degenerated to calculate the error rate at relays so that the overall BER at the destination is ob-tained. To find out the diversity of the network, the asymptotic upper and lower bounds are de-rived by considering only the worst case error events between coded sequences and signal point pairs with the shortest Euclidean distance in the constellation. By showing that both bounds achieve the same diversity order, the diversity order of the network is obtained. Numerical results show that our approximations are rather accurate (within a 0.4dB gap to the true BER) for differ-ent network setups.
Chapter 4 studies the BER performance and diversity order for block-fading Nakagami-m channels. For the BER analysis, the BER for a given active relay set can be obtained by first ob-taining the BER in AWGN channels before averaging it over channel realizations. Unfortunately, a direct integration may lead to a non-trivial gap to the exact performance due to the severe loss at low SNR regions in AWGN case. As a result, the BER in AWGN is modified and the Monte Carlo method is used for the channel averaging. The diversity of the cooperative BICM network is also derived for both S-DF/RT and S-DF/Idle. Numerical results are given to show that our ap-proximations are rather accurate for different network setups. An example is also provided to ver-ify our proof of the diversity order.
Chapter 5 investigates the power allocation of the cooperative BICM relaying network. Four relaying schemes are considered: AF, S-DF/RT, S-DF/Idle and S-DF/AF with the general formu-lation given in Chapter 2. For AF, by simplifying the union bound, an approximate BER is de-rived and shown to be monotonically decreasing with an equivalent channel gain so that the power allocation method named PA-EC is proposed which takes the equivalent channel as the cost function. For the S-DF relaying modes, two power allocation methods are proposed:
PA-ABER based on approximate BER and PA-MGEC based on minimum generalized equivalent
channel. In PA-ABER, the approximate BER is shown to be a convex function for S-DF/RT, S-DF/Idle and S-DF/AF. Therefore, gradient method can be adopted to find the solution. Then PA-MGEC transforms PA-ABER to a max-min problem, which can be optimized with even low-er complexity. Examples are given to demonstrate how powlow-er is allocated for PA-EC and PA-MGEC on the AF and S-DF relaying modes. Simulation results confirm that our proposed methods have the ability to properly allocate power according to the SNR value and the channel realizations. The proposed methods outperform the equal gain power allocation (PA-EG) with large margins for different network setups.
Chapter 6 considers S-DRF relays which are allowed to change the mappers before forward-ing so as to obtain an addition remappforward-ing gain. The proposed methods, PA-ABER and PA-MGEC, in Chapter 5 are extended in this chapter for the remapping case. Examples are provided to demonstrate how power is allocated, and numerical results confirm that the proposed method outperform PA-EG with large margins for different network setups.
Finally, Chapter 7 concludes the dissertation and discusses some possible extensions and fu-ture research topics.