The main theme of this thesis is the design and analysis of stochastic wireless networks, taking into account various factors including multiple antennas, different levels of feedback, and practical channel estimation (CE) schemes. This is facilitated by rigorous analytical and numerical analysis, and will involve the derivation of various performance expressions, such as the outage probability, and the transmission capacity. Several advanced mathematical tools, such as stochastic geometry and random matrix theory, are utilized for the derivations. Based on this analysis, new fundamental engineering design insights are obtained.

Firstly, multiple-antenna systems in wireless ad hoc networks are investigated, where perfect knowledge of the desired channel at each receiver is assum...[ Read more ]

The main theme of this thesis is the design and analysis of stochastic wireless networks, taking into account various factors including multiple antennas, different levels of feedback, and practical channel estimation (CE) schemes. This is facilitated by rigorous analytical and numerical analysis, and will involve the derivation of various performance expressions, such as the outage probability, and the transmission capacity. Several advanced mathematical tools, such as stochastic geometry and random matrix theory, are utilized for the derivations. Based on this analysis, new fundamental engineering design insights are obtained.

Firstly, multiple-antenna systems in wireless ad hoc networks are investigated, where perfect knowledge of the desired channel at each receiver is assumed. A general framework for the analysis of a broad class of point-to-point linear multiple input multiple output (MIMO) transmission schemes in decentralized wireless ad hoc networks is first developed. The results show that the transmission capacity scales linearly (but no faster) with the number of antennas under a set of mild conditions. The optimal number of data streams for maximizing the transmission capacity and the throughput in various asymptotic regimes is also characterized. To make the discussion concrete, the general framework is then applied to investigate three popular MIMO schemes, each requiring different levels of feedback. This reveals that significant performance gains are achieved by utilizing feedback under a range of network conditions.

Second, the performance of MIMO beamforming with quantized feedback in ad hoc networks is investigated. The primary findings are that a moderate number of feedback bits are sufficient to obtain significant transmission capacity gains compared to non-feedback schemes, whilst also achieving a high percentage of the transmission capacity obtained with unlimited feedback. Moreover, this achievable percentage is larger in high path loss environments, and that the number of feedback bits to maintain a fixed gain increases with the number of transmit antennas.

Finally, the impact of wireless ad hoc networks in the presence of CE errors is presented. The impact and design of two important application-dependent parameters—the effective per-node data rate and the outage constraint—for single-antenna point-to-point transmission in wireless ad hoc networks are investigated, in the presence of CE errors. In particular, for a fixed outage constraint, the optimal pilot-training length is shown to increase with the frame length according to a square root law. Consequently, for a sufficiently long coherence interval, it is shown that there is a negligible loss in transmission capacity when using this optimal pilot-training length. Moreover, for a fixed effective data rate constraint, the optimal outage constraint is shown to be quite large, and this result reveals that it is preferable to operate with a higher total data rate (aggregated across all nodes) at the expense of a lower reliability.