Part I Phase Noise Estimation in OFDM and OFDMA Uplink Communications
In this part, several phase noise models and the corresponding effects in OFDM and OFDMA systems are introduced. Among them, for the oscillator phase noise, we discussed the estimation of Wiener phase noise and stationary phase noise. Finally two multiuser phase noise estimation algorithms to mitigate the effects of multiple phase noise in uplink OFDMA systems are proposed. The LS approach provides acceptable performance with low complexity while the ML approach considers the second order statistics of the ICI to enhance the performance. The proposed schemes aim to compensate for CPE, the major effect of phase noise for medium to low phase noise levels where phase noise correction is applicable. Moreover, these two multiuser phase noise correction schemes are stable within a wide range of phase noise levels and applicable to any subcarrier assignment scheme, which shows its potential in practical applications.
Part II Characterizing The Wireless Ad Hoc Networks by Using The Distance
Distributions
To achieve pervasive computing in the absence of the existing infrastructure, nodes in the service area organize themselves into a wireless ad hoc network. Due to the random locations of the nodes, the distances between nodes are random. In this part, two distance distributions are presented and used as the foundations to characterize the organized wireless ad hoc networks. Given the prior knowledge of the order of the nearest neighbors, the nearest neighbor probability distribution in the theorem provides us how to use the optimum transmitting range to connect to the k-th nearest neighbor or, equivalently, how to power-efficiently deploy a k-connected wireless ad hoc network. Based on the marginal and the joint distribution of the distances between nodes and a RN in the theorem, we analytically show that the exact node degree of the wireless ad hoc network in the shadow fading environment is a binomial distribution.
With the exact distribution of node degree, we further obtain the probability of the minimum node degree of the wireless ad hoc network that is the necessary condition of the network connectivity. Our results also show that the connectivity of the
organized wireless ad hoc network in the shadow fading environment is improved due to the random fluctuation of signal strength.
Part III On The Distance Distributions of The Wireless Ad Hoc Networks
This part investigates the probability distributions of the random separation distances between node pairs in the wireless ad hoc networks. By using the concept of the Euclidean distance in the 2-dimensional Euclidean space, the first probability distribution, the distribution of the Euclidean distance to the k-th nearest neighbor, is obtained. Using the technique of function of random variables, we obtain the probability distribution of the Euclidean distance between two random selected wireless nodes. Then, through the computation of the union area on the node coverage area and a unit square, we derived the marginal cdf and pdf of a wireless node pair.
Since the Euclidean distances between wireless node pairs with the common reference wireless node are mutually independent, we can easily extend the marginal cdf and pdf to obtain the joint cdf and pdf of the Euclidean distances between wireless node pairs in the wireless ad hoc network with N wireless nodes that are randomly and uniformly distributed over a unit square.
Part IV Organizing an Optimal Cluster-Based Ad Hoc Network Architecture by the
Modified Quine-McCluskey Algorithm
To reduce the waste of precious bandwidth and the limited battery power in exchanging the cluster maintenance overheads, we propose a distributed Modified Quine-McCluskey (MQM) algorithm to organize the wireless ad hoc network into an optimal cluster-based network architecture that requires the minimum number of cluster maintenance overheads. Simulation results show that by minimizing the number of generated clusters and the variance of cluster members, the organized cluster-based network architecture requires the minimum number of cluster maintenance overheads. Thus, the optimal cluster-based network architecture is organized.
Part V A Clustering Algorithm to Produce Power-Efficient Architecture for
(N,B)-Connected Ad Hoc Networks
Through analyses, we find that reduction of the number of cluster maintenance overheads for a (N,B)-connected cluster-based wireless ad hoc network can be achieved by reducing the number of generated cluster and the variance of the number of the cluster members. By analyzing the cluster architectures generated by the ID-based and Degree-based clustering algorithms, we find that the number of generated cluster and the variance of the number cluster members can be reduced by reducing the number of orphan clusters generated by boundary nodes. Simulation results show that by assigning critical nodes the highest weights (or priorities) to be selected as CHs, the number of generated clusters and the variance of the number of cluster members for the cluster-based network architecture generated by the proposed DCA/CNF based approaches are reduced. As a consequence, the number of cluster maintenance overheads is reduced and the organized network architecture is power efficient.