附註:Includes bibliographical references (pages 157-163) and index.
Preface -- Acknowledgements -- 1: Introduction. 1.1. Book Outline -- 2: Mobility Management. 2.1. Background. 2.2. Mobility Management Techniques. 2.3. Profile Replicatio. 2.4. Related Work. 2.5. Summary -- 3: Off-Line Replication. 3.1. Related Work. 3.2. Optimization Objectives. 3.3. Replication for Unicast Replica Update (UR). 3.4. Replication for Multicast Replica Update (MR). 3.5. Comparison between UR and MR. 3.6. Summary -- 4: On-Line Replication. 4.1. Related Work. 4.2. Edge Algorithms. 4.3. Tree Algorithms. 4.4. Implementation Issues. 4.5. Summary -- 5: Computer Simulations. 5.1. Simulation Environment. 5.2. Off-Line Replication. 5.3. On-Line Replication. 5.4. Summary -- References -- Index.
摘要:This timely book is a comprehensive treatment of managing mobility in wireless networks. Significant new insight is also provided into solutions to file allocation problems, specifically data replication, faced in distributed database systems by Computer Scientists. Some of the solutions are applications from the general facility location problems in Operations Research. The authors thoroughly investigate replication for hierarchical mobility management where the underlying network forms a tree-like structure. Various problem formulations are considered that provide new insight into more comprehensive solutions. The off-line replication problem is reduced to problems in discrete location theory for which efficient dynamic programming solutions exist for tree networks. To solve the on-line replication problem, a unified framework is established that provides unique visual representation of the overall solution structure. This framework not only demonstrates the correspondence between two previously proposed on-line algorithms but also provides the basis for new algorithms and expands the applicability of the original problem formulation. All algorithms derived in this book have been implemented and simulated using realistic traffic models. Their performance is compared with previously proposed algorithms.
