Towards network coding enhanced routing protocols for group mobility

The routing strategies in mobile ad hoc networks (MANETs) have been mainly studied for entity mobility, where nodes move independently. In fact, many cooperative scenarios, especially those in military field, are better to be modeled as group mobility, in which nodes form different groups to move fo...

Full description

Saved in:
Bibliographic Details
Main Author: Xie, Lingfu
Other Authors: Chong Han Joo Peter
Format: Theses and Dissertations
Language:English
Published: 2014
Subjects:
Online Access:https://hdl.handle.net/10356/61756
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
Description
Summary:The routing strategies in mobile ad hoc networks (MANETs) have been mainly studied for entity mobility, where nodes move independently. In fact, many cooperative scenarios, especially those in military field, are better to be modeled as group mobility, in which nodes form different groups to move forward. Clearly, those strategies for entity mobility cannot perform well under group mobility. In this Thesis, we consider routing strategies for group mobility especially when any two groups are isolated most of time. Specifically, we first employ the delay/disruptive-tolerant network (DTN) routing for the packet delivery in group mobility; and then we consider a new technique called network coding to further improve the routing performance in group mobility. The Thesis begins with a brief review of existing DTN routing strategies proposed for entity mobility, particularly pointing out the pros and cons of each strategy. In fact, we find that when considering group mobility, the cons of each strategy can be effectively overcome while the pros can be preserved. First, we propose group-epidemic routing (G-ER) and leader based group routing (LBGR) to solve the resource-consuming problem in the packet replication based DTN routing. G-ER takes advantages of the cooperative nature of group mobility and incorporates two mechanisms to largely reduce the required resource such as the node buffer space and the bandwidth for the packet delivery. First, it treats each group as a single node and disseminates only one copy of one packet to each group. Second, it allows each node to share its buffer resource with other nodes in the same group. As compared to G-ER, LBGR further copes with group dynamics in group mobility by introducing the buffer resource allocation in each group. After that, we consider the mobility control based DTN routing in group mobility. We propose two mobility control techniques, the group formation enforcement and the group purposeful movement, both of which can be integrated into G-ER or LBGR to reduce the packet delivery delay. We analytically show the packet delivery delay in G-ER can be reduced by each technique. Then, we design a new DTN routing scheme called purposeful movement assisted routing (PMAR) based on the group purposeful movement, and show by simulation that PMAR improves G-ER under many practical network conditions. The second part of this Thesis is to study the new technique of network coding in wireless ad hoc networks. Our work here targets a general static network, but it can be extended to group mobility. To better benefit from network coding, we address the following two problems. We first address the re-encoding problem in network coding. We find that most proposed network coding solutions in wireless ad hoc networks may misidentify the coding opportunities when a packet is about to be re-encoded, which is detrimental to network performance. To solve this problem, we propose a principle called consistency of encoding and overhearing (CEO). The analytical and simulation results show that CEO is important to network coding to avoid performance degradation. Then, to find more coding opportunities, we address the limitation of overhearing in network coding. Overhearing is widely used to acquire the information for the packet decoding, but it only takes effect in the one-hop neighborhood of a node, which in turn restricts the identification of coding opportunities. We propose a new technique called virtual overhearing (VOH) to allow a node to obtain a packet sent by another node which can be two or more hops away. Thus, with VOH, more coding opportunities can be found. To make use of VOH, two network coding-aware routing schemes, distributed coding-aware routing with VOH (DCAR-VOH) and its enhanced version DCAR-VOH+, are proposed and studied. Both our analytical and simulation results indicate that VOH provides an effective way to discover more coding opportunities, resulting in improved network performance. Lastly, the discussion on how our studies on network coding can be applied to G-ER or MANETs is presented and elaborated.