Group Member:
Qi Difeng
Neda Edalat
Bao Haitao
Link
(1). Proposal (Qi Difeng, Neda Edalat, Bao Haitao)
(2). Related Work (Bao Haitao)
(3). Cell Formulate algorithm (Qi Difeng)
(4). In-Cell Network algorithm (Bao Haitao)
Performance(5). Performance (Neda Edalat)
Useful
Introduction
Routing protocols in mobile ad hoc network have been widely researched, yet the nodes in network are usually assumed to move with low probability. It is not true in some application such as battlefield network, navy fleet network and robot sensor network. We are designing a vehicle network which can connect all the taxi in Singapore together. Drivers or taxis can communicate with each other without using expensive cellular networks. Instead, they can relay for each other to send to packet to the destination. This is an extreme dynamic mobile ad hoc network. Nodes in this network moves so frequently that the topology of the network is changing every second. Traditional routing protocols cannot handle this situation.
We propose the Cell-Based Routing algorithm here to reduce the latency of the network, while keep the overhead low. We believe that in a mobile ad hoc network, the routing task should not relay on a single node since every node is moving. Instead, a cluster of nodes should cooperate to maintain the routing information. We assign a cluster according to the relative position of each node, i.e. all the nodes in a specific cell will composite a cluster. If other nodes moves in this cell, they will join in this cluster; on the other hand, if a node leaves the cell, it will leave that cluster too. While nodes in the network are always moving fast, the position of cell moves slowly, so the neighbor relationship of cells will keep for a long time. We can see that it is the same of Ad Hoc network if we regard a cell as a node. As a result, we can use traditional reactive routing algorithm such as DSR to route between cells.
Fig.1 Downtown of Singapore