Adhoc Wireless Networks

The term MANET (Mobile Adhoc Network)refers to a multihop packet based wireless network composed of a set of mobile nodes that can communicate and move at the same time , without using any kind of fixed wired infrastructure. MANET are actually self organizing and adaptive networks that can be formed and deformed on-the-fly without the need of any centralized administration.

As for other packet data networks, one to-one communication in a MANET is achieved by unicast routing each single packet. Routing in MANET is challenging due to the constraints existing on the transmission bandwidth battery power and CPU time and the requirement to cope with the frequent topological changes resulting from the mobility of the nodes. Nodes of a MANET cooperate in the task of routing packets to destination nodes since each node of the network is able to communicate only with those nodes located within its transmission radius R, while the source and destination nodes can be located at a distance much higher than R.

A first attempt to cope with the mobility is to use the specific techniques aimed to tailoring the conventional routing protocols to the mobile environment while preserving their nature. For this reason the protocol designed around such techniques are referred to as table-driven or proactive protocols. To guarantee that routing tables are up to date and reflect the actual network topology, nodes running a protocol continuously exchange route updates and re calculate paths to all possible destinations. The main advantage of the proactive protocols is that a route is immediately available when is needed for data transmissions. However if the user traffic is not generated, and then resources are wasted due the proactive route update mechanism.
A different approach in the design of the routing protocol is to calculate a path only when it is necessary for data transmission. These types of protocols are called as the reactive protocols or on-demand routing protocols. A reactive protocol is characterized by a path discovery procedure and a maintenance procedure. Path discovery is based on a query reply cycle that adopts flooding of queries. The destination is eventually reached by the query an at least one reply is generated. Path discovery procedure is called when there is a need for data transmission and the source does not the path to the destination. Discovered paths are maintained by the route maintenance procedure until they are no longer in use.
The main advantage of the reactive protocol is that if data traffic is not generated by nodes, and then routing activity is totally absent. The main drawback is the network-wide path discovery required to obtain routing information. Since discovery is based on flooding, such a procedure is very costly. The main strategy to reduce the cost is to defer path discovery as much as possible so that non-optimal but available routes are preferred to the effort of finding the current best path. The natural solution is route caching.
Proactive and reactive approaches are merged in hybrid protocols that aim to combine he advantages of both approaches. The Zone routing protocol (ZRP) is a well example of a hybrid protocol. ZRP is based on the notion of the zone. Each node n is the center of a zone with radius of k hopes, denoted Zk(n).nodes at a distance greater than or equal to k from n belong to Zk(n). A reactive protocol is used by n to reach node outsides its zone .Practically k is set to value much smaller then the network diameter to get a fast convergence of proactive component of zone routing protocol. Again caching is useful in ZRP to reduce the path discoveries.
The simplest form of caching is based on timeouts associated with the cach entries. When an entry is cached a timer starts, when the timeout elapses, the entry is removed from the cache. Each time the entry is used, the timer restarts. Therefore the effectiveness of such a scheme depends upon the timeout value associated with a cached route. If the timeout is well tuned the protocol performance is increase; otherwise that will create a problem of removing the cache entry too early or too late from the cache.
A cache scheme which is based on the notion of a caching-zone whose center is a node called as the cache leader. The leader is responsible for advertising routes, detected during some route discovery, inside its caching zone.
The leader must grantee the correctness of the advertised routes. The leader monitors each advertised route proactively and it sends the control messages inside its caching zone as soon as a route becomes stale. 

Post a Comment