Research On Broadcast In Mobile Ad Hoc Networks

Broadcasting in Ad Hoc network is not only a method of notification of messages, but also an important channel for route discovery. Currently, in the absence of node location information, broadcasting algorithms in Ad Hoc network can not avoid redundant broadcast. When algorithms get location information, most of them use the size of additional coverage by assessing broadcasting called EAC to determine whether to broadcast messages. This method has high computational burden, and in dense network, it is difficult to estimate the range of additional coverage after broadcast, what's more, the algorithms have weak ability of adapting networks.Counter-based broadcasting algorithm decides whether to forward packets by contrasting the threshold and size of duplicate messages that a node receives, so that this algorithm does not consider the impact of additional coverage area affected by distance between nodes. In order to better reflect counter value of networks, this paper proposed counter broadcasting algorithm based on the continuous increment function, when a node receives a duplicate packet, this algorithm yields the value according to the counter increment function generator to count values. Then determines to broadcast packets by the counter value that considers whether it is greater than the threshold; at the same time, the back off process using way of retreat between MAC layer and network layer reduces packet collision and the time delay.Location-based broadcasting algorithms and algorithms based on covering all angles need to calculate the available additional coverage areas called EAC after receiving duplicate broadcasting packets. It is a large amount of calculation in those algorithms, and no reduction in the maximum number of redundant broadcasting nodes. Therefore, this paper presents self-positioning role of broadcasting algorithm based on location information. This algorithm divides each node into six fan-shaped areas, and considers the middle point as the expected forward node. After a node sends a broadcasting message, this node which receives the message determines the location relationships together with the sending node coordinates, then compare it with the expected forward node in this area. If the node is closed to the expected forward node, broadcast the packet, or else do not need to broadcast packets. In order to prevent multiple nodes within the same region broadcasting messages in the same time, I use the strategy of" mark to wait for further notice". This algorithm can determine whether packets should be broadcast, so it reduces the time delay, and redundant broadcast nodes are greatly reduced through the "mark to wait for further notice" policy.This paper proposed a broadcast algorithm that considered the cost of clustering. The algorithm uses heuristic clustering method based on minimum ID to divide clusters. Clustering process also considered the factor of stability, and adding node into cluster that belongs to the nearest cluster head. By introducing cluster head counter, records and limits the number that cluster is selected as cluster head, this algorithm solves the problem of load imbalance based on the minimum ID heuristic clustering algorithm; to minimize gateway nodes between two clusters in the process of selecting gateways after clustering, this paper proposed "gateway node statement-cluster head license" mechanism. In order to reduce the cluster maintenance process, this paper proposed the concept of sub cluster head that when a cluster needs to re-cluster, firstly considers the competition qualification of sub cluster head, if it deserved, directly elected it as cluster head, or else continued to compete. This approach greatly reduces the cost of cluster maintenance. Broadcasting in the clustering algorithm considering the cost because there is not so many gateways between clusters, thus the number of redundant broadcasting nodes are reduced. Broadcasting in the backbone network composed of cluster heads and gateways reduces the time delay.The algorithms in this paper are all checked out under NS2-2.31 simulation. The simulation carried out in a number of scenarios, this paper compared and analyzed in aspects of average delay, arrival rate and forward rate with the existing algorithms and the proposed method. Results show that the proposed broadcasting algorithms can reach a higher rate, and reduce the average delay, what's more, adaptation of networks has greatly improved than existing algorithms.