Abstract
Broadcast storm occurs in radio-based wireless networks when a node attempts to broadcast a message to other nodes. All the 1-hop neighbors within the radio coverage range will relay the message by rebroadcasting it. Since radio signal is broadcast in nature, the rebroadcast communications can easily interfere with each other, causing the rebroadcast message to corrupt. The problem becomes especially serious in wireless sensor networks which often require very dense deployment. Unfortunately, most previous solutions assumed a perfect radio condition with a fixed, circular coverage region. However, in real practice, radio signals are very dynamic and irregular in their coverage. In this study, we consider dense wireless sensor networks with irregular radio coverage. The idea is to first find the rim neighbors, the neighbors on the rim of the irregular radio coverage region, according to signal strength. The forwarder set is then selected from the rim neighbors to cover all the 2-hop neighbors. Unlike previous works, the proposed algorithm, RIM, does not maintain neighbor sets, because neighbor sets will be very dynamic under irregular radio coverage while the maintenance overhead will be very high for dense sensor networks. Through simulation, we compare the proposed RIM algorithm with Double-Covered Broadcast (DCB) [10]. The results show that under dense deployment and irregular radio coverage, our algorithm performs better than DCB.