The capacity of wireless ad hoc networks can be increased by using multiple radio channels. But due to interference the capacity is still not fully utilized. This is caused by the limited number of available radio channels. The interference problem can be reduced using directional beams instead of omni-directional beams. This paper presents a novel cross-layer approach to use multiple radio channels with directional antennas. We are using three different radio channels. Each node has three fixed directional beams having fixed beamwidth and with different radio frequency. Two nodes can communicate when both the sending and receiving beams are pointing towards each other using the same frequency channel. In this study the directions of beams cannot be changed dynamically. A modified version of Ad hoc On-demand Distance Vector (AODV) routing protocol has been used. Simulation results show that our approach outperforms other methods using three different radio channels with omni-directional antennas
Wireless ad hoc networks are self-configurable distributed systems. One of the major problems in traditional wireless ad hoc networks is interference. The interference could be reduced using smart directional antennas. In this study, multibeam smart antennas have been used. When using this type of antenna, two nodes can communicate when both the sending and receiving beams are pointing towards each other. Also, a node can only communicate with a subset of nodes in its neighborhood depending on the number of beams and their beamwidth. Thus, the network topology needs to be dynamic in this case, and by controlling the topology network, performance can be increased. In this paper, we present a framework of a cross layer approach of topology control that interacts with the routing layer and MAC layer and meets the required QoS of different data streams. The approach is fully distributed. When the network is initialized, the algorithm builds an initial connected topology and the routing algorithm uses this topology to find paths for the current communications. Then, depending on the network scenario, current communications and the required QoS, the topology control layer changes the topology to optimize the network performance. This study concerns suburban ad hoc networks (SAHN) where nodes tend to be fixed and are aware of their locations.