No abstract

We consider a two user fading Multiple Access Channel with a wire-tapper (MAC-WT) where the transmitter has the channel state information (CSI) to the intended receiver but not to the eavesdropper (eve). We provide an achievable secrecy sum-rate with power control. We next provide a secrecy sum-rate with power control and cooperative jamming (CJ). We also study an achievable secrecy sum rate by employing an ON/OFF power control scheme which is more easily computable. Finally we employ CJ over this power control scheme. Results show that CJ boosts the secrecy sum-rate significantly even if we do not know the CSI of the eve's channel. At high SNR, the secrecy sum-rate (with CJ) without CSI of the eve exceeds the secrecy sum-rate (without CJ) with full CSI of the eve.

Abstruc/-A Mobile Ad-hoc'NETwork (MANET), also referred as Independent Basic Service Set (IBSS), has nodes which not only sendlreceive'packets associated with them but also co-operate in forwarding dataand control packets for the network. Efficient use of energy by the mobile nodes is very important to increase the lifetime of MANET. Since nodes seldom have packets to receive, they remain in Idle mode for long periods of time which cause a lot of energy drain. Coordination algorithms reduce the time a node is in the Idle mode by keeping it in Sleep mode as much as possible while coordinators buffer packets for these sleeping nodes.For a given transmit energy, directional antenna has larger .transmit range as compared to omni-directional antenna. This feature of directional antenna can be used in protocols which assumes a sparse network. We have chosen SPAN [ I ] to validate our above claim. SPAN is a protocol which elects a sparse network (i.e. backbone) of coordinators to buffer packets for sleeping nodes. Enhanced SPAN (E-SPAN) is the modified SPAN using directional antenna. Simulations prove the effectiveness of directional antenna and show that E-SPAN reduces average power consumed by the nodes in the network by about 20%.

No abstract

Abstracl-In mobile multimedia networks the traffic Bucluation is unpredictable and also due to limited resource availability, lhe resource allocation to multimedia applications of varying Quality of Service (QoS) requirement becomes B complex issue. This paper proposes an efficient resource allmation scheme based on resource reduction of running applications without hampering their QOS guarantee, in a single mobile cellular environment We propose a Linear Programming (LP) based resource reductiun far efficient Resource Allocation (RA). Artificial Neural Network model is used to solve lhe linear programming problem, which facilitates in real time control decision in the practical systems. The model is computationally less expensive and faster for resource allocation.The suggested scheme along with the LP-based resource reduction has shown that it is appropriate for the reduction of assigned resources for running applications during over load conditions and allocation of resources to requesting applicatioos. The simulation results for lhe proposed scheme yielded an improved resource utilization and lower percentage of rejection lo hand-off and new applications due to elfcient resource allocation.