Accepted Papers

  • A Study on the Role of Graphs in Anomaly Detection
    Debajit Sensarma and Samar Sen Sarma,University of Calcutta,India

    Anomaly detection is a very essential problem and research has been going on in diverse areas and application domains. It is also a very challenging task because now-a-days the size of the network data are very large and also increasing day by day. While working with this large scale data, interconnection between the entities often provides additional information that may be exploited for detection of anomalous events efficiently. This type of interconnectivity and inter relation between the data can be modeled using a unified structure namely ‘graph’. This paper contains a brief overview of existing and ongoing research works on graph based anomaly detection technique.

  • Novel Channel Access Mechanishm for Pcf and Tdma Based 802.11 Networks
    A Rakesh Kumar, Bharat Electronics Ltd,India

    Channel access algorithms at MAC layer play a vital role to attain maximum achievable throughput in wireless networks. Traditionally contention based, binary exponential backoff CSMA/CA (DCF) is default channel access mechanism for new nodes not associated with existing 802.11 networks. Even though contention free and polling based PCF is illustrated in IEEE 802.11 standards, it needs to be implemented along with DCF, for simple reason clients cannot be associated to Access point in PCF mode of operation. TDMA based MAC also implements RANDOM ACCESS PHASE to add new mobile stations to network. In this paper a novel contention free channel access mechanism for 802.11infrastructure-mode network, is proposed based on perceived RSSI and SNR metrics by generating signal-strength-snr graph. Mentioned channel access algorithm breaks the conventional model of contention based network access in 802.11 PCF and TDMA networks and allows multiple parallel wireless conversations possible in same channel.

  • Optimization of Average Distance Based Self-Relocation Algorithm Using Augmented Lagrangian Method
    Shivani Dadwal and T. S. Panag,BBSBEC,India

    Mobile robots with sensors installed on them are used in wireless sensor networks to generate information about the area. These mobile robotic sensors have to relocate themselves after initial location in the field to gain maximum coverage The average distance based algorithm for relocation process of mobile sensors does not require any GPS system for tracking the robotic sensors, thus avoiding cost, but increasing energy consumption. Augmented Lagrangian method is introduced in average distance based algorithm to reduce the extra energy consumption by sensors in average distance based relocation process. This modified average distance relocation scheme also improves the coverage area and the time taken by mobile robotic sensors to come to their final positions.